Computer Graphics (CS602)

VU

Computer Graphics
CS602

Virtual University of Pakistan

Knowledge beyond the boundaries

1 Copyright Virtual University of Pakistan

Computer Graphics (CS602)

VU

Table of Contents
Lecture No.1 Lecture No.2 Lecture No.3 Lecture No.4 Lecture No.5 Lecture No.6 Lecture No.7 Lecture No.8 Lecture No.9 Lecture No.10 Lecture No.11 Lecture No.12 Lecture No.13 Lecture No.14 Lecture No.15 Lecture No.16 Lecture No.17 Lecture No.18 Lecture No.19 Lecture No.20 Lecture No.21 Lecture No.22 Lecture No.23 Lecture No.24 Lecture No.25 Lecture No.26 Lecture No.27 Lecture No.28 Lecture No.29 Lecture No.30 Lecture No.31 Lecture No.32 Lecture No.33 Lecture No.34 Lecture No.35 Lecture No.36 Lecture No.37 Lecture No.38 Lecture No.39 Lecture No.40 Lecture No.41 Lecture No.42 Lecture No.43 Lecture No.44 Lecture No.45 Introduction to Computer Graphics....................................................3 Graphics Systems I ..............................................................................25 Graphics Systems II ............................................................................34 Point ......................................................................................................43 Line Drawing Techniques ...................................................................53 Circle Drawing Techniques ................................................................59 Ellipse and Other Curves....................................................................64 Filled-Area Primitives-I ......................................................................78 Filled-Area Primitives-II ....................................................................97 Mathematics Fundamentals .............................................................101 2D Transformations I ........................................................................113 2D Transformations II ......................................................................117 Drawing Example ..............................................................................126 Clipping-I ...........................................................................................136 Clipping-II ..........................................................................................141 3D Concepts .......................................................................................152 3D Transformations I ........................................................................168 3D Transformations II ......................................................................175 Projections ..........................................................................................188 Perspective Projection .......................................................................195 Triangles and Planes .........................................................................203 Triangle Rasterization ......................................................................211 Lighting I ............................................................................................222 Lighting II ..........................................................................................229 Mathematics of Lighting and Shading Part I .................................234 Mathematics of Lighting and Shading Part II Light Types and Shading Models..................................................................................238 Review II ............................................................................................242 Review III ...........................................................................................261 Mathematics of Lighting and Shading Part III ..............................276 Mathematics of Lighting and Shading Part IV ..............................285 Mathematics of Lighting and Shading Part V ................................291 Introduction to OpenGL ...................................................................296 OpenGL Programming - I ................................................................303 OpenGL Programming - II ..............................................................309 Curves .................................................................................................320 Space Curves ......................................................................................326 The Tangent Vector...........................................................................329 Bezier Curves .....................................................................................333 Building Polygonal Models of Surfaces ...........................................340 Fractals ...............................................................................................347 Viewing ...............................................................................................366 Examples of Composing Several Transformations ........................387 Real-World and OpenGL Lighting..................................................393 Evaluators, curves and Surfaces ......................................................413 Animations .........................................................................................423 2 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Lecture No.1
1.1 Definition

Introduction to Computer Graphics

Computers accept process, transform and present information. Computer Graphics involves technology to accept, process, transform and present information in a visual form that also concerns with producing images (or animations) using a computer. 1.2 Why Study Computer Graphics? There are certain important reasons to study computer graphics. We will discuss them under certain heads: Visualization I like to see what I am doing. Many a times it happens that you perform certain tasks which you cannot visualize; for example as a student of data structures, you implement trees, graphs and other Abstract Data Types (ADTs) but you cannot visualize them whereas you must be having an inner quest to see what these actually look like. I like to show people what I am doing. Similarly at certain times you would be performing certain tasks which you know but it would be difficult for others to understand them so there is very important requirement of showing the things in order to make them understandable. Graphics is interesting We are visual creatures and for us a picture is worth a thousand words. If we can get rid of text based static screen and get some graphics; its always interesting to see things in colours and motion on the screen. Therefore graphics is interesting because it involves simulation, algorithm, and architecture. Requirement Well there are certain areas which require use of computer graphics heavily. One example is drawing of machines. It is required to prepare drawing of a machine before the actual production. The other heavy requirement is for architects as they have to prepare a complete blue print of the building they have to build long before the actual construction work gets underway. AutoCAD and other applications of the kind are heavily used today for building architecture. Entertainment Merely a couple of decades back, the idea of a 24 hours Cartoons Network was really a far fetched one. That was the time when one would wait for a whole week long before getting an entertainment of mere 15 minutes. Well thanks to computer graphics that have enabled us to entertain ourselves with animated movies, cartoons etc.

3 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

1.3 Some History The term computer graphics was coined in 1960 by William Fetter to describe the new design methods that he was developing at Boeing. He created a series of widely reproduced images on a plotter exploring cockpit design using a 3D model of a human body.

Whirlwind: early graphics using Vector Scope (1951) Spacewars: first computer graphics game (MIT 1961)

4 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

First CAD system (IBM 1959)

First bump-mapped images (Blinn 1978)

Early texture-mapped image (Catmull 1974)

5 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

First distributed ray traced image (Cook 1984)

First ray traced image (Whitted 1980) 1.4 Graphics Applications Due to rapid growth in the field of computing, now computer is used as an economical and efficient tool for the production of pictures. Computer graphics applications are found in almost all areas. Here we will discuss some of the important areas including: i. User Interfaces ii. Layout and Design iii. Scientific Visualization and Analysis iv. Art and Design v. Medicine and Virtual Surgery vi. Layout Design & Architectural Simulations vii. History and cultural heritage viii. Entertainment ix. Simulations x. Games 6 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

User Interfaces Almost all the software packages provide a graphical interface. A major component of graphical interface is a window manager that allows a user to display multiple windows like areas on the screen at the same time. Each window can contain a different process that can contain graphical or non-graphical display. In order to make a particular window active, we simply have to click in that window using an interactive pointing device. Graphical Interface also includes menus and icons for fast selection of programs, processing operations or parameter values. An icon is a graphical symbol that is designed to look like the processing option it represents.

B205 Control Console (1960)

Impressive and Interactive 3D environment 7 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

3D Studio MAX

Layout and Design

8 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Scientific Visualization and Analysis Computer graphics is very helpful in producing graphical representations for scientific visualization and analysis especially in the field of engineering and medicine. It helps a lot in drawing charts and creating models.

9 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

ART AND DESIGN Computer graphics is widely used in Fine Arts as well as commercial arts for producing better as well as cost effective pictures. Artists use a variety of programs in their work, provided by computer graphics. Some of the most frequently used packages include: Artists paintbrush Pixel paint Super paint

10 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Medicine and Virtual Surgery Computer graphics has extensive use in tomography and simulations of operations. Tomography is the technique that allows cross-sectional views of physiological systems in X-rays photography. Moreover, recent advancement is to make model and study physical functions to design artificial limbs and even plan and practice surgery. Computer-aided surgery is currently a hot topic.

11 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Room Layout Design and Architectural Simulations

12 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Layout Design & Architectural Simulations

13 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

History and cultural heritage Another important application of computer graphics is in the field of history and cultural heritage. A lot of work is done in this area to preserve history and cultural heritage. The features so for provide are: Innovative graphics presentations developed for cultural heritage applications Interactive computer techniques for education in art history and archeology New analytical tools designed for art historians Computer simulations of different classes of artistic media 14 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

15 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Movies Computer graphics methods are now commonly used in making motions pictures, music videos and television shows. Sometimes the graphics scenes are displayed by themselves and sometimes graphics objects are combined with the actors and live scenes. A number of hit movies and shows are made using computer graphics technology. Some of them are: Star Trek- The Wrath of Khan Deep Space Nine Stay Tuned Reds Dreams Shes Mad

Tron (1980) First time computer graphics were used for live action sequences.

16 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Fully computer generated animated features

17 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Star Wars (1977)

18 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Star Trek II: The Wrath of Khan, genesis

The Last Starfighter (15 minutes) (1982) The Last Starfighter (15 minutes) (1982)

19 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Special Effects in Live Action Cinema

20 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Traditional Animated Features Some examples: Automating Keyframing in many Disney-type animations The flocking behaviour of the wild beast in Lion King Non photorealistic rendering: 3D effects in Futurama

21 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Behind the scenes on Antz Production

Number of frames in the movie Number of times the movie was rendered during production Number of feet of approved animation produced in a week Total number of hours of rendering per week Average size of the frame rendered Total number of Silicon Graphics servers used for rendering Number of desktop systems used in production Total Number of processors used for rendering Average amount of memory per processor Time it would have taken to render this movie on 1 processor Amount of storage required for the movie Amount of frames kept online at any given time Time to re-film out final cut beginning to end

119,592 15 (approx.) 107 ft. 275,000 hrs. 6 MB 270 166 700 256 MB 54 yrs., 222 days, 15 mins., 36 3.2 TB 75000 frames 41.5 days (997 hrs.)

22 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Simulations Simulation by all means is a very helpful tool to show the idea you have or the work you are doing or to see the results of your work. Given below is the picture in which you can see waves ripples on water; no doubt looking like original but is simply a simulation. A number of software packages are used for simulation including: Crackerjack Computer Skills Keen Artistic Eye Flash Maya

Game Thanks to computer graphics, real time games are now possible. Now game programming itself has become an independent field and game programmers are in high demand. Some of the famous games are: Quake Dooms Need For Speed Commandos 23 Copyright Virtual University of Pakistan

1-Introduction to Computer Graphics

VU

Related Disciplines

Data Processing Computer Vision DATA Computer Graphics IMAGES Image Processing

Interdisciplinary Science Physics: light, color, appearance, behavior Mathematics: Curves and Surfaces, Geometry and Perspective Engineering Hardware: graphics media and processors, input and output devices Software: graphics libraries, window systems Art, Perception and Esthetics Color, Composition, Lighting, Realism

24 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

Lecture No.2

Graphics Systems I

Introduction of Graphics Systems With the massive development in the field of computer graphics a broad range of graphics hardware and software systems is available. Graphics capabilities for both twodimensional and three-dimensional applications are now common on general-purpose computers, including many hand-held calculators. On personal computers there is usage of a variety of interactive input devices and graphics software packages; whereas, for higher-quality applications some special-purpose graphics hardware systems and technologies are employed. VIDEO DISPLAY DEVICES The primary output device in a graphics system is a video monitor. The operation of most video monitors is based on the standard cathode-ray-tube (CRT) design, but several other technologies exist and solid-state monitors may eventually predominate. Refresh Cathode-Ray Tubes Following figures illustrate the basic operation of a CRT. A beam of electrons (cathode rays) emitted by an electron gun, passes through focusing and deflection systems that direct the beam toward specified positions on the phosphor-coated screen. The phosphor then emits a small spot of light at each position contacted by the electron beam. The light emitted by the phosphor fades very rapidly therefore to keep the picture it is necessary to keep the phosphor glowing. This is achieved through redrawing the picture repeatedly by quickly directing the electron beam back over the same points and the display using this technique is called refresh CRT. The primary components of an electron gun in a CRT are the heated metal cathode and a control grid. Heat is supplied to the cathode by directing a current through filament (a coil of wire), inside the cylindrical cathode structure. Heating causes electrons to be boiled off the hot cathode surface. In the vacuum inside the CRT envelope, the free, negatively charged electrons are then accelerated toward the phosphor coating by a high positive voltage.

25 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

The accelerating voltage can be generated with a positively charged metal coating on the inside of the CRT envelope near the phosphor screen an accelerating anode can be used.

Intensity of the electron beam is controlled by setting voltage levels on the control grid, a metal cylinder that fits over the cathode. A high negative voltage applied to the control grid will shut off the beam by repelling electrons and stopping them from passing through the small hole at the end of the control grid structure. A smaller negative voltage on the control grid simply decreases the number of electrons striking the phosphor coating on the screen. It is the responsibility of focusing system to converge electron beam to a small spot where it strikes the phosphor. Otherwise the electrons will repel each other and the beam would disperse. This focusing is achieved through electric or magnetic fields. In electrostatic focusing the electron beam passes through a positively charged metal cylinder that forms an electrostatic lens. Then electrostatic lens focuses the electron beam at the center of the screen. Similar task can be achieved with a magnetic field setup by a coil mounted around the outside of the CRT envelope. Magnetic lens focusing produces the smallest spot size on the screen and is used in special purpose devices. The distance that the electron beam must travel from gun to the exact location of the screen that is small spot is different from the distance to the center of the screen in most CRTs because of the curvature therefore some additional focusing hardware is required in high precision systems to take beam to all positions of the screen. This procedure is achieved in two steps in first step beam is conveyed through the exact center of the screen 26 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

and then additional focusing system adjust the focusing according to the screen position of the beam. Cathode-ray tubes are now commonly constructed with magnetic deflection coils mounted on the outside of the CRT envelope. Two pairs of coils are used, with the coils in each pair mounted on opposite sides of the neck of the CRT envelope. One pair is mounted on the top and bottom of the neck and the other pair is mounted on opposite sides of the neck. The magnetic field produced by each pair of coils results in a traverse deflection force that is perpendicular both to the direction of the magnetic field and to the direction of travel of the electron beam. Horizontal deflection is achieved with one pair of coils, and vertical deflection by the other pair. The proper deflection amounts are attained by adjusting the current through the coils. When electrostatic deflection is used, two pairs of parallel plates are mounted inside the CRT envelope. One pair of plates is mounted horizontally to control the vertical deflection, and the other pair is mounted vertical to control horizontal deflection.

Phosphor is available in different kinds. One variety is available in color but a major issue is their persistence. Persistence is defined as the time it takes the emitted light from the phosphor to decay to one-tenth of its original intensity. Lower persistence phosphors require higher refresh rates to maintain a picture on the screen without flicker. A phosphor with low persistence is useful for displaying highly complex, static pictures. Monitors normally come with persistence in the range from 10 to 60 microseconds. The maximum number of points (that can be uniquely identified) on a CRT is referred to as the resolution. A more precise definition of resolution is the number of points per centimeter that can be plotted horizontally and vertically, although it is often simply stated as the total number of points in each direction. Finally aspect ratio; is the ratio of vertical points to horizontal points necessary to produce equal-length lines in both directions on the screen. An aspect ratio of 3/4 means that a vertical line plotted with three points has the same length as a horizontal line plotted with four points. RASTER-SCAN SYSTEMS Raster scan is the most common type of monitors using CRT. In raster scan picture is stored in the area called refresh buffer or frame buffer. First of all why information is stored; because picture have to be refreshed again and again for this very reason it is stored. Second is how it is stored; so picture is stored in a two dimensional matrix where each element corresponds to each pixel on the screen. If there arise a question what is a pixel? The very simple answer is a pixel (short for picture element) represents the shortest 27 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

possible unique position/ element that can be displayed on the monitor without overlapping. The frame buffer stores information in a two dimensional matrix; the question is that how many bits are required for each pixel or element. If there is black and white picture then there is only one bit required to store 0 for black or 1 for white and in this case buffer will be referred as bitmap. In colour pictures obviously multiple bits are required for each pixel position depending on the possible number of colours for example to show 256 colours 8 bits will be required for each pixel and in case if multiple bits are used for one pixel frame buffer will be referred as pixmap. Now with the information in frame buffer, let us see how an image is drawn. The drawing is done in a line-by-line fashion. After drawing each line from left to right it reaches at the left end of the next line to draw next line; which is called horizontal retrace. Similarly after completing all lines in horizontal fashion it again reaches the top left corner to start redrawing the image (that is for refreshing) and this is called vertical retrace. Normally each vertical retrace takes 1/60th of a second to avoid flickering. There are two further methods to scan the image: interlaced and non-interlaced. In interlaced display beam completes scanning in two passes. In one pass only odd lines are drawn and in the second pass even lines are drawn. Interlacing provides effect of double refresh rate by completing half of the lines in half of the time. Therefore, in systems with low refresh rates interlacing helps avoid flickering.

RANDOM-SCAN Displays In random-scan displays a portion of the screen can be displayed. Random-scan displays draw a picture one line at a time and are also called vector displays (or stroke-writing or calligraphic displays). In these systems image consists of a set of line drawing commands referred to as Refresh Display File. Random-scan can refresh the screen in any fashion by repeating line drawing mechanism. Random-scan displays are designed to draw all the component lines of a picture 30 to 60 times each second. High-quality vector systems are capable of handling approximately 100,000 short lines at this refresh rate. When a small set of lines is to be displayed, each refresh cycle is delayed to avoid refresh rates greater than 60 frames per second. Otherwise, faster refreshing of the set of lines could burn out the phosphor. Random-scan displays are designed for line-drawing applications and cannot display complex pictures. The lines drawn in vector displays are smoother whereas in raster-scan lines often become jagged. Color CRT Monitors A CRT monitor displays colour pictures by using a combination of phosphors that emit different coloured light. With the combination of phosphor a range of colours can be displayed. There are two techniques used in colour CRT monitors: Beam Penetration Method Shadow Mask Method 28 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

In beam penetration method two layers of phosphor, usually coated onto the inside of the CRT screen, and the displayed colour depend on how far the electron beam penetrates into the phosphor layers. At intermediate beam speeds, combinations of red and green light are emitted to show two additional colours, orange and yellow. Beam penetration is an inexpensive way to produce colours as only a few colours are possible and the quality of picture is also not impressive. Shadow mask methods can display a wide range of colours. In this technique each pixel position is made up of three phosphor dots called triads as shown in the following figure. Three phosphor dots have different colors i.e. red, green and blue and the display colour is made by the combination of all three dots. Three guns are used to throw beam at the three dots of the same pixel. By varying intensity at each dot a wide range of colours can be generated. A shadow-mask is used which has holes aligned with the dots so that each gun can fire beam to corresponding dot only. CRT Displays Advantages Fast response (high resolution possible) Full colour (large modulation depth of E-beam) Saturated and natural colours Inexpensive, matured technology Wide angle, high contrast and brightness Disadvantages Large and heavy (typ. 70x70 cm, 15 kg) High power consumption (typ. 140W) Harmful DC and AC electric and magnetic fields Flickering at 50-80 Hz (no memory effect) Geometrical errors at edges Direct View Storage Devices A direct view storage tube stores the picture information as a charge distribution just behind the phosphor-coated screen. Two electron guns are used in this system as shown in the following figure. They are: Primary Gun Flood Gun

Primary gun is used to store the picture pattern whereas flood gun maintains the picture display. 29 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

DVST has advantage that no refresh is required so very complex pictures can be displayed at very high resolutions without flicker. Whereas, it has disadvantage that ordinarily no colors can be displayed and that selected parts of a picture cannot be erased. To eliminate a picture section, the entire screen must be erased and the modified picture redrawn. The erasing and redrawing process can take several seconds for a complex picture. Flat-Panel Displays This is emerging technology slowly replacing CRT monitors. The flat-panel displays have following properties: Little Volume Light Weight Lesser Power consumption

Flat panels are used in calculators, pocket video games and laptop computers. There are two categories of flat panel displays: Emissive Display (Plasma Panels) Non-Emissive Display (Liquid Crystal Display) The emissive displays (emitters) are devices that convert electrical energy into light. Plasma panels, thin-film electro-luminescent displays, and light-emitting diodes are examples of emissive displays. Non-emissive displays (non-emitters) use optical effects to convert sunlight or light from some other source into graphics patterns. The most important example of a non-emissive flat-panel display is a liquid-crystal device. Plasma-panel Displays Plasma panels also called gas-discharge displays are constructed by filling the region between two glass plates with a mixture of gases that usually includes neon. A series of vertical conducting ribbons is placed on one glass panel, and a set of horizontal ribbons is built into the other glass panel. Firing voltages applied to a pair of horizontal and vertical conductors cause the gas at the intersection of the two conductors to break down into glowing plasma of electrons and ions. Picture definition is stored in a refresh buffer, and the firing voltages are applied to refresh the pixel positions 60 times per second. Advantages Large viewing angle 30 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

Good for large-format displays Fairly bright Disadvantages Expensive Large pixels (~1 mm versus ~0.2 mm) Phosphors gradually deplete Less bright as compared to CRTs, using more power Liquid Crystal Displays Liquid crystal refers to the fact that these compounds have a crystalline arrangement of molecules, yet they flow like a liquid. Flat panel displays use nematic liquid crystal, as demonstrated in the following figures. Two glass plates, each containing a light polarizer at right angles to the other plate, sandwich the liquid-crystal material. Rows of horizontal transparent conductors are built into one glass plate, and columns of vertical conductors are put into the other plate. The intersection of two conductors defines a pixel position. Polarized light passing through the material is twisted so that it will pass through the opposite polarizer. The light is then reflected back to the viewer. To turn off the pixel, we apply a voltage to the two intersecting conductors to align the molecules so that the light is not twisted. LCD Displays Advantages Small footprint (approx 1/6 of CRT) Light weight (typ. 1/5 of CRT) Low power consumption (typ. 1/4 of CRT) Completely flat screen - no geometrical errors Crisp pictures - digital and uniform colours No electromagnetic emission Fully digital signal processing possible Large screens (>20 inch) on desktops Disadvantages High price (presently 3x CRT) Poor viewing angle (typ. +/- 50 degrees) Low contrast and luminance (typ. 1:100) Low luminance (Natural light) (typ. 200 cd/m2) 31 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

Three-Dimensional Viewing Devices Graphics monitors for the display of three-dimensional scenes have been devised using a technique that reflects a CRT image from a vibrating, flexible mirror. In this system when varifocal mirror vibrates it changes focal length. These vibrations are synchronized with the display of an object on a CRT so that each point on the object is reflected from the mirror into spatial position corresponding to the distance of that point from a specified viewing position. This allows user to walk around an object or scene and view it from different sides. Virtual Reality Devices Virtual reality system enables users to move and react in a computer-simulated environment. Various types of devices allow users to sense and manipulate virtual objects much as they would real objects. This natural style of interaction gives participants the feeling of being immersed in the simulated world. Virtual reality simulations differ from other computer simulations in that they require special interface devices that transmit the sights, sounds, and sensations of the simulated world to the user. These devices also record and send the speech and movements of the participants to the simulation program. To see in the virtual world, the user wears a head-mounted display (HMD) with screens directed at each eye. The HMD contains a position tracker to monitor the location of the user's head and the direction in which the user is looking. Using this information, a computer recalculates images of the virtual world to match the direction in which the user is looking and displays these images on the HMD. Users hear sounds in the virtual world through earphones in the HMD. The hepatic interface, which relays the sense of touch and other physical sensations in the virtual world, is the least developed feature. Currently, with the use of a glove and position tracker, the user can reach into the virtual world and handle objects but cannot actually feel them. 32 Copyright Virtual University of Pakistan

2-Graphics Systems I

VU

Another interesting simulation is interactive walk through. A sensing system in the headset keeps track of the viewers opposition, so that the front and back of objects can be seen as the viewer walks and interacts with the displays. Similarly given below is a figure using a headset and a data glove worn on the right hand?

33 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

Lecture No.3
Raster-Scan Systems

Graphics Systems II

Interactive raster graphics systems typically employ several processing units. In addition to the CPU, a special purpose processor, called the video controller or display controller is used to control the operation of the display device. Organization of a simple raster system is shown in following figure. Here the frame buffer can be anywhere in the system memory, and the video controller accesses the frame buffer to refresh the screen.

Architecture of a simple raster graphics system In addition to the video controller more sophisticated raster systems employ other processors as coprocessors and accelerators to implement various graphics operations. Video Controller Following figure shows a commonly used organization for raster systems. A fixed area of the system memory is reserved for the frame buffer, and the video controller is given direct access to the frame-buffer memory. Frame-buffer locations, and the corresponding screen positions, are referenced in Cartesian coordinates.

34 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

Architecture of a raster system with a fixed portion of a system memory reserved for the frame buffer. In the following figure the basic refresh operations of the video controller are diagrammed. Two registers are used to store the coordinates of the screen pixels. Initially, the x register is set to 0 and the y register is set to ymax. The value stored in the frame buffer for this pixel position is then retrieved and used to set the intensity of the CRT beam. Then the x register is incremented by 1, and the process repeated for the next pixel on the top scan line. This procedure is repeated for each pixel along the next line by resetting x register to 0 and decrementing the y register by 1. Pixels along this scan line are then processed in turn, and the procedure is repeated for each successive scan line. After cycling through all pixels along the bottom scan line y=0, the video controller resets to the first pixel position on the top scan line and the refresh process starts over.
Raster Scan Generator

Horizontal and Vertical Deflection Voltages

Register X

Register Y

Memory Addresses

Pixel Register

Intensity

Frame Buffer
Basic Video Controller Refresh Operations Since the screen must be refreshed at the rate of 60 frames per second, the simple procedure illustrated in above figure cannot be accommodated by typical RAM chips. The cycle time is too large making the process very slow. To speed up pixel processing, video controllers can retrieve multiple pixel values from the refresh buffer on each pass. 35 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

The multiple pixel intensities are then stored in a separate register and used to control the CRT beam intensity for a group of adjacent pixels. When that group of pixels has been processed, the next block of pixel values is retrieved from the frame buffer. Raster Scan Display Processor Following figure shows one way to setup the organization of a raster system containing a separate display processor, sometimes referred to as a graphics controller or a display coprocessor. The purpose of the display processor is to free the CPU from the graphics chores. In addition to the system memory, a separate display processor memory area can also be provided. A major task of the display processor is digitizing a picture definition given in an application program into a set of pixel-intensity values for storage in the frame buffer. This digitization process is called scan conversion.
Display Processor Memory Frame Buffer Video Controller Monitor

CPU

Display Processor

System Memory

System Bus

I/O Devices

Architecture of a raster graphics system with a display processor Raster-Scan Characters Graphics commands specifying straight lines and other geometric objects are scan converted into a set of discrete intensity points. Scan converting a straight-line segment, for example, means that we have to locate the pixel positions closest to the line path and store the intensity for each position in the frame buffer. Similar methods are used for scan converting curved lines and polygon outlines. Characters can be defined with rectangular grids, as shown in following figure, or they can be defined with curved outlines shown in the right hand side figure given below. The array size for character grids can vary from about 5 by 7 to 9 by 12 or more for higherquality displays. A character grid is displayed by superimposing the rectangular grid pattern into the frame buffer at a specified coordinate position. With characters that are defined as curve outlines, character shapes are scan converted into the frame buffer.

36 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

Defined as a grid of pixel positions

Defined as a curve outline

Random-Scan Systems The organization of a simple random scan system is shown in following figure. An application program is input and stored in the system memory along with a graphics package. Graphics commands in the application program are translated by the graphics package into a display file stored in the system memory. This display file is then accessed by the display processor to refresh the screen. The display processor cycles through each command in the display file program once during every refresh cycle. Sometimes the display processor in a random scan system is referred to as a display processing unit or graphics controller.

Architecture of a simple random scan system Graphics Card or Display Adapters A video card is typically an adapter, a removable expansion card in the PC. Thus, it can be replaced! A video display adapter which is the special printed circuit board that plugs into one of the several expansion slots present on the mother board of the computer. A video display adapter is simply referred as a video card. The video card can also be an integral part of the system board; this is the case in certain brands of PCs and is always the case in laptops and clear preference for the replaceable video card in some PCs. 37 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

A number of display adapters are available with varying capabilities especially Intel systems support following adapters: Monochrome Adapter (MA) Hercules Adapter (HA) Color Graphics Adapter (CGA) Enhanced Graphics Adapter (EGA) Multicolor Graphics Adapter (MCGA) Video Graphics Adapter (VGA) Super Video Graphics Adapter (SVGA) Extended Graphics Adapter (XGA)

Monochrome Adapter The simplest and the first available adapter is MA. This adapter can display only text in single color and has no graphics displaying capability. Originally this drawback only prevented the users from playing video games, but today, even the most serious business software uses graphics and color to great advantage. Hence, MA is no longer suitable, though it offers clarity and high resolution. Hercules Adapter The Hercules card emulates the monochrome adapter but also operates in a graphics mode. Having graphics capabilities the Hercules card became somewhat of a standard for monochrome systems. Color Graphics Adapter This adapter can display text as well as graphics. In text mode it operates in 25 rows by 80 column mode with 16 colors. In graphics mode two resolutions are available: Medium resolution graphics mode 320 * 200 with 4 colors available from palette of 16 colors and 640 * 200 with 2 colors

One drawback of CGA card is that it produces flicker and snow. Flicker is the annoying tendency of the text to flash as it moves up or down. Snow is the flurry of bright dots that can appear anywhere on the screen. Enhanced Graphics Adapter The EGA was introduced by IBM in 1984 as alternative to CGA card. The EGA could emulate most of the functions and all the display modes of CGA and MA. The EGA offered high resolution and was not plagued with the snow and flicker problems of CGA. In addition EGA is designed to use the enhanced color monitor capable of displaying 640 * 350 in 16 colors from a palette of 64. The EGA card has several internal registers. A serious limitation of the EGA card is that it supports write operations to most of its internal registers, but no read operation. The result is it is not possible for software to detect and preserve the state of the adapter, which makes EGA unsuited for memory resident application or for multitasking like windows and OS/2.

38 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

Multicolor Graphics Adapter The MCGA was designed to emulate the CGA card and to maintain compatibility with all the CGA modes. In addition to the text and graphics modes of the CGA, MCGA has two new graphics modes: 640 * 480 with 2 colors 320 * 200 in with 256 colors Video Graphics Adapter The VGA supports all the display modes of MA, CGA and MCGA. In addition VGA supports a graphics mode of 640 * 480 with 16 colors. Super Video Graphics Adapter The SVGA designation refers to enhancements to the VGA standard by independent vendors. Unlike display adapters discussed earlier SVGA does not refer to a card that meets a particular specification but to a group of cards that have different capabilities. For example one card may have resolutions 800 * 600 and 1024 * 768, whereas, another card may have same resolution but more colors. These cards have different capabilities, but still both of them are classified as SVGA. Since each SVGA card has different capabilities, you need special device driver programs for driving them. This means that unlike VGA cards which can have a single driver that works with all VGA cards, regardless of the vendor, each SVGA card must have a corresponding driver. Extended Graphics Adapter The XGA evolved from the VGA and provides greater resolution, more colors and much better performance. The XGA has a graphics processor bus mastering. Being a bus master adapter means that the XGA can take control of the system as though it were the mother board. In essence, a bus master is an adapter of the mother board. The XGA offers 2 new modes: 640 * 480 with 16 bit colors (65536 colors) 1024 * 768 with 8 bit colors (256 colors) Video Card Supports the CPU The video card provides a support function for the CPU. It is a processor like the CPU. However it is especially designed to control screen images.

39 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

RAM on the Video Card Video cards always have a certain amount of RAM. This RAM is also called the frame buffer. Today video cards hold plenty of RAM, but earlier it was more important: How much RAM? That is significant for color depth at the highest resolutions. Which type of RAM? This is significant for card speed. Video card RAM is necessary to keep the entire screen image in memory. The CPU sends its data to the video card. The video processor forms a picture of the screen image and stores it in the frame buffer. This picture is a large bit map. It is used to continually update the screen image. 3D - lots of RAM Supporting the demand for high quality 3D performance many new cards come with a frame buffer of 16 or 32 MB RAM and they use the AGP interface for better bandwidth and access to the main memory. VRAM Briefly, in principle all common RAM types can be used on the video card. Most cards use very fast editions of ordinary RAM (SDRAM or DDR). Some high end cards (like Matrox Millennium II) earlier used special VRAM (Video RAM) chips. This was a RAM type, which only was used on video cards. In principle, a VRAM cell is made up of two ordinary RAM cells, which are "glued" together. Therefore, you use twice as much RAM than otherwise. VRAM also costs twice as much. The smart feature is that the double cell allows the video processor to simultaneously read old and write new data on the same RAM address. Thus, VRAM has two gates which can be active at the same time. Therefore, it works significantly faster. With VRAM you will not gain speed improvements increasing the amount of RAM on the graphics controller. VRAM is already capable of reading and writing simultaneously due to the dual port design. UMA and DVMT On some older motherboards the video controller was integrated. Using SMBA (Shared Memory Buffer Architecture) or UMA (Unified Memory Architecture) in which parts of the system RAM were allocated and used as frame buffer. But sharing the memory was very slow and the standards never became very popular. A newer version of this is found in Intel chip set 810 and the better 815, which also integrates the graphics controller and use parts of the system RAM as frame buffer. Here the system is called Dynamic Video Memory Technology (D.V.M.T.). The RAMDAC All traditional graphics cards have a RAMDAC chip converting the signals from digital to analog form. CRT monitors work on analog signals. The PC works with digital data which are sent to the graphics adapter. Before these signals are sent to the monitor they have to be converted into analog output and this is processed in the RAMDAC: 40 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

The recommendation on a good RAMDAC goes like this: External chip, not integrated in the VGA chip Clock speed: 250 - 360 MHz. Heavy Data Transport The original VGA cards were said to be "flat." They were unintelligent. They received signals and data from the CPU and forwarded them to the screen, nothing else. The CPU had to make all necessary calculations to create the screen image. As each screen image was a large bit map, the CPU had to move a lot of data from RAM to the video card for each new screen image. The graphic interfaces, like Windows, gained popularity in the early nineties. That marked the end of the "flat" VGA cards. The PC became incredibly slow, when the CPU had to use all its energy to produce screen images. You can try to calculate the required amount of data. A screen image in 1024 x 768 in 16 bit color is a 1.5 MB bit map. That is calculated as 1024 x 768 x 2 bytes. Each image change (with a refresh rate of 75 HZ there is 75 of them each second) requires the movement of 1.5 MB data. That zaps the PC energy, especially when we talk about games with continual image changes. Furthermore, screen data have to be moved across the I/O bus. In the early nineties, we did not have the PCI and AGP buses, which could move large volumes of data. The transfer took place through the ISA bus, which has a very limited width. Additionally the CPUs were 386s and early 486s, which also had limited power. Accelerator Cards In the early nineties the accelerator video cards appeared. Today all cards are accelerated and they are connected to the CPU through high speed buses like PCI and AGP. With accelerated video chips, Windows (and with that the CPU) need not calculate and design the entire bit map from image to image. The video card is programmed to draw lines, Windows and other image elements. The CPU can, in a brief code, transmit which image elements have changed since the last transmission. This saves the CPU a lot of work in creating screen images. The video chip set carries the heavy load: All video cards are connected to the PCI or the AGP bus, this way providing maximum data transmission. The AGP bus is an expanded and improved version of the PCI bus used for video cards only. 41 Copyright Virtual University of Pakistan

3-Graphics Systems II

VU

Modern video cards made for 3D gaming use expensive high-end RAM to secure a sufficient bandwidth. If you for example want to see a game in a resolution of 1280 x 1024 at 80 Hz, you may need to move 400 MB of data each second - that is quite a lot. The calculation goes like this: 1280 X 1024 pixels x 32 bit (color depth) x 80 = 419,430,400 bytes 419,430,400 bytes = 409,600 kilobytes = 400 megabytes.

Graphics Libraries Graphics developers some time use 2D or 3D libraries to create graphics rapidly and efficiently. These developers include game developers, animators, designers etc. The following libraries are commonly used among developers: FastGL OpenGL DirectX Others Advantages of Graphics Libraries These libraries help developers to create fast and optimized animations and also help to access features that are available in video hardware. Hardware manufacturers give support in hardware for libraries. Famous manufacturers include SIS, NVIDIA, ATI, INTEL etc. Graphics Software There is a lot of 2D and 3D software available in the market. These software provide visual interface for creation of 2D and 3D animation / models image creation. These tools are under use of movie makers, professional animators and designers. These tools are flash, Maya, 3D studio max, adobe photo shop, CorelDraw, image viewer, paintbrush etc.

42 Copyright Virtual University of Pakistan

4-Point

VU

Lecture No.4

Point

Pixel: The smallest dot illuminated that can be seen on screen. Picture: Composition of pixels makes picture that forms on whole screen Resolution We know that Graphics images on the screen are built up from tiny dots called picture elements or pixels. The display resolution is defined by the number of rows from top to bottom, and number of pixels from left to right on each scan line. Since each mode uses a particular resolution. For example mode 19 uses a resolution of 200 scan lines, each containing 320 pixels across. This is often referred to as 320*200 resolution. In general, higher the resolution, more pleasing is the picture. Higher resolution means a sharper, clearer picture, with less pronounced staircase effect on lines drawn diagonally and better looking text characters. On the other hand, higher resolution also means more memory requirement for the display. Text and Graphics Modes We discussed different video hardware devices that include VGA cards and monitors. Video cards are responsible to send picture data to monitor each time it refresh itself. Video cards support both different text and graphics modes. Modes consist of their own refresh rate, number of colors and resolutions (number of rows multiply by number of columns). The following famous video modes that we can set in todays VGA cards on different refresh rate: 25 * 80 with 16 colors support (text mode) 320 * 200 with 8 bit colors support (graphics mode) 640 * 480 with 16 colors support (graphics mode) 640 * 480 with 8, 16, 24, 32 bit color support (graphics mode) 800 * 600 with 8, 16, 24, 32 bit color support (graphics mode) Text and Graphics All modes are fundamentally of two types, text or graphics. Some modes display only text and some are made only for graphics. As seen earlier, the display adapter continuously dumps the contents of the VDU (video display unit) memory on the screen. The amount of memory required representing a character on screen in text mode and a pixel in graphics mode varies from mode to mode. Mode No. 3 6 7 18 19 Type Text Graphics Text Graphics Graphics Resolution 80 x 25 640 x 200 80 x 25 640 x 480 320 x 200 Memory Required 2 bytes per char 1 bit per pixel 2 bytes per char 1 bit per pixel 1 byte per pixel 43 Copyright Virtual University of Pakistan

4-Point

VU

In mode 6 each pixel displayed on the screen occupies one bit in VDU memory. Since this bit can take only two values, either 0 or 1, only two colors can be used with each pixel. How text displays As seen previously text modes need two bytes in VDU memory to represent one character on screen; of these two bytes, the first byte contains the ASCII value of the character being displayed, whereas the second byte is the attribute byte. The attribute byte controls the color in which the character is being displayed. The ASCII value present in VDU memory must be translated into a character and drawn on the screen. This drawing is done by a character generator this is part of the display adapter or in VBIOS. The CGA has a character generator that uses 8 scan lines and 8 pixels in each of these scan lines to produce a character on screen; whereas the MAs character generator uses 9 scan lines and 14 pixels in each of these scan lines to produce a character. This larger format of MA makes the characters generated by MA much sharper and hence easier to read. On older display adapters like MA and CGA, the character generator is located in ROM (Read Only Memory). EGA and VGA do not have a character generator ROM. Instead, character generator data is loaded into plane 2 of display RAM. This feature makes it easy for custom character set to be loaded. Multiple character sets (up to 4 for EGA and up to 8 for VGA) may reside in RAM simultaneously. A set of BIOS services is available for easy loading of character sets. Each character set can contain 256 characters. Either one or two character sets may be active giving these adapters on the screen simultaneously. When two character sets are active, a bit in each character attribute byte selects which character set will be used for that character. Using a ROM-BIOS service we can select the active character set. Each character in the standard character set provided with the EGA is 8 pixels wide and 14 pixels tall. Since VGA has higher resolution, it provides a 9 pixel wide by 16 pixels tall character set. Custom character set can also be loaded using BIOS VDU services. The graphics modes can also display characters, but they are produced quite differently. The graphics modes can only store information bit by bit. The big advantage of this method is that you design characters of desired style, shape and size. Text mode colors In mode 3, for each character on screen there are two bytes in VDU memory, one containing the ACCII value of the character and other containing its attribute. The attribute byte controls the color of the character. The attribute byte contains three components: the foreground color (color of the character itself), the background color (color of the area not covered by the character) and the blinking component of the character. The next slide shows the breakup of the attribute byte.

44 Copyright Virtual University of Pakistan

4-Point

VU

7 X X X X X X X 1

Bits 6 x x x x x x 1 x

5 x x x x x 1 x x

x x x x 1 x x x

4 x x x 1 x x x x

3 x x 1 x x x x x

2 x 1 x x x x x x

1 1 x x x x x x x

Purpose Blue component of foreground color Green component of foreground color Red component of foreground color Intensity component of foreground color Blue component of background color Green component of background color Red component of background color Blinking component

Graphics Mode colors So far we have seen how to set color in text modes. Setting color in graphics modes is quite different. In the graphics mode each pixel on the screen has a color associated with it. There are important differences here as compared to setting color in text mode. First, the pixels cannot blink. Second, each pixel is a discrete dot of color, there is no foreground and background. Each pixel is simply one color or another. The number of colors that each adapter can support and the way each adapter generates these colors is drastically different. But we will only discuss here colors in VGA. Colors in VGA IBM first introduced the VGA card in April 1987. VGA has 4 color planes red, green, blue and intensity, with one bit from each of these planes contributing towards 1 pixel value. There are lots of ways that you can write pixel on screen. You can write pixel on screen by using one of the following methods: Using video bios services to write pixel Accessing memory and registers directly to write pixel on screen. Using library functions to write pixel on screen Practical approach to write pixel on screen As we have discussed three ways to write pixel on screen. Here we will discuss all these ways practically and see how the pixel is displayed on screen. For that we will have to write code in Assembly and C languages. So get ready with these languages Writing pixel Using Video BIOS The following steps are involved to write pixel using video BIOS services. Setting desired video mode Using bios service to set color of a screen pixel Calling bios interrupt to execute the process of writing pixel.

Source code Below are the three lines written in assembly language that can set graphics mode 19(13h). You can use this for assembler or you can embed this code in C language using asm keyword 45 Copyright Virtual University of Pakistan

4-Point

VU

MOV AH,0 MOV AL,13h ;;mode number from 0-19 INT 10H To insert in C language above code will be inserted with key word asm and curly braces. asm{ MOV AH,0 MOV AL,13h ;;mode number from 0-19 INT 10H } Description Line #1: mov ah,0 is the service number for setting video mode that is in register ah Line #2: mov al,13h is the mode number that is in register al Line #3: int 10h is the video bios interrupt number that will set mode 13h Source code for writing pixel The following code can be used to write pixel using video bios interrupt 10h and service number 0ch. MOV AH,0Ch MOV AL,COLOR_NUM MOV BH,0 MOV CX,ROW_NUM MOV DX,COLUMN_NUM INT 10h Description Line#1: service number in register Ah Line#2: color value, since it is 13h mode so it has 0-255 colors range. You can assign any color number from 0 to 255 to all register. Color will be selected from default palette setting against the number you have used. Line#3: page number in Bh register. This mode supports only one page. So 0 is used in Bh register. 0 mean default page. Line#4: column number will be used in CX register Line#5: row number will be used in DX register Line#6: BIOS interrupt number 10h

Writing pixel by accessing memory directly So far we used BIOS to draw pixel. Here we will draw pixel by accessing direct pointer to the video memory and write color value. The following steps are involved to write direct pixel without using BIOS: Set video mode by using video BIOS routine as discussed earlier Set any pointer to the video graphics memory address 0x0A0000. Now write any color value in the video memory addressing

46 Copyright Virtual University of Pakistan

4-Point

VU

Direct Graphics Memory Access Code Mov ax,0a000h Mov ds,ax ;;segment address changed Mov si,10 ;; column number Mov [si],COLOR_NUM Work to do: Write pixel at 12th row and 15th column Hint: use formula (row * 320 + column) in si register. Writing character directly on screen You can also write direct text by setting any text mode using BIOS service and then setting direct pointer at text memory address 0x0b8000. Example Set mode Number 3. using BIOS service and then use this code to write character Mov ax,0b8000h Mov ds,ax Mov si,10 Mov [si],a

;;column number ;;character to write

Using Library functions While working in C language, you can use graphics library functions to write pixel on screen. These graphics library functions then use BIOS routines or use direct memory access drivers to draw pixel on screen. initgraph(&gdriver, &gmode, ""); /* read result of initialization */ errorcode = graphresult(); if (errorcode != grOk) /* an error occurred */ { printf("Graphics error: %s\n", getch()); /* return with error code */ } /* draw a pixel on 10th row and 10 column */ putpixel(10, 10, BLUE); /* clean up */ closegraph();

exit(1);

Steps in C language First call Initgraph() function and then call putpixel() function to draw pixel on screen. It takes row, column and color value as parameters. after drawing pixel use closegraph() function to close the graphics routines provided by built in driver by Borland.

47 Copyright Virtual University of Pakistan

4-Point

VU

Discussion on pixel drawing methods BIOS routines are standard routines built in VGA cards but these routines are very much slow. You will use pixel to draw filled triangle, rectangles and circles and these all will be much slower than direct memory access method. Direct memory access method allows you to write pixel directly by passing the complex BIOS routines. It is easy and faster but its programming is only convenient in mode 13h. Library functions are easier to use and even faster because these are optimized and provided with special drivers by different companies. Drawing pixel in Microsoft Windows So far we have been discussing writing pixel in DOS. Here we will discuss briefly how to write pixel in Microsoft Windows. Microsoft windows are a complete graphical operating system but it does not allow you to access BIOS or direct memory easily. It provides library functions (APIs) that can be used to write graphics. By working in graphics in windows one must have knowledge about Windows GDI (graphics device interface) system. Windows GDI functions Here are some windows GDI functions that can be used to draw pixel e.g SetPixel and SetPixelV. Both are used to draw pixel on screen. The example and source code of writing pixel in windows will be available. Window Code Example: // a.cpp : Defines the entry point for the application. // #include "stdafx.h" #include "resource.h" #define MAX_LOADSTRING 100 // Global Variables: HINSTANCE hInst; instance TCHAR szTitle[MAX_LOADSTRING]; // The title bar text TCHAR szWindowClass[MAX_LOADSTRING]; // The title bar text

//

current

// Foward declarations of functions included in this code module: ATOM MyRegisterClass(HINSTANCE hInstance); BOOL InitInstance(HINSTANCE, int); LRESULT CALLBACK WndProc(HWND, UINT, WPARAM, LPARAM); LRESULT CALLBACK About(HWND, UINT, WPARAM, LPARAM); int APIENTRY WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, LPSTR lpCmdLine, int nCmdShow) 48 Copyright Virtual University of Pakistan

4-Point

VU

{ // TODO: Place code here. MSG msg; HACCEL hAccelTable; // Initialize global strings LoadString(hInstance, IDS_APP_TITLE, szTitle, MAX_LOADSTRING); LoadString(hInstance, IDC_A, szWindowClass, MAX_LOADSTRING); MyRegisterClass(hInstance); // Perform application initialization: if (!InitInstance (hInstance, nCmdShow)) { return FALSE; } hAccelTable = LoadAccelerators(hInstance, (LPCTSTR)IDC_A); // Main message loop: while (GetMessage(&msg, NULL, 0, 0)) { if (!TranslateAccelerator(msg.hwnd, hAccelTable, &msg)) { TranslateMessage(&msg); DispatchMessage(&msg); } } return msg.wParam; } // FUNCTION: MyRegisterClass() // // PURPOSE: Registers the window class. // // COMMENTS: // // This function and its usage is only necessary if you want this code // to be compatible with Win32 systems prior to the 'RegisterClassEx' // function that was added to Windows 95. It is important to call this function // so that the application will get 'well formed' small icons associated // with it. // ATOM MyRegisterClass(HINSTANCE hInstance) { WNDCLASSEX wcex; wcex.cbSize = sizeof(WNDCLASSEX); wcex.style = CS_HREDRAW | CS_VREDRAW; 49 Copyright Virtual University of Pakistan

4-Point

VU

wcex.lpfnWndProc wcex.cbClsExtra wcex.cbWndExtra wcex.hInstance wcex.hIcon wcex.hCursor wcex.hbrBackground wcex.lpszMenuName wcex.lpszClassName wcex.hIconSm

= (WNDPROC)WndProc; = 0; = 0; = hInstance; = LoadIcon(hInstance, (LPCTSTR)IDI_A); = LoadCursor(NULL, IDC_ARROW); = (HBRUSH)(COLOR_WINDOW+1); = (LPCSTR)IDC_A; = szWindowClass; = LoadIcon(wcex.hInstance, (LPCTSTR)IDI_SMALL);

return RegisterClassEx(&wcex); } // // FUNCTION: InitInstance(HANDLE, int) // // PURPOSE: Saves instance handle and creates main window // // COMMENTS: // // In this function, we save the instance handle in a global variable and // create and display the main program window. // BOOL InitInstance(HINSTANCE hInstance, int nCmdShow) { HWND hWnd; hInst = hInstance; // Store instance handle in our global variable hWnd = CreateWindow(szWindowClass, szTitle, WS_OVERLAPPEDWINDOW, CW_USEDEFAULT, 0, CW_USEDEFAULT, 0, NULL, NULL, hInstance, NULL); if (!hWnd) { return FALSE; } ShowWindow(hWnd, nCmdShow); UpdateWindow(hWnd); return TRUE; } // // FUNCTION: WndProc(HWND, unsigned, WORD, LONG) // // PURPOSE: Processes messages for the main window. // // WM_COMMAND - process the application menu 50 Copyright Virtual University of Pakistan

4-Point

VU

// WM_PAINT - Paint the main window // WM_DESTROY - post a quit message and return // // LRESULT CALLBACK WndProc(HWND hWnd, UINT message, WPARAM wParam, LPARAM lParam) { int wmId, wmEvent; PAINTSTRUCT ps; HDC hdc; TCHAR szHello[MAX_LOADSTRING]; LoadString(hInst, IDS_HELLO, szHello, MAX_LOADSTRING); switch (message) { case WM_COMMAND: wmId = LOWORD(wParam); wmEvent = HIWORD(wParam); // Parse the menu selections: switch (wmId) { case IDM_ABOUT: DialogBox(hInst, (LPCTSTR)IDD_ABOUTBOX, hWnd, (DLGPROC)About); break; case IDM_EXIT: DestroyWindow(hWnd); break; default: return DefWindowProc(hWnd, message, wParam, lParam); } break; case WM_PAINT: { hdc = BeginPaint(hWnd, &ps); // TODO: Add any drawing code here... RECT rt; GetClientRect(hWnd, &rt); int j=0; //To draw some pixels of RED colour on the screen for(int i=0;i<100;i++) { SetPixel(hdc,i+j,10,RGB(255,0,0)); j+=6; } EndPaint(hWnd, &ps); } break; 51 Copyright Virtual University of Pakistan

4-Point

VU

case WM_DESTROY: PostQuitMessage(0); break; default: return DefWindowProc(hWnd, message, wParam, lParam); } return 0; } // Mesage handler for about box. LRESULT CALLBACK About(HWND hDlg, UINT message, WPARAM wParam, LPARAM lParam) { switch (message) { case WM_INITDIALOG: return TRUE; case WM_COMMAND: if (LOWORD(wParam) == IDOK || LOWORD(wParam) == IDCANCEL) { EndDialog(hDlg, LOWORD(wParam)); return TRUE; } break; } return FALSE; }

52 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

Lecture No.5

Line Drawing Techniques

Line A line, or straight line, is, roughly speaking, an (infinitely) thin, (infinitely) long, straight geometrical object, i.e. a curve that is long and straight. Given two points, in Euclidean geometry, one can always find exactly one line that passes through the two points; this line provides the shortest connection between the points and is called a straight line. Three or more points that lie on the same line are called collinear. Two different lines can intersect in at most one point; whereas two different planes can intersect in at most one line. This intuitive concept of a line can be formalized in various ways. A line may have three forms with respect to slope i.e. it may have slope = 1 as shown in following figure (a), or may have slope < 1 as shown in figure (b) or it may have slope > 1 as shown in figure (c). Now if a line has slope = 1 it is very easy to draw the line by simply starting form one point and go on incrementing the x and y coordinates till they reach the second point. So that is a simple case but if slope < 1 or is > 1 then there will be some problem.

figure (a)

figure (b)

figure (c)

There are three techniques to be discussed to draw a line involving different time complexities that will be discussed later. These techniques are: Incremental line algorithm DDA line algorithm Bresenham line algorithm

Incremental line algorithm This algorithm exploits simple line equation y = m x + b Where m = dy / dx and b=ymx Now check if |m| < 1 then starting at the first point, simply increment x by 1 (unit increment) till it reaches ending point; whereas calculate y point by the equation for each x and conversely if |m|>1 then increment y by 1 till it reaches ending point; whereas calculate x point corresponding to each y, by the equation. Now before moving ahead let us discuss why these two cases are tested. First if |m| is less than 1 then it means that for every subsequent pixel on the line there will be unit increment in x direction and there will be less than 1 increment in y direction and vice versa for slope greater than 1. Let us clarify this with the help of an example: 53 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

Suppose a line has two points p1 (10, 10) and p2 (20, 18) Now difference between y coordinates that is dy = y2 y1 = 18 10 = 8 Whereas difference between x coordinates is dx = x2 x1 = 20 10 = 10 This means that there will be 10 pixels on the line in which for x-axis there will be distance of 1 between each pixel and for y-axis the distance will be 0.8. Consider the case of another line with points p1 (10, 10) and p2 (16, 20) Now difference between y coordinates that is dy = y2 y1 = 20 10 = 10 Whereas difference between x coordinates is dx = x2 x1 = 16 10 = 6 This means that there will be 10 pixels on the line in which for x-axis there will be distance of 0.6 between each pixel and for y-axis the distance will be 1. Now having discussed this concept at length let us learns the algorithm to draw a line using above technique, called incremental line algorithm: Incremental_Line (Point p1, Point p2) dx = p2.x p1.x dy = p2.y p1.y m = dy / dx x = p1.x y = p1.y b=ym*x if |m| < 1 for counter = p1.x to p2.x drawPixel (x, y) x=x+1 y=m*x+b else for counter = p1.y to p2.y drawPixel (x, y) y=y+1 x=(yb)/m Discussion on algorithm: Well above algorithm is quite simple and easy but firstly it involves lot of mathematical calculations that is for calculating coordinate using equation each time secondly it works only in incremental direction. We have another algorithm that works fine in all directions and involving less calculation mostly only addition; which will be discussed in next topic. Digital Differential Analyzer (DDA) Algorithm: DDA abbreviated for digital differential analyzer has very simple technique. Find difference dx and dy between x coordinates and y coordinates respectively ending points of a line. If |dx| is greater than |dy|, than |dx| will be step and otherwise |dy| will be step. if |dx|>|dy| then step = |dx| else 54 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

step = |dy|

Now very simple to say that step is the total number of pixel required for a line. Next step is to divide dx and dy by step to get xIncrement and yIncrement that is the increment required in each step to find next pixel value. xIncrement = dx/step yIncrement = dy/step Next a loop is required that will run step times. In the loop drawPixel and add xIncrement in x1 by and yIncrement in y1. To sum-up all above in the algorithm, we will get, DDA_Line (Point p1, Point p2) dx = p2.x p1. x dy = p2.y p1. y x1=p1.x y1=p1.y if |dx|>|dy| then step = |dx| else step = |dy| xIncrement = dx/step yIncrement = dy/step for counter = 1 to step drawPixel (x1, y1) x1 = x1 + xIncrement y1 = y1 + yIncrement Criticism on Algorithm: There is serious criticism on the algorithm that is use of floating point calculation. They say that when we have to draw points that should have integers as coordinates then why to use floating point calculation, which requires more space as well as they have more computational cost. Therefore there is need to develop an algorithm which would be based on integer type calculations. Therefore, work is done and finally we will come up with an algorithm Bresenham Line Drawing algorithm which will be discussed next. Bresenham's Line Algorithm Bresenham's algorithm finds the closest integer coordinates to the 55 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

actual line, using only integer math. Assuming that the slope is positive and less than 1, moving 1 step in the x direction, y either stays the same, or increases by 1. A decision function is required to resolve this choice. If the current point is (xi, yi), the next point can be either (xi+1,yi) or (xi+1,yi+1) . The actual position on the line is (xi+1, m(xi+1)+c) . Calculating the distance between the true point, and the two alternative pixel positions available gives: d1 d2 = = = = y - yi m * (x+1)+b-yi yi + 1 - y yi + 1 m ( xi + 1 ) - b

Let us magically define a decision function p, to determine which distance is closer to the true point. By taking the difference between the distances, the decision function will be positive if d1 is larger, and negative otherwise. A positive scaling factor is added to ensure that no division is necessary, and only integer math need be used. pi = pi = pi = pi = where k=2 dy + dx (2b-1) dx (d1-d2) dx (2m * (xi+1) + 2b 2yi-1 ) 2 dy (xi+1) 2 dx yi + dx (2b-1 ) ------------------ (i) 2 dy xi 2 dx yi + k ------------------ (ii)

Then we can calculate pi+1 in terms of pi without any xi , yi or k . pi+1 = 2 dy xi+1 2 dx yi+1 + k = 2 dy (xi + 1) - 2 dx yi+1 + k since xi+1= xi + 1 pi+1 pi+1 = 2 dy xi + 2 dy- 2 dx yi+1 + k ------------------ (iii) Now subtracting (ii) from (iii), we get pi+1 - pi = 2 dy - 2 dx (yi+1 - yi ) pi+1 = pi + 2 dy - 2 dx (yi+1 - yi ) If the next point is: (xi+1,yi) then d1<d2 => => => d1-d2<0 pi<0 pi+1= pi + 2 dy

If the next point is: (xi+1,yi+1) then d1>d2 => => => d1-d2>0 pi>0 pi+1= pi + 2 dy - 2 dx

The pi is our decision variable, and calculated using integer arithmetic from pre-computed constants and its previous value. Now a question is remaining how to calculate initial value of pi. For that use equation (i) and put values (x1, y1) pi = 2 dy (x1+1) 2 dx yi + dx (2b-1 ) where b = y m x implies that 56 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

pi pi pi

= = =

2 dy x1 +2 dy 2 dx yi + dx ( 2 (y1 mx1) -1 ) 2 dy x1 +2 dy 2 dx yi + 2 dx y1 2 dy x1 - dx 2 dy x1 +2 dy 2 dx yi + 2 dx y1 2 dy x1 - dx

there are certain figures will cancel each other shown in same different colour pi = 2 dy - dx

Thus Bresenham's line drawing algorithm is as follows: dx = x2-x1 dy = y2-y1 p = 2dy-dx c1 = 2dy c2 = 2(dy-dx) x = x1 y = y1 plot (x,y,colour) while (x < x2 ) x++; if (p < 0) p = p + c1 else p = p + c2 y++ plot (x,y,colour) Again, this algorithm can be easily generalized to other arrangements of the end points of the line segment, and for different ranges of the slope of the line. Improving performance Several techniques can be used to improve the performance of line-drawing procedures. These are important because line drawing is one of the fundamental primitives used by most of the other rendering applications. An improvement in the speed of line-drawing will result in an overall improvement of most graphical applications. Removing procedure calls using macros or inline code can produce improvements. Unrolling loops also may produce longer pieces of code, but these may run faster. The use of separate x and y coordinates can be discarded in favour of direct frame buffer addressing. Most algorithms can be adapted to calculate only the initial frame buffer address corresponding to the starting point and to replaced: X++ with Addr++ Y++ with Addr+=XResolution Fixed point representation allows a method for performing calculations using only integer arithmetic, but still obtaining the accuracy of floating point values. In fixed point, the fraction part of a value is stored separately, in another integer: M = Mint.Mfrac 57 Copyright Virtual University of Pakistan

5-Line Drawing Techniques

VU

Mint = Mfrac =

Int(M) Frac(M) MaxInt

Addition in fixed point representation occurs by adding fractional and integer components separately, and only transferring any carry-over from the fractional result to the integer result. The sequence could be implemented using the following two integer additions: ADD Yfrac,Mfrac ; ADC Yint,Mint Improved versions of these algorithms exist. For example the following variations exist on Bresenham's original algorithm: Symmetry (forward and backward simultaneously) Segmentation (divide into smaller identical segments - GCD(D x,D y) ) Double step, triple step, n step. Setting a Pixel Initial Task: Turning on a pixel (loading the frame buffer/bit-map). Assume the simplest case, i.e., an 8-bit, non-interlaced graphics system. Then each byte in the frame buffer corresponds to a pixel in the output display.

To find the address of a particular pixel (X,Y) we use the following formula: addr(X, Y) = addr(0,0) + Y rows * (Xm + 1) + X (all in bytes) addr(X,Y) = the memory address of pixel (X,Y) addr(0,0) = the memory address of the initial pixel (0,0) Number of rows = number of raster lines. Number of columns = number of pixels/raster line. Example: For a system with 640 480 pixel resolution, find the address of pixel X = 340, Y = 150 addr(340, 150) = addr(0,0) + 150 * 640 (bytes/row) + 340 = base + 96,340 is the byte location Graphics system usually have a command such as set_pixel (x, y) where x, y are integers.

58 Copyright Virtual University of Pakistan

6-Circle Drawing Techniques

VU

Lecture No.6
Circle

Circle Drawing Techniques

A circle is the set of points in a plane that are equidistant from a given point O. The distance r from the center is called the radius, and the point O is called the center. Twice the radius is . The angle a circle subtends known as the diameter from its center is a full angle, equal to 360 or radians. A circle has the maximum possible area for a given perimeter, and the minimum possible perimeter for a given area. The perimeter C of a circle is called the circumference, and is given by C=2r

Circle Drawing Techniques First of all for circle drawing we need to know what the input will be. Well the input will be one center point (x, y) and radius r. Therefore, using these two inputs there are a number of ways to draw a circle. They involve understanding level very simple to complex and reversely time complexity inefficient to efficient. We see them one by one giving comparative study. Circle drawing using Cartesian coordinates This technique uses the equation for a circle on radius r centered at (0, 0) given as: x2 + y2 = r2, an obvious choice is to plot y=

Obviously in most of the cases the circle is not centered at (0, 0), rather there is a center point (xc, yc); other than (0, 0). Therefore the equation of the circle having center at point (xc, yc): (x- xc) 2 + (y-yc)2 = r2, this implies that , y = yc Using above equation a circle can easily be drawn. The value of x varies from r-xc to r+xc. and y will be calculated using above formula. Using this technique a simple algorithm will be: Circle1 (xcenter, ycenter, radius) for x = radius - xcenter to radius + xcenter 59 Copyright Virtual University of Pakistan

6-Circle Drawing Techniques

VU

y = xc + drawPixel (x, y) y = xc drawPixel (x, y) This works, but is inefficient because of the multiplications and square root operations. It also creates large gaps in the circle for values of x close to r as shown in the above figure.

Circle drawing using Polar coordinates A better approach, to eliminate unequal spacing as shown in above figure is to calculate points along the circular boundary using polar coordinates r and . Expressing the circle equation in parametric polar form yields the pair of equations x = xc + r cos y = yc + r sin Using above equation circle can be plotted by calculating x and y coordinates as takes values from 0 to 360 degrees or 0 to 2. The step size chosen for depends on the application and the display device. Larger angular separations along the circumference can be connected with straight-line segments to approximate the circular path. For a more continuous boundary on a raster display, we can set the step size at 1/r. This plots pixel positions that are approximately one unit apart. Now let us see how this technique can be sum up in algorithmic form. Circle2 (xcenter, ycenter, radius) for = 0 to 2 step 1/r x = xc + r * cos y = yc + r * sin drawPixel (x, y) Again this is very simple technique and also solves problem of unequal space but unfortunately this technique is still inefficient in terms of calculations involves especially floating point calculations. Calculations can be reduced by considering the symmetry of circles. The shape of circle is similar in each quadrant. We can generate the circle section in the second quadrant of the xy-plane by noting that the two circle sections are symmetric with respect to the y axis and circle sections in the third an fourth quadrants can be obtained from sections in the first and second quadrants by considering symmetry about the x axis. We can take this one step further and note that there is also symmetry between octants. Circle sections in adjacent octants within one quadrant are symmetric with respect to the 45o line dividing the two octants. These symmetry conditions are illustrated in above figure. 60 Copyright Virtual University of Pakistan

6-Circle Drawing Techniques

VU

Therefore above algorithm can be optimized by using symmetric octants. Lets see: Circle2 (xcenter, ycenter, radius) for = 0 to / 4 step 1/r x = xc + r * cos y = yc + r * sin DrawSymmetricPoints (xcenter, ycenter, x, y) DrawSymmeticPoints (xcenter, ycenter, x, y) Plot ( x + xcenter, y + ycenter ) Plot ( y + xcenter, x + ycenter ) Plot ( y + xcenter, -x + ycenter ) Plot ( x + xcenter, -y + ycenter ) Plot ( -x + xcenter, -y + ycenter) Plot ( -y + xcenter, -x + ycenter) Plot ( -y + xcenter, x + ycenter) Plot ( -x + xcenter, y + ycenter)

Hence we have reduced half the calculations by considering symmetric octants of the circle but as we discussed earlier inefficiency is still there and that is due to the use of floating point calculations. In next algorithm we will try to remove this problem. Midpoint circle Algorithm As in the Bresenham line drawing algorithm we derive a decision parameter that helps us to determine whether or not to increment in the y coordinate against increment of x coordinate or vice versa for slope > 1. Similarly here we will try to derive decision parameter which can give us closest pixel position. 2 2 2 Let us consider only the first octant of a circle of x +y -r =0 x=y radius r centred on the origin. We begin by plotting point (r, 0) and end when x < y. The decision at each step is whether to choose the pixel directly above the current pixel or the pixel; which is above and to the left (8-way stepping). Assume: Pi = (xi, yi) Ti = (xi, yi +1) Si = (xi -1, yi +1)

is the current pixel. is the pixel directly above is the pixel above and to the left.

To apply the midpoint method, we define a circle function: fcircle(x, y) = x2 + y2 r2 Therefore following relations can be observed: 61 Copyright Virtual University of Pakistan

6-Circle Drawing Techniques

VU

f circle (x, y)

< 0, = 0, > 0,

if (x, y) is inside the circle boundary if (x, y) is on the circle boundary if (x, y) is outside the circle boundary

Yk The circle function tests given above are performed for the X2+Y2-r2=0 midpoints between pixels near the circle path at each Yk-1 sampling step. Thus, the circle function is the decision parameter in the midpoint algorithm, and we can set up incremental calculations for this function as we did in the line algorithm. Xk Xk+1 Figure given above shows the midpoint between the two candidate pixels at sampling position xk+1. Assuming we have just plotted the pixel at (xk, yk), we next need to determine whether the pixel at position (xk + 1, yk), we next need to determine whether the pixel at position (xk+1, yk) or the one at position (xk+1, yk-1) is closer to the circle. Our decision parameter is the circle function evaluated at the midpoint between these two pixels: Pk = f circle ( xk + 1, yk - ) Pk = ( xk + 1 ) 2 + ( yk - ) 2 r 2

( 1 )

If pk < 0, this midpoint is inside the circle and the pixel on scan line yk is closer to the circle boundary. Otherwise, the mid position is outside or on the circle boundary, and we select the pixel on scan-line yk-1. Successive decision parameters are obtained using incremental calculations. We obtain a recursive expression for the next decision parameter by evaluating the circle function at sampling position xk+2= xk+1+1=xk+1+1 = xk+2: Pk+1 = f circle ( xk+1 + 1, yk+1 - ) Pk+1 = [ ( xk + 1 ) + 1 ] 2 + ( yk+1 - ) 2 r 2( 2 ) Subtracting (1) from (2), we get
2

Pk+1 - Pk = [ ( xk + 1 ) + 1 ] 2 + ( yk+1 - ) 2 r 2 ( xk + 1 ) 2 - ( yk - ) 2 + Pk+1 = Pk + 2( xk + 1 ) + ( y2k+1 - y2k ) ( yk+1 - yk ) + 1

r or

Where yk+1 is either yk or yk-1, depending on the sign of Pk. Therefore, if Pk < 0 or negative then yk+1 will be yk and the formula to calculate Pk+1 will be: Pk+1 = Pk + 2( xk + 1 ) + ( y2k - y2k ) ( yk - yk ) + 1 Pk+1 = Pk + 2( xk + 1 ) + 1 Otherwise, if Pk > 0 or positive then yk+1 will be yk-1and the formula to calculate Pk+1 will be: Pk+1 = Pk + 2( xk + 1 ) + [ (y k -1)2 - y2k ] ( yk -1- yk ) + 1 Pk+1 = Pk + 2( xk + 1 ) + (y2 k - 2 y k +1 - y2k ] ( yk -1- yk ) + 1 62 Copyright Virtual University of Pakistan

6-Circle Drawing Techniques

VU

Pk+1 = Pk + 2( xk + 1 ) - 2 y k + 1 + 1 +1 Pk+1 = Pk + 2( xk + 1 ) - 2 y k + 2 +1 Pk+1 = Pk + 2( xk + 1 ) - 2 ( y k 1 ) + 1 Now a similar case that we observe in line algorithm is that how would starting Pk be evaluated. For this at the start pixel position will be ( 0, r ). Therefore, putting this value is equation , we get P0 = ( 0 + 1 ) 2 + ( r - ) 2 r 2 P 0 = 1 + r2 - r + r 2 P0 = 5/4 r If radius r is specified as an integer, we can simply round p0 to: P0 = 1 r Since all increments are integer. Finally sum up all in the algorithm: MidpointCircle (xcenter, ycenter, radius) y = r; x = 0; p = 1 - r; do DrawSymmetricPoints (xcenter, ycenter, x, y) x=x+1 If p < 0 Then p=p+2*(x+1)+1 else y=y-1 p=p+2*(x+1)2*(y -1)+1 while ( x < y ) Now let us consider an example to calculate first octant of the circle using above algorithm; while one quarter is displayed where you can observe that exact circle is passing between the points calculated in a raster circle. Example: xcenter= 0 ycenter= 0 radius= 10 p -9 -6 -1 6 -3 8 5 6 x 0 1 2 3 4 5 6 7 Y 10 10 10 10 9 9 8 7 63 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Lecture No.7
Ellipse

Ellipse and Other Curves

An ellipse is a curve that is the locus of all points in the plane the sum of whose distances r1 and r2 from two fixed points F1 and F2, (the foci) separated by a distance of is a given positive constant . This results in the two-center bipolar coordinate equation: r1 + r2 = 2a where a is the semi-major axis and the origin of the coordinate system is at one of the foci. The corresponding parameter b is known as the semiminor axis. The ellipse was first studied by Menaechmus, investigated by Euclid, and named by Apollonius. The focus and conic section directrix of an ellipse were considered by Pappus. In 1602, Kepler believed that the orbit of Mars was oval; he later discovered that it was an ellipse with the Sun at one focus. In fact, Kepler introduced the word "focus" and published his discovery in 1609. In 1705 Halley showed that the comet now named after him moved in an elliptical orbit around the Sun (MacTutor Archive). An ellipse rotated about its minor axis gives an oblate spheroid, while an ellipse rotated about its major axis gives a prolate spheroid. Let an ellipse lie along the x-axis and find the equation of the figure given above where F1 and F2 are at (-c, 0) and (c, 0). In Cartesian coordinates,

Bring the second term to the right side and square both sides,

Now solve for the square root term and simplify

Square one final time to clear the remaining square root, 64 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Grouping the x terms then gives

this can be written in the simple form

Defining a new constant

puts the equation in the particularly simple form

The parameter b is called the semi-minor axis by analogy with the parameter a, which is called the semi-major axis (assuming b < a). The fact that b as defined at right is actually the semi-minor axis is easily shown by letting r1 and r2 be equal. Then two right triangles are produced, each with hypotenuse a, base c, and height b = a2 - c2. Since the largest distance along the minor axis will be achieved at this point, b is indeed the semi-minor axis. If, instead of being centered at (0, 0), the center of the ellipse is at (x0, y0), at right equation becomes:

Ellipse Drawing Techniques Now we already understand circle drawing techniques. One way to draw ellipse is to use the following equation:

where x0 may be replaced by xc in case of center other than origin and same in case of y. Another way is to use polar coordinates r and , for that we have parametric equations: 65 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

x = xc + rx cos y = yc + ry sin Four-way symmetry Symmetric considerations can be had to further reduce computations. An ellipse in standard position is symmetric between quadrants, but unlike a circle, it is not symmetric between the two octants of a quadrant. Thus, we must calculate pixel positions along the elliptical arc throughout one quadrant, and then we obtain positions in the remaining three quadrants by symmetry as shown in at right figure. Midpoint ellipse algorithm Consider an ellipse centered at the origin:

To apply the midpoint method, we define an ellipse function: f ellipse (x, y) = ry2x2 + rx2y2 rx2 ry2 Therefore following relations can be observed: < 0, = 0, > 0, if (x, y) is inside the circle boundary if (x, y) is on the circle boundary if (x, y) is outside the circle boundary

f ellipse (x, y)

Now as you have some idea that ellipse is different from circle. Therefore, a similar approach that is applied in circle can be applied here using some different sampling direction as shown in the figure at right. There are two regions separated in one octant. Therefore, idea is that in region 1 sampling will be at x direction; whereas y coordinate will be related to decision parameter. In region 2 sampling will be at y direction; whereas x coordinate will be related to decision parameter. So consider first region 1. We will start at (0, r y); we take unit steps in the x direction until we reach the boundary between region 1 and region 2. Then we switch to unit steps in the y direction over the remainder of the curve in the first quadrant. At each step, we 66 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

need to test the value of the slope of the curve. The ellipse slope is calculated from following equation: dy / dx = -2 ry2x2 / 2 rx2y2 At the boundary region 1 and region 2, dy/ dx = -1 and 2 rx2y2 = 2 ry2x2 Therefore, we move out of region 1 whenever 2 ry2x2 >= 2 rx2y2 Figure at right shows the midpoint between the two candidate pixels at sampling position xk + 1 in the first region. Assuming position (xk, yk) has been selected at the previous step; we determine the next position along the ellipse path by evaluating the decision parameter at this midpoint: P1k = f ellipse ( xk + 1, yk ) f ellipse (xk +1, yk ) = r2y ( xk + 1)2 + rx2 ( yk )2 rx2 ry2 -------( 1 ) If pk < 0, this midpoint is inside the ellipse and the pixel on scan line yk is closer to the ellipse boundary. Otherwise, the mid position is outside or on the ellipse boundary, and we select the pixel on scan-line yk-1. Successive decision parameters are obtained using incremental calculations. We obtain a recursive expression for the next decision parameter by evaluating the ellipse function at sampling position xk+1=xk+2: fellipse (xk+1 +1, yk+1 ) = ry2 [( xk + 1) + 1 ] 2 + rx2 ( yk+1 ) 2 rx2 ry2 ---( 2 ) Subtracting (1) from (2), and by simplification, we get Pk+11 = Pk1 + 2 ry2 ( xk + 1) + rx2 ( yk+12 yk2 ) - rx2 (yk+1 - yk ) + ry2 Where yk+1 is either yk or yk-1, depending on the sign of Pk. Therefore, if Pk < 0 or negative then yk+1 will be yk and the formula to calculate Pk+1 will be: P1k+1 = Pk1 + 2 ry2 ( xk + 1) + rx2 ( y k2 yk2 ) - rx2 (yk - yk ) + ry2 Pk+11 = Pk1 + 2 ry2 ( xk + 1 ) + ry2 Otherwise, if Pk > 0 or positive then yk+1 will be yk-1and the formula to calculate Pk+1 will be:

ry2x2 + rx2y2 rx2 ry2=0

67 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Pk+11 = Pk1 + 2 ry2 ( xk + 1) + r2x ( (yk 1)2 yk2 ) - rx2 (yk 1 yk ) + ry2 P1k+1 = P1k + 2 r y 2 ( x k + 1) + r x 2 ( 2 y k + 1 ) - r x 2 ( 1 ) + r y 2 P1k+1 = P1k + 2 ry2 ( x k + 1) 2 rx2 y k + r2x + rx2 + ry2 P1k+1 = P1k + 2 r y2 ( x k + 1) 2 r x2 ( y k 1 ) + r y2 Now a similar case that we observe in line algorithm is from where starting Pk will evaluate. For this at the start pixel position will by ( 0, ry ). Therefore, putting this value is equation , we get P10 = r y2 ( 0 + 1) 2 + r x2 (ry ) 2 rx2 ry2 P10 = r2y + rx2 ry2 r2x ry + rx2 rx2 ry2 P10 = r2y rx2 r y + rx2 Similarly same procedure will be adapted for region 2 and decision parameter will be calculated, here we are giving decision parameter and there derivation is left as an exercise for the students. Pk+12 = Pk2 2 rx2 ( yk + 1 ) + rx2 if pk2 > 0 ,

Pk+12 = Pk2 + 2 ry2 ( xk + 1) 2 rx2 yk + rx2 otherwise The initial parameter for region 2 will be calculated by following formula using the last point calculated in region 1 as: P02 = ry2 (x0 + ) + rx2 (y0 - 1 )2 rx2ry2 Since all increments are integer. Finally sum up all in the algorithm: MidpointEllipse (xcenter, ycenter, r x, r y) x =0 x =0 y = ry do DrawSymmetricPoints (xcenter, ycenter, x, y) P01 = ry2 rx2 ry + rx2 x = x +1 If p1k < 0 Pk+11 = Pk1 + 2 ry2 ( xk + 1 ) + ry2 else Pk+11 = Pk1 + 2 ry2 ( xk + 1) 2 rx2 ( yk 1 ) + r2y y = y -1 P02 = ry2 (x 0 + ) + rx2 (y 0 - 1 )2 rx2 ry2 If p2k > 0 y = y -1

68 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Pk+12 = Pk2 2 rx2 ( yk + 1 ) + rx2 else Pk+12 = Pk2 + 2 ry2 ( xk + 1) 2 rx2 yk + rx2 while ( 2 ry2x2 >= 2 rx2y2 )

x=x+1

Other Curves Various curve functions are useful in object modeling, animation path specifications, data, function graphing, and other graphics applications. Commonly encountered curves include conics, trigonometric and exponential functions, probability distributions, general polynomials, and spline functions. Displays of these curves can be generated with methods similar to those discussed for the circle and ellipse. We can obtain positions along curve paths directly from explicit representations y = f(x) or from parametric forms. Alternatively, we could apply the incremental midpoint method to plot curves described with implicit functions f(x,y) = 0. Conic Sections A conic section is the intersection of a plane and a cone. Circle Ellipse (h)Parabola (h)Hyperbola (h)

Ellipse

(v)Parabola

(v)Hyperbola

(v)

The general equation for a conic section: Ax2 + Bxy + Cy2 + Dx + Ey + F = 0

The type of section can be found from the sign of: B2-4AC If B2 - 4AC is then the curve is a... < 0 ellipse, circle, point or no curve. = 0 parabola, 2 parallel lines, 1 line or no curve. > 0 hyperbola or 2 intersecting lines. For any of the below with a center (j, k) instead of (0, 0), replace each x term with (x-j) and each y term with (y-k).

69 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Circle

Ellipse
2 2 2 2

Parabola

Hyperbola

Equation vertex): Equations Asymptotes: Equation vertex): Equations Asymptotes:

(horiz. of (vert. of

x2 + y2 = r2

x /a +y /b 4px = y2 =1

x2 / a2 - y2 / b2 = 1 y = (b/a)x

x2 + y2 = r2

y2 / a2 + x2 / b2 4py = x2 =1

y2 / a2 - x2 / b2 = 1 x = (b/a)y

Variables:

a = major radius (= 1/2 length major axis) r = circle b = minor radius radius (= 1/2 length minor axis) c = distance center to focus 0 p=0 a -b =c
2 2 2

p = distance from vertex to focus (or directrix)

a = 1/2 length major axis b = 1/2 length minor axis c = distance center to focus

Eccentricity: Relation to Focus:

c/a p=p

c/a a2 + b2 = c2

Definition: is the sum of distance to distance to difference between locus of all points distances to the origin is focus = distance distances to each which meet the each focus is constant to directrix foci is constant condition... constant
Hyperbola

We begin this section with the definition of a hyperbola. A hyperbola is the set of all points (x, y) in the plane the difference of whose distances from two fixed points is some constant. The two fixed points are called the foci. Each hyperbola consists of two branches. The line segment; which connects the two foci intersects the hyperbola at two points, called the vertices. The line segment; which ends at these vertices is called the transverse axis and the midpoint of this line is called the center of the hyperbola. See figure at right for a sketch of a hyperbola with these pieces identified. Note that, as in the case of the ellipse, a hyperbola can have a vertical or horizontal orientation. We now turn our attention to the standard equation of a hyperbola. We say that the standard equation of a hyperbola centered at the origin is given by 70 Copyright Virtual University of Pakistan Focus

Center Transverse Axis Vertex

7-Ellipse and Other Curves

VU

if the transverse axis is horizontal, or

if the transverse axis is vertical. Notice a very important difference in the notation of the equation of a hyperbola compared to that of the ellipse. We see that a always corresponds to the positive term in the equation of the ellipse. The relationship of a and b does not determine the orientation of the hyperbola. (Recall that the size of a and b was used in the section on the ellipse to determine the orientation of the ellipse.) In the case of the hyperbola, the variable in the ``positive'' term of the equation determines the orientation of the hyperbola. Hence, if the variable x is in the positive term of the equation, as it is in the equation

then the hyperbola is oriented as follows:

If the variable y is in the positive term of the equation, as it is in the equation

then we see the following type of hyperbola:

Note that the vertices are always a units from the center of the hyperbola, and the distance c of the foci from the center of the hyperbola can be determined using a, b, and the following equality:

We will use this relationship often, so keep it in mind. The next question you might ask is this: ``what happens to the equation if the center of the hyperbola is not (0, 0)?'' As in the case of the ellipse, if the center of the hyperbola is (h, k), then the equation of the hyperbola becomes

if the transverse axis is horizontal, or 71 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

if the transverse axis is vertical. A few more terms should be mentioned here before we move to some examples. First, as in the case of an ellipse, we say that the eccentricity of a hyperbola, denoted by e, is given by

or we say that the eccentricity of a hyperbola is given by the ratio of the distance between the foci to the distance between the vertices. Now in the case of a hyperbola, the distance between the foci is greater than the distance between the vertices. Hence, in the case of a hyperbola,

Recall that for the ellipse,

Two final terms that we must mention are asymptotes and the conjugate axis. The two branches of a hyperbola are bounded by two straight lines, known as asymptotes. These asymptotes are easily drawn once one plots the vertices and the points (h, k+b) and (h, k-b) and draws the rectangle which goes through these four points. The line segment joining (h, k+b) and (h, k-b) is called the conjugate axis. The asymptotes then are simply the lines which go through the corners of the rectangle. But what are the actual equations of these asymptotes? Note that if the hyperbola is oriented horizontally, then the corners of this rectangle have the following coordinates:

and

Here I have paired these points in such a way that each asymptote goes through one pair of the points. Consider the first pair of points:

Given two points, we can find the equation of the unique line going through the points using the point--slope form of the line. First, let us determine the slope of our line. We 72 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

find this as ``change in y over change in x'' or ``rise over run''. In this case, we see that this slope is equal to

or simply

Then, we also know that the line goes through the center (h, k). Hence, by the point-slope form of a line, we know that the equation of this asymptote is

or

The other asymptote in this case has a negative slope; which is given by

Using the same argument, we see that this asymptote has equation

What if the hyperbola is vertically oriented? Then one of the asymptote will go through the corners of the rectangle given by

Then the slope in this case will not be b/a but will be a/b. Hence, analogous to the work we just performed, we can show that the asymptotes of a vertically oriented hyperbola are determined by

and

Parabola

A parabola is the set of all points (x, y) that are the same distance from a fixed line (called the directrix) and a fixed point (focus) not on the directrix. See figure for the view of a parabola and its related focus and directrix. 73 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

Note that the graph of a parabola is similar to one branch of a hyperbola. However, you should realize that a parabola is not simply one branch of a hyperbola. Indeed, the branches of a hyperbola approach linear asymptotes, while a parabola does not do so. Several other terms exist which are associated with a parabola. The midpoint between the focus and directrix of the parabola is called the vertex and the line passing through the focus and vertex is called the axis of the parabola. (This is similar to the major axis of the ellipse and the transverse axis of the hyperbola.) See figure at right.

Now let's move to the standard algebraic equations for parabolas and note the four types of parabolas that exist. As we discuss the four types, you should notice the differences in the equations that are related to each of the four parabolas. The standard form of the equation of the parabola with vertex at (0, 0) with the focus lying d units from the vertex is given by

if the axis is vertical and

if the axis is horizontal. See figure below for an example with vertical axis and figure below for an example with horizontal axis.

Note here that we have assumed that

74 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

It is also the case that d could be negative, which flips the orientation of the parabola. (See Figures) Thus, we see that there are four different orientations of parabolas, which depend on a) which variable is squared (x or y) and b) whether d is positive or negative. One last comment before going to some examples; if the vertex of the parabola is at (h, k), then the equation of the parabola does change slightly. The equation of a parabola with vertex at (h, k) is given by

if the axis is vertical and

Rotation of Axes

Note that in the sections at right dealing with the ellipse, hyperbola, and the parabola, the algebraic equations that appeared did not contain a term of the form xy. However, in our Algebraic View of the Conic Sections,'' we stated that every conic section is of the form

where A, B, C, D, E, and F are constants. In essence, all of the equations that we have studied have had B=0. So the question arises: ``what role, if any, does the xy term play in conic sections? If it were present, how would that change the geometric figure?'' First of all, the answer is NOT that the conic changes from one type to another. That is to say, if we introduce a xy term, the conic does NOT change from an ellipse to a hyperbola. If we start with the standard equation of an ellipse and insert an extra term, a xy term, we still have an ellipse.

v y g

So what does the xy term do? The xy term actually rotates the graph in the plane. For example, in the case of an ellipse, the major axis is no longer parallel to the x-axis or y-axis. Rather, depending on the constant in front of the xy term, we now have the major axis rotated.
Animated Applications Ellipses, hyperbolas, and parabolas are particularly useful in certain animation applications. These curves describe orbital and other motions for objects subjected to gravitational, electromagnetic, or nuclear forces. Planetary orbits in the solar system, for

75 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

example, are ellipses; and an object projected into a uniform gravitational field travels along a parabolic trajectory. Figure at right shows a parabolic path in standard position for gravitational field acting in the negative y direction. The explicit equation for the parabolic trajectory of the object shown can be written as: y = y0 + a (x - x0 )2 + b ( x - x0 ) With constants a and b determined by the initial velocity v0 of the object and the acceleration g due to the uniform gravitational force. We can also describe such parabolic motions with parametric equations using a time parameter t, measured in seconds from the initial projection point: x = x0 + vx0 t y = y0 + vy0 t g t 2 Here vx0 and vy0 are the initial velocity components, and the value of g near the surface of the earth is approximately 980 cm/ sec2. Object positions along the parabolic path are then calculated at selected time steps.
Some related real world applications are given below. Parabolic Reflectors

One of the ``real--world'' applications of parabolas involves the concept of a 3-dimensional parabolic reflector in which a parabola is revolved about its axis (the line segment joining the vertex and focus). The shape of car headlights, mirrors in reflecting telescopes, and television and radio antennae (such as the one at right) all utilize this property. In terms of a car headlight, this property is used to reflect the light rays emanating from the focus of the parabola (where the actual light bulb is located) in parallel rays. This property is used in a converse fashion when one considers parabolic antennae. Here, all incoming rays parallel to the axis of the parabola are reflected through the focus.
Elliptical Orbits

At one time, it was thought that the planets in our solar system revolve around the sun in a circular 76 Copyright Virtual University of Pakistan

7-Ellipse and Other Curves

VU

orbit. It was later discovered, however, that the orbits are not circular, but were actually very round elliptical shapes. (Recall the discussion of the eccentricity of an ellipse mentioned at right.) The eccentricity of the orbit of the Earth around the sun is approximately 0.0167, a fairly small number. Pluto's orbit has the highest eccentricity of all the planets in our solar system at 0.2481. Still, this is not a very large value. As a matter of fact, the sun acts as one of the foci in the ellipse. This phenomenon was first noted by Apollonius in the second century B.C. Kepler later studied this in a more rigorous fashion and developed the scientific view of planetary motion.
Whispering Galleries

In rooms whose ceilings are elliptical, a sound made at one focus of the ellipse will be reflected to the other focus (across the room), allowing people standing at the two foci to hear one another very clearly. This has been called the ``whispering gallery'' effect and has been used by many in the design of special rooms. In particular, St. Paul's Cathedral and one of the rooms at the United States Capitol were built with this in mind.
Polynomials and Spline Curves A polynomial function of nth degree in x is defined as

y =

a kx k

y = a 0 x 0 + a 1 x 1 + ----------------+ a n-1 x n-1 + a n x n Where n is a nonnegative integer and the a k are constants, with a n not equal to 0. We get a quadratic when n = 2; a cubic polynomial when n = 3; a quadratic when n = 4; and so forth. And obviously a straight line when n = 1. Polynomials are useful in a number of graphics applications, including the design of object shapes, the specifications of animation paths, and the graphing of data trends in a discrete set of data points. Designing object shapes or motion paths is typically done by specifying a few points to define the general curve contour, then fitting the selected points with a polynomial. One way to accomplish the curve fitting is to construct a cubic polynomial curve section between each pair of specified points. Each curve section is then described in parametric form as x = a x0 + a x1 u + a x2 u 2 + a x3 u 3 y = a y0 + a y1 u + a y2 u 2 + a y3 u 3 Where parameter u varies over the interval 0 to1. A curve is shown below calculated using at right equations.

Continuous curves that are formed with polynomial pieces are called spline curves, or simply splines. Spline is a detailed topic; which will be discussed later in 3 dimensions. 77 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Lecture No.8

Filled-Area Primitives-I

So far we have covered some output primitives that is drawing primitives like point, line, circle, ellipse and some other variations of curves. Also we can draw certain other shapes with the combinations of lines like triangle, rectangle, square and other polygons (we will have some discussion on polygons coming ahead). Also we can draw some shapes using mixture of lines and curves or circles etc. So we are able to draw outline/ sketch of certain models but need is there to make a solid model. Therefore, in this section we will see what are filled area primitives and what are the different issues related to them. There are two basic approaches to area filling on raster systems. One way is to draw straight lines between the edges of polygon called scan-line polygon filling. As said earlier there are several issues related to scan line polygon, which we will discuss in detail. Second way is to start from an interior point and paint outward from this point till we reach the boundary called boundary-fill. A slight variation of this technique is used to fill an area specified by cluster (having no specific boundary). The technique is called flood-fill and having almost same strategy that is to start from an interior point and start painting outward from this point till the end of cluster. Now having an idea we will try to see each of these one by one, starting from scan-line polygon filling.
Scan-line Polygon Fill

Before we actually start discussion on scan-line polygon filling technique, it is useful to discuss what is polygon? Besides polygon definition we will discuss the following topics one by one to have a good understanding of the concept and its implementation.

Polygon Definition Filled vs. Unfilled Polygons Parity Definition Scan-Line Polygon Fill Algorithm Special Cases Handled By the Fill Polygon Fill Example

Polygon

A polygon can be defined as a shape that is formed by line segments that are placed end to end, creating a continuous closed path. Polygons can be divided into three basic types: convex, concave, and complex. I. Convex polygons are the simplest type of polygon to fill. To determine whether or not a polygon is convex, ask the following question: Does a straight line connecting ANY two points that are inside the polygon intersect any edges of the polygon?

78 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

If the answer is no, the polygon is convex. This means that for any scan-line, the scan-line will cross at most two polygon edges (not counting any horizontal edges). Convex polygon edges also do not intersect each other. II. Concave polygons are a superset of convex polygons, having fewer restrictions than convex polygons. The line connecting any two points that lie inside the polygon may intersect more than two edges of the polygon. Thus, more than two edges may intersect any scan line that passes through the polygon. The polygon edges may also touch each other, but they may not cross one another.

Complex polygons are just what their name suggests: complex. Complex polygons are basically concave polygons that may have self-intersecting edges. The complexity arises from distinguishing which side is inside the polygon when filling it.

Difference between Filled and Unfilled Polygon When an unfilled polygon is rendered, only the points on the perimeter of the polygon are drawn. Examples of unfilled polygons are shown in the next page. However, when a polygon is filled, the interior of the polygon must be considered. All of the pixels within the boundaries of the polygon must be set to the specified color or pattern. Here, we deal only with solid colors. The following figure shows the difference between filled and unfilled polygons.

79 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

In order to determine which pixels are inside the polygon, the odd-parity rule is used within the scan-line polygon fill algorithm. This is discussed next.
Parity

What is parity? Parity is a concept used to determine which pixels lie within a polygon, i.e. which pixels should be filled for a given polygon. The Underlying Principle: Conceptually, the odd parity test entails drawing a line segment from any point that lies outside the polygon to a point P that we wish to determine whether it is inside or outside of the polygon. Count the number of edges that the line crosses. If the number of polygon edges crossed is odd, then P lies within the polygon. Similarly, if the number of edges is even, then P lies outside of the polygon. There are special ways of counting the edges when the line crosses a vertex. This will be discussed in the algorithm section. Examples of counting parity can be seen in the following demonstration.

Using the Odd Parity Test in the Polygon Fill Algorithm The odd parity method creates a problem: How do we determine whether a pixel lies outside of the polygon to test for an inside one, if we cannot determine whether one lies within or outside of the polygon in the first place? If we assume our polygon lies entirely within our scene, then the edge of our drawing surface lies outside of the polygon. Furthermore, it would not be very efficient to check each point on our drawing surface to see if it lies within the polygon and, therefore, needs to be colored. So, we can take advantage of the fact that for each scan-line we begin with even parity; we have NOT crossed any polygon edges yet. Then as we go from left to right across our scan line, we will continue to have even parity (i.e., will not use the fill color) until we cross the first polygon edge. Now our parity has changed to odd and we will start using the fill color. How long will we continue to use the fill color? Well, our parity won't change until we cross the next edge. Therefore, we want to color all of the pixels from when we crossed the first edge until we cross the next one. Then the parity will become even again. 80 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

So, you can see if we have a sorted list of x-intersections of all of the polygon edges with the scan line, we can simply draw from the first x to the second, the third to the forth and so on.
Polygon Filling

In order to fill a polygon, we do not want to have to determine the type of polygon that we are filling. The easiest way to avoid this situation is to use an algorithm that works for all three types of polygons. Since both convex and concave polygons are subsets of the complex type, using an algorithm that will work for complex polygon filling should be sufficient for all three types. The scan-line polygon fill algorithm, which employs the odd/even parity concept previously discussed, works for complex polygon filling.
Reminder: The basic concept of the scan-line algorithm is to draw points from edges of odd parity to even parity on each scan-line. b) What is a scan-line?

A scan-line is a line of constant y value, i.e., y=c, where c lies within our drawing region, e.g., the window on our computer screen. The scan-line algorithm is outlined next.
Algorithm

When filling a polygon, you will most likely just have a set of vertices, indicating the x and y Cartesian coordinates of each vertex of the polygon. The following steps should be taken to turn your set of vertices into a filled polygon.
1. Initializing All of the Edges:

The first thing that needs to be done is determine how the polygon's vertices are related. The all_edges table will hold this information. Each adjacent set of vertices (the first and second, second and third, similarly last and first) defines an edge. In above figure vertices are shown by small lines and edges are numbered from 1 to 9 each between successive vertices. For each edge, the following information needs to be kept in a table: 1. The minimum y value of the two vertices 2. The maximum y value of the two vertices 3. The x value associated with the minimum y value 4. The slope of the edge The slope of the edge can be calculated from the formula for a line: y = mx + b; where m = slope, b = y-intercept, 81 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

y0 = maximum y value, y1 = minimum y value, x0 = maximum x value, x1 = minimum x value The formula for the slope is as follows: m = (y0 - y1) / (x0 - x1). For example, the edge values may be kept as follows, where N is equal to the total number of edges - 1 (starting from 0) and each index into the all_edges array contains a pointer to the array of edge values.
Index 0 1 Y-min Y-max X-val 1/m

10 16 -

16 20 16

10 10 28

0 1.5 0

10

Table: All_edges 2. Initializing the Global Edge Table:

The global edge table will be used to keep track of the edges that are still needed to complete the polygon. Since we will fill the edges from bottom to top and left to right. To do this, the global edge table should be inserted with edges grouped by increasing minimum y values. Edges with the same minimum y values are sorted on minimum x values as follows: 1. Place the first edge with a slope that is not equal to zero in the global edge table. 2. If the slope of the edge is zero, do not add that edge to the global edge table. 3. For every other edge, start at index 0 and increase the index of the global edge table once each time the current edge's y value is greater than that of the edge at the current index in the global edge table. Next, Increase the index to the global edge table once each time the current edge's x value is greater than and the y value is less than or equal to that of the edge at the current index in the global edge table. If the index, at any time, is equal to the number of edges currently in the global edge table, do not increase the index. Place the edge information for minimum y value, maximum y value, x value, and 1/m in the global edge table at the index.

82 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

The global edge table should now contain all of the edge information necessary to fill the polygon in order of increasing minimum y and x values.
3. Initializing Parity

The initial parity is even since no edges have been crossed yet.
4. Initializing the Scan-Line

The initial scan-line is equal to the lowest y value for all of the global edges. Since the global edge table is sorted, the scan-line is the minimum y value of the first entry in this table.
5. Initializing the Active Edge Table

The active edge table will be used to keep track of the edges that are intersected by the current scan-line. This should also contain ordered edges. This is initially set up as follows: Since the global edge table is ordered on minimum y and x values, search, in order, through the global edge table and, for each edge found having a minimum y value equal to the current scan-line, append the edge information for the maximum y value, x value, and 1/m to the active edge table. Do this until an edge is found with a minimum y value greater than the scan line value. The active edge table will now contain ordered edges of those edges that are being filled as such:
Index 0 1 Y-max X-val 1/m

16 20 -

10 10 28
Active

0 1.5 0

16

6. Filling the Polygon

Filling the polygon involves deciding whether or not to draw pixels, adding to and removing edges from the active edge table, and updating x values for the next scan-line. Starting with the initial scan-line, until the active edge table is empty, do the following: 1. Draw all pixels from the x value of odd to the x value of even parity edge pairs. 2. Increase the scan-line by 1. 3. Remove any edges from the active edge table for which the maximum y value is equal to the scan line.

83 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

4. Update the x value for each edge in the active edge table using the formula x1 = x0 + 1/m. (This is based on the line formula and the fact that the next scan-line equals the old scan-line plus one.) 5. Remove any edges from the global edge table for which the minimum y value is equal to the scan-line and place them in the active edge table. 6. Reorder the edges in the active edge table according to increasing x value. This is done in case edges have crossed. Special Cases There are some special cases, the scan-line polygon fill algorithm covers these cases, but you may not understand how or why. The following will explain the handling of special cases to the algorithm.
1. Horizontal Edges:

Here we follow the minimum y value rule during scan-line polygon fill. If the edge is at the minimum y value for all edges, it is drawn. Otherwise, if the edge is at the maximum y value for any edge, we do not draw it. (See the next section containing information about top vs. bottom edges.) This is easily done in the scan-line polygon fill implementation. Horizontal edges are removed from the edge table completely. Question arises that how horizontal lines are are filled then? Since each horizontal line meets exactly two other edge end-points on the scan-line, the algorithm will allow a fill of the pixels between those two end-point vertices when filling on the scan-line which the horizontal line is on, if it meets the top vs. bottom edge criteria.

-->

-->

As can be seen above, if we start with a polygon with horizontal edges, we can remove the horizontal edges from the global edge table. The two endpoints of the edge will still exist and a line will be drawn between the lower edges following the scan-line polygon fill algorithm. (The blue arrowed line is indicating the scan-line for the bottom horizontal edge.)
2. Bottom and Left Edges vs. Top and Right Edges:

If polygons, having at least one overlapping edge the other, were filled completely from edge to edge, these polygons would appear to overlap and/or be distorted. This would be especially noticeable for polygons in which edges have limited space between them. In order to correct for this phenomenon, our algorithm does not allow fills of the right or top edges of polygons. This distortion problem could also be corrected by not drawing either the left or right edges and not drawing either the top or bottom edges of the polygon. Either way, a consistent method should be used with all polygons. If some polygons are filled with the left and bottom edges and others with the bottom and right edges, this problem will still occur. 84 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

-->

As can be seen above, if we remove the right and top edges from both polygons, the polygons no longer appear to be different shapes. For polygons with more overlap than just one edge, the polygons will still appear to overlap as was meant to happen.
3. How do we deal with two edges meeting at a vertex when counting parity? This is a scenario which needs to be accounted for in one of the following ways:

1. When dealing with two edges; which meet at a vertex and for both edges the vertex is the minimum point, the pixel is drawn and is counted twice for parity. Essentially, the following occurs. In the scan-line polygon fill algorithm, the vertex is drawn for the first edge, since it is a minimum value for that edge, but not for the second edge, since it is a right edge and right edges are not drawn in the scan-line fill algorithm. The parity is increased once for the first edge and again for the second edge.

2. When dealing with two edges; which meet at a vertex and for both edges the vertex is the maximum point, the pixel is not drawn and is counted twice for parity. Basically, this occurs because the vertex is not drawn for the first edge, since it is a maximum point for that edge, and parity is increased. The vertex is then not drawn for the 85 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

second edge, since it is a right edge, and parity is The point should not be drawn since maximum y values for edges are not drawn in the scan-line polygon fill implementation. 3. When dealing with two edges; which meet at a vertex and for one edge the vertex is the maximum point and for the other edge the vertex is the minimum point, we must also consider whether the edges are left or right edges. Two edges meeting in such a way can be thought of as one edge; which is "bent".

If the edges are on the left side of the polygon, the pixel is drawn and is counted once for parity purposes. This is due to the fact that left edges are drawn in the scan-line polygon fill implementation. The vertex is drawn just once for the edge; which has this vertex as its minimum point. Parity is incremented just once for this "bent edge".

4. If both edges are on the right, the pixel is not drawn and is counted just once for parity purposes. This is due to the fact that right edges are not drawn in the scan-line polygon fill implementation.
A Simple Example

0 1 2 3 4 5

(10, 10) (10, 16) (16, 20) (28, 10) (28, 16) (22, 10)
Ordered Vertices

Just to reiterate the algorithm, the following simple example of scanline polygon filling will be outlined. Initially, each vertex of the polygon is given in the form of (x, y) and is in an ordered array as such:

We will now walk through the steps of the algorithm to fill in the polygon:

86 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

1. Initializing All of the Edges:

We want to determine the minimum y value, maximum y value, x value, and 1/m for each edge and keep them in the all_edges table. We determine these values for the first edge as follows:
Y-min: Since the first edge consists of the first and second vertex in the array, we use the y values of those vertices to choose the lesser y value. In this case it is 10. Y-max: In the first edge, the greatest y value is 16. X-val: Since the x value associated with the vertex with the highest y value is 10, 10 is the x value for this edge. 1/m: Using the given formula, we get (10-10)/ (16-10) for 1/m.

The edge value results are in the form of Y-min, Y-max, X-val, and Slope for each edge array pointed to in the all_edges table. As a result of calculating all edge values, we get the following in the all_edges table.
Index 0 1 2 3 4 5 Y-min Y-max X-val 1/m

10 16 10 10 10 10

16 20 20 16 16 10

10 10 28 28 22 10

0 1.5 -1.2 0 1 Inf

Table: All_edges 2. Initializing the Global Edge Table:

We want to place all the edges in the global edge table in increasing y and x values, as long as slope is not equal to zero. For the first edge, the slope is not zero so it is placed in the global edge table at index=0.

Index 0

Y-min

Y-max

X-val

1/m

10

16
Table: global

10

For the second edge, the slope is not zero and the minimum y value is greater than that at zero, so it is placed in the global edge table at index=1.

87 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Index 0 1

Y-min

Y-max

X-val

1/m

10 16

16 20

10 10

0 1.5

Table: global

For the third edge, the slope is not zero and the minimum y value is equal the edge's at index zero and the x value is greater than that at index 0, so the index is increased to 1. Since the third edge has a lesser minimum y value than the edge at index 2 of the global edge table, the index for the third edge is not increased again. The third edge is placed in the global edge table at index=1.

Index 0 1 2

Y-min

Y-max

X-val

1/m

10 10 16

16 20 20
Table: global

10 28 10

0 -1.2 1.5

We continue this process until we have the following:

Index 0 1 2 3 4 Y-min Y-max X-val 1/m

10 10 10 10 16

16 16 16 20 20
Table: global

10 22 28 28 10

0 1 0 -1.2 1.5

Notice that the global edge table has only five edges and the all_edges table has six. This is due to the fact that the last edge has a slope of zero and, therefore, is not placed in the global edge table.
3. Initializing Parity

Parity is initially set to even.

4. Initializing the Scan-Line

Since the lowest y value in the global edge table is 10, we can safely choose 10 as our initial scan-line. 88 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

5. Initializing the Active Edge Table

Since our scan-line value is 10, we choose all edges which have a minimum y value of 10 to move to our active edge table. This results in the following.
Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 22 28 28

0 1 0 -1.2

16

20

10

1.5

Table: active 6. Filling the Polygon

Table: global

Starting at the point (0, 10), which is on our scan-line and outside of the polygon, will want to decide which points to draw for each scan-line.
1. Scan-line = 10:

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=22. Parity is then changed to even. The next edge is reached at x=28 and the point is drawn once on this scan-line due to the special parity case. We are now done with this scan-line. The polygon is now filled as follows:

89 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

First, we update the x values in the active edge table using the formula x1 = x0 + 1/m to get the following:
Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 23 28 26.8

0 1 0 -1.2

16

20

10

1.5

Table: active

Table: global

The edges then need to be reordered since the edge at index 3 of the active edge table has a lesser x value than that of the edge at index 2. Upon reordering, we get:
Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 23 26.8 28

0 1 -1.2 0

16

20

10

1.5

Table: active 2. Scan-line = 11:

Table: global

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=23. Parity is then changed to even. The next edge is reached at x=27 and parity is changed to odd. The points are then drawn until the next edge is reached at x=28. We are now done with this scan-line. The polygon is now filled as follows:

Upon updating the x values, the edge tables are as follows: 90 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Index 0 1 2 3

Ymax

Xval

1/m

Index 0

Y-min

Y-max

X-val

1/m

16 16 16 20

10 24 25.6 28

0 1 -1.2 0

16

20

10

1.5

Table: active

Table: global

It can be seen that no reordering of edges is needed at this time.

3. Scan-line = 12:

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=24. Parity is then changed to even. The next edge is reached at x=26 and parity is changed to odd. The points are then drawn until the next edge is reached at x=28. We are now done with this scan-line. The polygon is now filled as follows:

Updating the x values in the active edge table gives us:

Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 25 24.4 28

0 1 -1.2 0

16

20

10

1.5

Table: active

Table: global

We can see that the active edges need to be reordered since the x value of 24.4 at index 2 is less than the x value of 25 at index 1. Reordering produces the following:

91 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Index 0 1 2 3

Ymax

Xval

1/m

Index 0

Y-min

Y-max

X-val

1/m

16 16 16 20

10 24.4 25 28

0 1 0 -1.2

16

20

10

1.5

Table: active 4. Scan-line = 13:

Table: global

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=25 Parity is then changed to even. The next edge is reached at x=25 and parity is changed to odd. The points are then drawn until the next edge is reached at x=28. We are now done with this scan-line. The polygon is now filled as follows:

Upon updating the x values for the active edge table, we can see that the edges do not need to be reordered.
Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 23.2 26 28

0 1 0 -1.2

16

20

10

1.5

Table: active 5. Scan-line = 14:

Table: global

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=24. Parity is then changed to even. The next edge is reached at x=26 and parity is changed to odd. The points are then drawn until the 92 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

next edge is reached at x=28. We are now done with this scan-line. The polygon is now filled as follows:

Upon updating the x values for the active edge table, we can see that the edges still do not need to be reordered.
Index 0 1 2 3 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

16 16 16 20

10 22 27 28

0 -1.2 1 0

16

20

10

1.5

Table: active 6. Scan-line = 15:

Table: global

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is encountered at x=22. Parity is then changed to even. The next edge is reached at x=27 and parity is changed to odd. The points are then drawn until the next edge is reached at x=28. We are now done with this scan-line. The polygon is now filled as follows:

93 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Since the maximum y value is equal to the next scan-line for the edges at indices 0, 2, and 3, we remove them from the active edge table. This leaves us with the following:
Index 0 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

20

22

-1.2

16

20

10

1.5

Table: active

Table: global

We then need to update the x values for all remaining edges.

Index 0 Ymax Xval 1/m Index 0 Y-min Y-max X-val 1/m

20

20.8

-1.2

16

20

10

1.5

Table: active

Table: global

Now we can add the last edge from the global edge table to the active edge table since its minimum y value is equal to the next scan-line. The active edge table now looks as follows (the global edge table is now empty):
Index 0 1 Ymax Xval 1/m

20 20

20.8 10

-1.2 1.5

Table: active

These edges obviously need to be reordered. After reordering, the active edge table contains the following:
Index 0 1 Ymax Xval 1/m

20 20

10 20.8

1.5 -1.2

Table: active 7.

Scan-line = 16:

Once the first edge is encountered at x=10, parity = odd. All points are drawn from this point until the next edge is reached at x=21. We are now done with this scan-line. The polygon is now filled as follows:

94 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

The x values are updated and the following is obtained:

Index 0 1 Ymax Xval 1/m

20 20

11.5 19.6

1.5 -1.2

Table: active 8. Scan-line = 17:

Once the first edge is encountered at x=12, parity = odd. All points are drawn from this point until the next edge is reached at x=20. We are now done with this scan-line. The polygon is now filled as follows:

We update the x values and obtain:

Index 0 1 Ymax Xval 1/m

20 20

13 18.4

1.5 -1.2

Table: active 9. Scan-line = 18:

95 Copyright Virtual University of Pakistan

8-Filled-Area Primitives-I

VU

Once the first edge is encountered at x=13, parity = odd. All points are drawn from this point until the next edge is reached at x=19. We are now done with this scan-line. The polygon is now filled as follows:

Upon updating the x values we get:

Index 0 1 Ymax Xval 1/m

20 20

14.5 17.2

1.5 -1.2

Table: active 10. Scan-line = 19:

Once the first edge is encountered at x=15, parity = odd. All points are drawn from this point until the next edge is reached at x=18. We are now done with this scan-line. Since the maximum y value for both edges in the active edge table is equal to the next scan-line, we remove them. The active edge table is now empty and we are now done. The polygon is now filled as follows:

Now that we have filled the polygon, let's see what it looks like to the human eye:

96 Copyright Virtual University of Pakistan

9-Filled-Area Primitives-II

VU

Lecture No.9

Filled-Area Primitives-II

Boundary fill Another important class of area-filling algorithms starts at a point known to be inside a figure and starts filling in the figure outward from the point. Using these algorithms a graphic artist may sketch the outline of a figure and then select a color or pattern with which to fill it. The actual filling process begins when a point inside the figure is selected. These routines are like the paint-scan function seen in common interactive paint packages. The first such method that we will discuss is called the boundary-fill algorithm. The boundary-fill method requires the coordinates of a starting point, a fill color, and a boundary color as arguments. Boundary fill algorithm: The Boundary fill algorithm performs the following steps: Check the pixel for boundary color Check the pixel for fill color Set the pixel in fill color Run the process for neighbors

The pseudo code for Boundary fill algorithm can be written as: boundaryFill (x, y, fillColor, boundaryColor) if ((x < 0) || (x >= width)) return if ((y < 0) || (y >= height)) return current = GetPixel(x, y) if ((current != boundaryColor) && (current != fillColor)) setPixel(fillColor, x, y) boundaryFill (x+1, y, fillColor, boundaryColor) boundaryFill (x, y+1, fillColor, boundaryColor) boundaryFill (x-1, y, fillColor, boundaryColor) boundaryFill (x, y-1, fillColor, boundaryColor) Note that this is a recursive routine. Each invocation of boundaryFill () may call itself four more times. The logic of this routine is very simple. If we are not either on a boundary or already filled we first fill our point, and then tell our neighbors to fill themselves.

97 Copyright Virtual University of Pakistan

9-Filled-Area Primitives-II

VU

Process of Boundary Fill Algorithm

By the way, sometimes the boundary fill algorithm doesn't work. Can you think of such a case?
Flood Fill Sometimes we need an area fill algorithm that replaces all connected pixels of a selected color with a fill color. The flood-fill algorithm does exactly that. Flood-fill algorithm An area fill algorithm that replaces all connected pixels of a selected color with a fill color.

Before Applying Flood-fill algorithm (Light color)

After Applying Flood-fill algorithm (Dark color) Flood-fill algorithm in action

98 Copyright Virtual University of Pakistan

9-Filled-Area Primitives-II

VU

The pseudo code for Flood fill algorithm can be written as: public void floodFill(x, y, fillColor, oldColor) if ((x < 0) || (x >= width)) return if ((y < 0) || (y >= height)) return if ( getPixel (x, y) == oldColor) setPixel (fillColor, x, y) floodFill (x+1, y, fillColor, oldColor) floodFill (x, y+1, fillColor, oldColor) floodFill (x-1, y, fillColor, oldColor) floodFill (x, y-1, fillColor, oldColor)

It's a little awkward to kick off a flood fill algorithm because it requires that the old color must be read before it is invoked. The following implementation overcomes this limitation, and it is also somewhat faster, a little bit longer. The additional speed comes from only pushing three directions onto the stack each time instead of four. fillFast (x, y, fillColor) if ((x < 0) || (x >=width)) return if ((y < 0) || (y >=height)) return int oldColor = getPixel (x, y) if ( oldColor == fill ) return setPixel (fillColor, x, y) fillEast (x+1, y, fillColor, oldColor) fillSouth (x, y+1, fillColor, oldColor) fillWest (x-1, y, fillColor, oldColor) fillNorth (x, y-1, fillColor, oldColor) fillEast (x, y, fillColor, oldColor) if (x >= width) return if ( getPixel(x, y) == oldColor) setPixel( fillColor, x, y) fillEast (x+1, y, fillColor, oldColor) fillSouth (x, y+1, fillColor, oldColor) fillNorth (x, y-1, fillColor, oldColor) fillSouth(x, y, fillColor, oldColor) if (y >=height) return if (getPixel (x, y) == oldColor) setPixel (fillColor, x, y) fillEast (x+1, y, fillColor, oldColor) fillSouth (x, y+1, fillColor, oldColor) fillWest (x-1, y, fillColor, oldColor) fillWest(x, y, fillColor, oldColor) { 99 Copyright Virtual University of Pakistan

9-Filled-Area Primitives-II

VU

if (x < 0) return if (getPixel (x, y) == oldColor) setPixel (fillColor, x, y) fillSouth (x, y+1, fillColor, oldColor) fillWest (x-1, y, fillColor, oldColor) fillNorth (x, y-1, fillColor, oldColor) fillNorth (x, y, fill, old) if (y < 0) return if (getPixel (x, y) == oldColor) setPixel (fill, x, y) fillEast (x+1, y, fillColor, oldColor) fillWest (x-1, y, fillColor, oldColor) fillNorth (x, y-1, fillColor, oldColor)

A final consideration when writing an area-fill algorithm is the size and connectivity of the neighborhood around a given pixel.

The eight-connected neighborhood is able to get into nooks and crannies that an algorithm based on a four-connected neighborhood cannot. Here's the code for an eight-connected flood fill. floodFill8 (x, y, fill, old) if ((x < 0) || (x >=width)) return if ((y < 0) || (y >=height)) return if (getPixel (x, y) == oldColor) setPixel (fill, x, y); floodFill8 (x+1, y, fillColor, oldColor) floodFill8 (x, y+1, fillColor, oldColor) floodFill8 (x-1, y, fillColor, oldColor) floodFill8 (x, y-1, fillColor, oldColor) floodFill8 (x+1, y+1, fillColor, oldColor) floodFill8 (x-1, y+1, fillColor, oldColor) floodFill8 (x-1, y-1, fillColor, oldColor) floodFill8 (x+1, y-1, fillColor, oldColor)

100 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Lecture No.10

Mathematics Fundamentals

Matrices and Simple Matrix Operations In many fields matrices are used to represent objects and operations on those objects. In computer graphics matrices are heavily used especially their major role is in case of transformations (we will discuss in very next lecture), but not only transformation there are many areas where we use matrices and we will see in what way matrices help us. Anyhow today we are going to discuss matrix and their operation so that we will not face any problem using matrices in coming lectures and in later lectures. Today we will cover following topics: What a Matrix is? Dimensions of a Matrix Elements of a Matrix Matrix Addition Zero Matrix Matrix Negation Matrix Subtraction Scalar multiplication of a matrix The transpose of a matrix Definition of Matrix A matrix is a collection of numbers arranged into a fixed number of rows and columns. Usually the numbers are real numbers. In general, matrices can contain complex numbers but we won't see those here. Here is an example of a matrix with three rows and three columns:

The top row is row 1. The leftmost column is column 1. This matrix is a 3x3 matrix because it has three rows and three columns. In describing matrices, the format is: rows X columns Each number that makes up a matrix is called an element of the matrix. The elements in a matrix have specific locations. The upper left corner of the matrix is row 1 column 1. In the above matrix the element at row 1 column 1 is the value 1. The element at row 2 columns 3 is the value 4.6. Matrix Dimensions The numbers of rows and columns of a matrix are called its dimensions. Here is a matrix with three rows and two columns: 101 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Sometimes the dimensions are written off to the side of the matrix, as in the above matrix. But this is just a little reminder and not actually part of the matrix. Here is a matrix with different dimensions. It has two rows and three columns. This is a different "data type" than the previous matrix.

Question: What do you suppose a square matrix is? Here is an example:

Answer: The number of rows == the number of columns Square Matrix In a square matrix the number of rows equals the number of columns. In computer graphics, square matrices are used for transformations. A column matrix consists of a single column. It is a N x 1 matrix. These notes, and most computer graphics texts, use column matrices to represent geometrical vectors. At left is a 4 x 1 column matrix. A row matrix consists of a single row. A column matrix is also called column vector and call a row matrix a row vector. Question: What are square matrices used for? Answer: Square matrices are used (in computer graphics) to represent geometric transformations. Names for Matrices Try to remember that matrix starts from rows never from columns so if order of matrix is 3*2 that means there are three rows and two columns. A matrix can be given a name. In printed text, the name for a matrix is usually a capital letter in bold face, like A or M. Sometimes as a reminder the dimensions are written to the right of the letter, as in B3x3. The elements of a matrix also have names, usually a lowercase letter the same as the matrix name, with the position of the element written as a subscript. So, for example, the 3x3 matrix A might be written as: 102 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Sometimes you write A = [aij] to say that the elements of matrix A are named aij. Question: (Thought Question:) If two matrices contain the same numbers as elements, are the two matrices equal to each other? Answer: No, to be equal, two matrices must have the same dimensions, and must have the same values in the same positions. Matrix Equality For two matrices to be equal, they must have The same dimensions. Corresponding elements must be equal. In other words, say that An x m = [aij] and that Bp x q = [bij]. Then A = B if and only if n=p, m=q, and aij=bij for all i and j in range.

Here are two matrices which are not equal even though they have the same elements.

Matrix Addition If two matrices have the same number of rows and same number of columns, then the matrix sum can be computed: If A is an MxN matrix, and B is also an MxN matrix, then their sum is an MxN matrix formed by adding corresponding elements of A and B Here is an example of this:

Of course, in most practical situations the elements of the matrices are real numbers with decimal fractions, not the small integers often used in examples. Question: What 3x2 matrix could be added to a second 3x2 matrix without changing that second matrix? Answer: The 3x2 matrix that has all its elements zero. 103 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Zero Matrix A zero matrix is one; which has all its elements zero. Here is a 3x3 zero matrix:

The name of a zero matrix is a boldface zero: 0, although sometimes people forget to make it bold face. Here is an interesting problem:

Question: Form the above sum. No electronic calculators allowed! Answer: Of course, the sum is the same as the non-zero matrix. Rules for Matrix Addition You should be happy with the following rules of matrix addition. In each rule, the matrices are assumed to all have the same dimensions. A+B=B+A A+0=0+A=A 0+0=0 These look the same as some rules for addition of real numbers. (Warning!! Not all rules for matrix math look the same as for real number math.) The first rule says that matrix addition is commutative. This is because ordinary addition is being done on the corresponding elements of the two matrices, and ordinary (real) addition is commutative:

Question: Do you think that (A + B) + C = A + (B + C) Answer: Yes this is another rule that works like real number math. Practice with Matrix Addition Here is another matrix addition problem. Mentally form the sum (or use a scrap of paper):

104 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Hint: this problem is not as tedious as it might at first seem. Question: What is the sum? Answer: Each element of the 3x3 result is 10. Multiplication of a Matrix by a Scalar A matrix can be multiplied by a scalar (by a real number) as follows: To multiply a matrix by a scalar, multiply each element of the matrix by the scalar. Here is an example of this. (In this example, the variable a is a scalar.)

Question: Show the result if the scalar a in the above is the value -1. Answer: Each element in the result is the negative of the original, as seen below. Negative of a Matrix The negation of a matrix is formed by negating each element of the matrix: -A = -1A So, for example:

It will not surprise you that A + (-A) = 0 Question: Look at the above fact. Can you think of a way to define matrix subtraction? Answer: It seems like subtraction could be defined as adding a negation of a matrix. Matrix Subtraction If A and B have the same number of rows and columns, then A - B is defined as A + (-B). Usually you think of this as: To form A - B, from each element of A subtract the corresponding element of B. Here is a partly finished example:

Notice in particular the elements in the first row of the answer. The way the result was calculated for the elements in row 1 column 2 is sometimes confusion. Question: Mentally fill in the two question marks. Answer:

105 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Transpose The transpose of a matrix is a new matrix whose rows are the columns of the original (which makes its columns the rows of the original). Here is a matrix and its transpose:

The superscript "T" means "transpose". Another way to look at the transpose is that the element at row r column c if the original is placed at row c column r of the transpose. We will usually work with square matrices, and it is usually square matrices that will be transposed. However, non-square matrices can be transposed, as well:

Question: What is the transpose of:

Answer:

A Rule for Transpose If a transposed matrix is itself transposed, you get the original back:

This illustrates the rule (AT)T = A. Question: What is the transpose of:

Answer:

106 Copyright Virtual University of Pakistan

10-3D Concepts

VU

The transpose of a row matrix is a column matrix. And the transpose of a column matrix is a row matrix. Rule Summary Here are some rules that cover what has been discussed. You should check that they seem reasonable, rather than memorize them. For each rule the matrices have the same number of rows and columns. A+0=A A + (B + C) = (A + B) + C a0 = 0 (AT)T = A A+B=B+A (ab)A = a(bA) (-1)A = -A 0T = 0 0+0=0 a(A + B) = aB + aA A - A= 0

In the above, a and b are scalars (real numbers). A and B are matrices, and 0 is the zero matrix of appropriate dimension. Question: If A = B and B = C, then does A = C? Answer: Yes Vectors Another important mathematical concept used in graphics is the Vector. If P1 = (x1, y1, z1) is the starting point and P2 = (x2, y2, z2) is the ending point, then the vector V = (x2 x1, y2 y1, z2 z1)

and if P2 = (x2, y2, z2) is the starting point and P1 = (x1, y1, z1) is the ending point, then the vector V = (x1 x2, y1 y2, z1 z2)

107 Copyright Virtual University of Pakistan

10-3D Concepts

VU

This just defines length and direction, but not position. Vector Projections Projection of v onto the x-axis

Projection of v onto the xz plane

2D Magnitude and Direction The magnitude (length) of a vector: |V| = sqrt ( Vx2 + Vy2 ) The equation is derived from the Pythagorean theorem. The direction of a vector: tan = Vy / Vx = tan-1 (Vy / Vx) Where is angular displacement from the x-axis. 3D Magnitude and Direction 3D magnitude is a simple extension of 2D |V| = sqrt( Vx2 + Vy2 + Vz2 ) 108 Copyright Virtual University of Pakistan

10-3D Concepts

VU

3D direction is a bit harder than in 2D. Particularly it needs 2 angles to fully describe direction. Latitude/ longitude is a real-world example. Direction Cosines are often used: , , and are the positive angles that the vector makes with each positive coordinate axes x, y, and z, respectively cos = Vx / |V| cos = Vy / |V| cos = Vz / |V| Vector Normalization Normalizing a vector means shrinking or stretching it so its magnitude is 1. A simple way is normalize by dividing by its magnitude: V = (1, 2, 3) |V| = sqrt( 12 + 22 + 32) = sqrt(14) + 3.74 Vnorm = V / |V| = (1, 2, 3) / 3.74 = (1 / 3.74, 2 / 3.74, 3 / 3.74) = (.27, .53, .80) |Vnorm| = sqrt( .272 + .532 + .802) = sqrt( .9 ) = .95 Note that the last calculation doesnt come out to exactly 1. This is because of the error introduced by using only 2 decimal places in the calculations above. Vector Addition Equation: V3 = V1 + V2 = (V1x + V2x , V1y + V2y , V1z + V2z)

Vector Subtraction Equation: V3 = V1 - V2 = (V1x - V2x , V1y - V2y , V1z - V2z) 109 Copyright Virtual University of Pakistan

10-3D Concepts

VU

Dot Product The dot product of 2 vectors is a scalar

V1 . V2 = (V1x V2x) + (V1y V2y ) + (V1z V2z ) Or, perhaps more importantly for graphics: V1 . V2 = |V1| |V2| cos() where is the angle between the 2 vectors and is in the range 0 Why is dot product important for graphics? It is zero if and only if the 2 vectors are perpendicular cos (90) = 0

110 Copyright Virtual University of Pakistan

10-3D Concepts

VU

The Dot Product computation can be simplified when it is known that the vectors are unit vectors V1 . V2 = cos() because |V1| and |V2| are both 1 Saves 6 squares, 4 additions, and 2 sqrts. Cross Product The cross product of 2 vectors is a vector V1 x V2 = ( V1y V2z - V1z V2y , V1z V2x - V1x V2z , V1x V2y - V1y V2x ) Note that if you are big into linear algebra there is also a way to do the cross product calculation using matrices and determinants Again, just as with the dot product, there is a more graphical definition: V1 x V2 = u |V1| |V2| sin () where is the angle between the 2 vectors and is in the range 0 and u is the unit vector that is perpendicular to both vectors Why u? |V1| |V2| sin() produces a scalar and the result needs to be a vector.

The direction of u is determined by the right hand rule. The perpendicular definition leaves an ambiguity in terms of the direction of u Note that you cant take the cross product of 2 vectors that are parallel to each other Sin (0) = sin (180) = 0 produces the vector (0, 0, 0) Forming Coordinate Systems Cross products are great for forming coordinate system frames (3 vectors that are perpendicular to each other) from 2 random vectors. 1) Cross V1 and V2 to form V3. V3 is now perpendicular to both V1 and V2 2) Cross V2 and V3 to form V4 V4 is now perpendicular to both V2 and V3 Then V2, V4, and V3 form your new frame

111 Copyright Virtual University of Pakistan

10-3D Concepts

VU

V1 and V2 are in the new xy plane

112 Copyright Virtual University of Pakistan

11-2D Transformation I

VU

Lecture No.11

2D Transformations I

In the previous lectures so far we have discussed output primitive as well as filling primitives. With the help of them we can draw an attractive 2D drawing but that will be static whereas in most of the cases we require moving pictures for example games, animation, and different model; where we show certain objects moving or rotating or changing their size. Therefore, changes in orientation that is displacement, rotation or change in size is called geometric transformation. Here, we have certain basic transformations and some special transformation. We start with basic transformation. Basic Transformations Translation Rotation Scaling Above are three basic transformations. Where translation is independent of others whereas rotation and scaling depends on translation in most of cases. We will see how in their respective sections but here we will start with translation. Translation A translation is displacement from original place. This displacement happens to be along a straight line; where two distances involves one is along x-axis that is tx and second is along y-axis that is ty. The same is shown in the figure also we can express it with following equation as well as by matrix: x = x + tx , y = y + ty Here (tx, ty) is translation vector or shift vector. We can express above equations as a single matrix equation by using column vectors to represent coordinate positions and the translation vector: P = P + T Where P = P= T=

Translation is a rigid-body transformation that moves objects without deformation. That is, every point on the object is translated by the same amount. A straight line can be translated by applying the above transformation equation to each of the line endpoints and redrawing the line between the new coordinates. Similarly a 113 Copyright Virtual University of Pakistan

11-2D Transformation I

VU

polygon can be translated by applying the above transformation equation to each vertices of the polygon and redrawing the polygon with new coordinates. Similarly curved objects can be translated. For example to translate circle or ellipse, we translate the center point and redraw the same using new center point. Rotation A two dimensional rotation is applied to an object by repositioning it along a circular path in the xy plane. To rotate a point, its coordinates and rotation angle is required. Rotation is performed around a fixed point called pivot point. In start we will assume pivot point to be the origin or in other words we will find rotation equations for the rotation of object with respect to origin, however later we will see if we change our pivot point what should be done with the same equations. Another thing is to be noted that for a positive angle the rotation will be anti-clockwise where for negative angle rotation will be clockwise. Now for the rotation around the origin as shown in the above figure we required original position/ coordinates which in our case is P(x,y) and rotation angle . Now using polar coordinates assume point is already making angle from origin and distance of point from origin is r, therefore we can represent x and y in the form: x = r cos and y = r sin Now if we want to rotate point by an angle , we have new angle that is (+ ), therefore now point P(x,y) can be represented as: x = r cos( + ) = r cos cos r sin sin and y = r sin( + ) = r cos sin + r sin cos Now replacing r cos = x and r sin = y in above equations we get: x = x cos y sin and y= x sin + y cos Again we can represent above equations with the help of column vectors: P = R . P Where

114 Copyright Virtual University of Pakistan

11-2D Transformation I

VU

When coordinate positions are represented as row vectors instead of column vectors, the matrix product in rotation equation is transposed so that the transformed row coordinate vector [x,y] is calculated as: = (R . P)T = PT . RT T Where P and the other transpose matrix can be obtained by interchanging rows and columns. Also, for rotation matrix, the transpose is obtained by simply changing the sign of the sine terms. Rotation about an Arbitrary Pivot Point: As we discussed above that pivot point may be any point as shown in the above figure, however for the sake of simplicity we assume above that pivot point is at origin. Anyhow, the situation can be dealt easily as we have equations of rotation with respect to origin. We can simply involve another transformation already read that is translation so simply translate pivot point to origin. By translation, now points will make angle with origin, therefore apply the same rotation equations and what next? Simply retranslate the pivot point to its original position that is if we subtract xr,yr now add them therefore we get following equations: x = xr + (x - xr) cos (y - yr) sin y = yr + (x - xr) sin (y - yr) cos As it is discussed in translation rotation is also rigid-body transformation that moves the object along a circular path. Now if we want to rotate a point we already achieved it. But what if we want to move a line along its one end point very simple treat that end point as pivot point and perform rotation on the other end point as discussed above. Similarly we can rotate any polygon with taking some pivot point and recalculating vertices and then redrawing the polygon. Scaling A scaling transformation changes the size of an object. Scaling may be in any terms means either increasing the original size or decreasing the original size. An exemplary scaling is shown in the above figure where scaling factors used Sx=3 and Sy=2. So, what are these scaling factors and how they work very simple, simply we multiply each coordinate with its respective scaling factor. Therefore, scaling with respect to origin is achieved by multiplying x coordinate with factor Sx and y coordinate with Copyright Virtual University of Pakistan Sx=3 Sy=2 115 P T

11-2D Transformation I

VU

factor Sy. Therefore, following equations can be expressed: x = x.Sx y = y.Sy In matrix form it can be expressed as: P = S.P

Now we may have different values for scaling factor. Therefore, as it is multiplying factor therefore, if we have scaling factor > 1 then the object size will be increased than original size; whereas; in reverse case that is scaling factor < 1 the object size will be decreased than original size and obviously there will be no change occur in size for scaling factor equal 1. Two variations are possible in scaling that is having scaling factors to be kept same that is to keep original shape; which is called uniform scaling having Sx factor equal Sy factor. Other possibility is to keep Sx and Sy factor unequal that is called differential scaling and that will alter the original shape that is a square will no more remain square. Now above equation of scaling can be applied to any line, circle and polygon etc. However, as in case of line and polygon we will scale ending points or vertices then redraw the object but in circle or ellipse we will scale the radius. Now coming to the point when scaling with respect to any point other then origin, then same methodology will work that is to apply translation before scaling and retranslation after scaling. So here if we consider fixed/ pivot point (xf,yf), then following equations will be achieved: x = xf + (x - xf)Sx y = yf + (y - yf)Sy These can be rewritten as: x = x. Sx + xf (1 Sx) y = y. Sy + yf (1 Sy) Where the terms xf (1 Sx) and yf (1 Sy) are constant for all points in the object.

116 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

Lecture No.12

2D Transformations II

Before starting our next lecture just recall equations of three basic transformations i.e. translation, rotation and scaling: Translation: P= P + T Rotation: P= R. P

Scaling: P= S. P In many cases of computer graphics applications we require sequence of transformations. For example in animation on each next move we may have object to be translated than scaled. Similarly in games an object in a particular moment may have to be rotated as well as translated. That means we have to perform sequence of matrix operations but the matrix we have seen in the previous lecture have order which restrict them to be operated in sequence. However, with slight reformulated we can bring them into the form where they can easily be operated in any sequence thus efficiency can be achieved. Homogeneous Coordinates Again considering our previous lecture all the three basic transformations covered in that lecture can be expressed by following equation: P= M1.P + M2 With coordinate positions P and P represented as column vectors. Matrix M1 is a 2 by 2 array containing multiplicative factors, and M2 is a two-element column matrix containing translation terms. For translation, M1 is a the identity matrix, For rotation or scaling, M2 contains the translational terms associated with the pivot point or scaling fixed point. To produce a sequence of transformations with these equations, such as scaling followed by rotation then translation, we must calculate the transformed coordinates one step at a time. First, coordinate positions are scaled, then these scaled coordinates are rotated, and finally the rotated coordinates are translated. Now the question is can we find a way to eliminate the matrix addition associated with translation? Yes, we can but for that M1 will have to be rewritten as a 3x3 matrix and also the coordinate positions will have to be expressed as a homogeneous coordinate triple: (x, y) as (xh, yh, h) where xh , y = yh h h We can choose the h as any non-zero value. However, a convenient choice is 1, thus (x, y) has homogeneous coordinates as (x, y, 1). Expressing positions in homogeneous coordinates allows us to represent all geometric transformation equations as matrix 117 Copyright Virtual University of Pakistan x=

12-2D Transformation II

VU

multiplications. Coordinates are represented with three-element column vectors, and transformation operations are written as 3 by 3 matrices. Translation with Homogeneous Coordinates The translation can now be expressed using homogeneous coordinates as:

Abbreviated as: P = T (tx, ty) . P Rotation with Homogeneous Coordinates The rotation can now be expressed using homogeneous coordinates as:

Abbreviated as: P = R () . P Scaling with Homogeneous Coordinates The scaling can now be expressed using homogeneous coordinates as:

Abbreviated as: P = S (Sx, Sy) . P Matrix representations are standard methods for implementing transformations in graphics systems. In many systems, rotation and scaling functions produce transformations with respect to the coordinate origin as expressed in the equation above. Rotations and scaling relative to other reference positions are then handled as a succession of transformation operations. Composite Transformations As in the previous section we achieved homogenous matrices for each of the basic transformation, we can find a matrix for any sequence of transformation as a composite transformation matrix by calculating the matrix product of the individual transformations. 118 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

Translations If two successive translations vectors (tx1, ty1) and (tx2, ty2) are applied to a coordinate position P, the final transformed location P is calculated as P = T(tx2,ty2) . {T(tx1,ty1) . P} = {T(tx2,ty2) . T(tx1,ty1)} . P where P and P are represented as homogeneous-coordinate column vectors. The composite transformation matrix for this sequence of translations is

or T(tx2,ty2) . T(tx1,ty1) = T(tx1 + tx2 , ty1 + ty2) Which means that two successive translations are additive. Hence,

Composite Rotations Two successive Rotations applied to a point P produce the transformed position = R(2) . {R(1) . P} = {R(2) . R(1)} . P By multiplying the two rotation matrices, we can verify that two successive rotations are additive: R(2) . R(1) = R(1 + 2) so that the final rotated coordinates can be calculated with the composite rotation matrix as P = R(1 + 2) . P Composite Scaling Concatenating transformation matrices for two successive scaling operations produces the following composite scaling matrix: P

119 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

or S (sx2,sy2).S(sx1,sy1) = S(sx1.sx2, sy1.sy2) The resulting matrix in the case indicates that successive scaling operations are multiplicative. That is, if we were to triple the size of an object twice in succession, the final size would be nine times that of the original. General Pivot Point Rotation With a graphics package that only provides a rotate function for revolving object about the coordinate origin, we can generate rotations about any selected pivot point (xr, yr) by performing the following sequence of translate-rotate-translate operations: Translate the object so that the pivot-point positions is moved to the coordinate origin Rotate the object about the coordinate origin Translate the object so that the pivot point is returned to its original position

which can be expressed in the form T(xr , yr) . R() . T(-xr ,-yr) = R(xr, yr , ) where T(-xr , -yr) = T-1(xr , yr). General Fixed Point Scaling Following figure is showing a transformation sequence to produce scaling with respect to a selected fixed point (xf, yf) using a scaling function that can only scale relative to the coordinate origin. Translate object so that the fixed point coincides with the coordinate origin Scale the object with respect to the coordinate origin Use the inverse translation of step 1 to return the object to its original position Concatenating the matrices for these three operations produces the required scaling matrix. 120 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

T(xf,yf).S(sx,sy).T(-xf,-yf) = S(xf,yf , sx,sy) This transformation is automatically generated on systems that provide a scale function that accepts coordinates for the fixed point. Concatenation Properties Matrix multiplication is associative. For any three matrices A, B and C, the matrix product A. B. C can b3e performed by first multiplying A and B or by first multiplying B and C: A . B . C = (A . B) . C = A . (B . C) Therefore, we can evaluate matrix products using a left-to-right or a right-to-left associative grouping. On the other hand, transformation products may not be commutative. The matrix product A. B is not equal to B. A, in general. This means that if we want to translate and rotate an object, we must be careful about the order in which the composite matrix is evaluated as show in following figure.

121 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

Reversing the order in which a sequence of transformations is performed may affect the transformed position of an object. In above figure an object is first translated, and then rotated. Whereas, in this figure an object is rotated first, then translated.

For some special cases, such as a sequence of transformations all of same kind, the multiplication of transformation matrices is commutative. As an example, two successive rotations could be performed in either order and the final position would be the same. This commutative property holds also for two successive translations or two successive scalings. Another commutative pair of operation is rotation and uniform scaling (Sx = Sy). General Composite Transformations and Computational Efficiency A general two-dimensional transformation, representing a combination of translations, rotations, and scaling, can be expressed as

The four elements rsij are the multiplicative rotation-scaling terms in the transformation that involve only rotating angles and scaling factors. Elements trsx and trsy are the translational terms containing combinations of translation distances, pivot-point and fixed-point coordinates, and rotation angles and scaling parameters. For example, if an object is to be scaled and rotated about its centroid coordinates (xc, yc) and then translated, the values for the elements of the composite transformation matrix are T (tx,ty) . R(xc,yc, ) . S(xc,yc,sx,sy)

122 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

Although matrix given before above matrix requires nine multiplications and six additions, the explicit calculations for the transformed coordinates are x = x.rsxx + y.rsxy + trsx y = x.rsyx + y.rsyy + trsy Thus, we only need to perform four multiplications and four additions to transform coordinate positions. This is the maximum number of computations required for any transformation sequence, once the individual matrices have been concatenated and the elements of the composite matrix evaluated. Without concatenation, the individual transformations would be applied one at a time and the number of calculations could be significantly increased. An efficient implementation for the transformation operations, therefore, is to formulate transformation matrices, concatenate any transformation sequence, and calculate transformed coordinates using above equations. Other Transformations Basic transformations such as translation, rotation, and scaling are included in most graphics packages. Some packages provide a few additional transformations that are useful in certain applications. Two such transformations are reflection and shear. Reflection A reflection is a transformation that produces a mirror image of an object. The mirror image for a two-dimensional reflection is generated relative to an axis of reflection by rotating the object 180o about the reflection axis. We can choose an axis of reflection in the xy plane or perpendicular to the xy plane. When the reflection axis is a line in the xy plane; the rotation path about this axis is in a plane perpendicular to the xy plane. For reflection axes that are perpendicular to the xy plane, the rotation path in the xy plane. Following are examples of some common reflections. Reflection about the line y=0, the x-axis, relative to axis of reflection can be achieved by rotating the object about axis of reflection by 180o.

1 0 0 0 1 0 0 0 1

123 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

The transformation matrix is

Similarly in case of reflection about y-axis the transformation matrix will be, also the reflection is shown in following figure:

Shear A transformation that distorts the shape of an object such that the transformed shape appears as if the object were composed of internal layers that had been caused to slide over each other is called a shear. Two common shearing transformations are those that shift coordinate x values and those that shift y values.
An x direction shear relative to the x axis is produced with the transformation matrix

124 Copyright Virtual University of Pakistan

12-2D Transformation II

VU

which transforms coordinate position as x = x + shx . y y = y

Any real number can be assigned to the shear parameter shx. A coordinate position (x,y) is then shifted horizontally by an amount proportional to its distance (y value) from the x axis (y=0). Setting shx to 2, for example, changes the square in following figure into a parallelogram. Negative values for shx shift coordinate positions to the left.

Similarly y-direction shear relative to the y-axis is produced with the transformation matrix

and coordinate positions transformed as x = x y = shy . x + y Another similar transformation may be in x and y direction shear, where matrix will be

and coordinate positions transformed as x = x + shx . y y = shy . x + y 125 Copyright Virtual University of Pakistan

13-Drawing Example

VU

Lecture No.13

Drawing Example

Let us now learn some of the implementation techniques. So far we already have done with learning drawing primitives including output primitives as well as filling primitives. Also we have studied transformations. So we should be in position to make use of them in two-dimensional drawing. Though most of you will think that they can draw twodimensional drawing very easily yet it may not be true due to lack of knowledge of some implementation techniques, which are very useful in drawing as well as in transformation. So we will cover this with some examples.

Drawing Table First of all we are going to draw a very simple drawing that is a table. Yes, a simple rectangular table with four legs. So, in order to draw such table we have to draw table top plus four legs connecting four edges of the rectangle. Design Here we will first design the table like an ordinary student. So, what we will do we will see the location of the table. For example assume that our screen has dimensions 640*480 and initially we want to draw table right in the middle of the center. Also, another factor is important that is y axis travels from top to bottom. That is y=0 will be the top edge of the screen and 480 will be the lower edge of the screen. Another thing is the dimension of the table we want to draw. Therefore, we have table that has width 20, length 14 and height 10.
Therefore, we have to find out four vertices that make the corners of the table. So, first of all consider x coordinate. Left edge of the table will be 10 less from the center of the screen that is 320. Therefore, x1 and x4 values will be 310. Similarly right edge of the table will be 10 plus center of the screen. Therefore, x2 and x3 values will be 330. Similarly, top edge of the table will be 10 less from the center of the screen that is 240. Therefore, y1 and y2 values will be 233 and y3 and y4, which are lying on the lower edge, will be 247. Finally, last parameter is required to define the length of legs. Having length of legs we have to simply draw vertical lines of that length starting from each corner respectively. Therefore, following code will be required to draw such a table: void translate(int tx, int ty) { xc+=tx; yc+=ty; x1+=tx; x2+=tx; x3+=tx; x4+=tx; y1+=ty; y2+=ty; y3+=ty; y4+=ty; } 126 Copyright Virtual University of Pakistan

13-Drawing Example

VU

void rotate (float angle) { int tempx=x1; x1=xc+(tempx-xc)*cos(angle)-(y1-yc)*sin(angle); y1=yc+(tempx-xc)*sin(angle)+(y1-yc)*cos(angle); tempx=x2; x2=xc+(tempx-xc)*cos(angle)-(y2-yc)*sin(angle); y2=yc+(tempx-xc)*sin(angle)+(y2-yc)*cos(angle); tempx=x3; x3=xc+(tempx-xc)*cos(angle)-(y3-yc)*sin(angle); y3=yc+(tempx-xc)*sin(angle)+(y3-yc)*cos(angle); tempx=x4; x4=xc+(tempx-xc)*cos(angle)-(y4-yc)*sin(angle); y4=yc+(tempx-xc)*sin(angle)+(y4-yc)*cos(angle); } void scale(int sx, int sy) { x1=xc+(x1-xc)*sx; x2=xc+(x2-xc)*sx; x3=xc+(x3-xc)*sx; x4=xc+(x4-xc)*sx; y1=yc+(y1-yc)*sy; y2=yc+(y2-yc)*sy; y3=yc+(y3-yc)*sy; y4=yc+(y4-yc)*sy; legLength*=sy; } x1=310, x2=330, x3=330, x4=310; y1=233, y2=233, y3=247, y4=247; legLength=10; So, what I want that you should observe the issue. Now consider first of all translation. In translation you have to translate all the points one by one and redraw the picture. Instruction to translate the table will be of the form:

Now in this design drawing is pretty simple. We have to draw 4 lines between corner points. That is from (x1, y1) to (x2, y2), from (x2, y2) to (x3, y3), from (x3, y3) to (x4, y4) and from (x4, y4) to (x1, y1). That will suffice our table top. Next simply draw four lines each starting from one of the corner of the table in the vertical direction having length 10. Now let us see the simple code of drawing such table:
//Table Top line (x1, y1, x2, y2); line (x2, y2, x3, y3); line (x3, y3, x4, y4); line (x4, y4, x1, y1); //Table Legs 127 Copyright Virtual University of Pakistan

13-Drawing Example

VU

line (x1, y1, x1, y1+legLength); line (x2, y2, x2, y2+legLength); line (x3, y3, x3, y3+legLength); line (x4, y4, x4, y4+legLength); Now is not that easy to draw table in the same manner? We will discuss the problem after take a bit look at basic transformations (translation, rotation, scaling). The code is: void translate(int tx, int ty) { xc+=tx; yc+=ty; x1+=tx; x2+=tx; x3+=tx; x4+=tx; y1+=ty; y2+=ty; y3+=ty; y4+=ty; } Now seeing the code you can easily understand the idea. In translation we have to translate all points one by one and then redrawing the table at new calculated points. Later you will see in the other method that translation will only involve one line. Anyhow next we will move to other transformation (rotation). Having pivot point at the center of the screen, we have to perform translation in three steps. That is translation then rotation and then translation. So take a look at the code: void rotate (float angle) { int tempx=x1; x1=xc+(tempx-xc)*cos(angle)-(y1-yc)*sin(angle); y1=yc+(tempx-xc)*sin(angle)+(y1-yc)*cos(angle); tempx=x2; x2=xc+(tempx-xc)*cos(angle)-(y2-yc)*sin(angle); y2=yc+(tempx-xc)*sin(angle)+(y2-yc)*cos(angle); tempx=x3; x3=xc+(tempx-xc)*cos(angle)-(y3-yc)*sin(angle); y3=yc+(tempx-xc)*sin(angle)+(y3-yc)*cos(angle); tempx=x4; x4=xc+(tempx-xc)*cos(angle)-(y4-yc)*sin(angle); y4=yc+(tempx-xc)*sin(angle)+(y4-yc)*cos(angle); } So, here calculations required each time and for each pixel; whereas; you will observe that we can make that rotation pretty simple. A similar problem is lying in the case of scaling that is to perform three steps; translation then scaling and then translation. So look at the code given below: 128 Copyright Virtual University of Pakistan

13-Drawing Example

VU

void scale(int sx, int sy) { x1=xc+(x1-xc)*sx; x2=xc+(x2-xc)*sx; x3=xc+(x3-xc)*sx; x4=xc+(x4-xc)*sx; y1=yc+(y1-yc)*sy; y2=yc+(y2-yc)*sy; y3=yc+(y3-yc)*sy; y4=yc+(y4-yc)*sy; legLength*=sy; } Therefore, same heavy calculations involves in scaling. So, here we will conclude our first method and will start next method so that we can judge how calculations become simple. Now having all discussion on table drawing, let us now consider the complete implementation of class Table: /******************************************************************** Table is designed without considering pivot point simply taking points according to the requirement. Therefore, translation of table involves translation of all points. Scaling and Rotation will be done after translation. ********************************************************************/ #include <graphics.h> #include <iostream.h> #include <conio.h> #include <math.h> float round(float x) { return x+0.5; } class Table { private: int xc, yc;//Center of the figure int xp, yp;//Pivot point for this figure int x1, x2, x3, x4; int y1, y2, y3, y4; int legLength; public: Table() { 129 Copyright Virtual University of Pakistan

13-Drawing Example

VU

xc=320, yc=240;//Center of the figure xp=0; yp=0;//Pivot point for this figure x1=310, x2=330, x3=330, x4=310; y1=233, y2=233, y3=247, y4=247; legLength=10; } void translate(int tx, int ty) { xc+=tx; yc+=ty; x1+=tx; x2+=tx; x3+=tx; x4+=tx; y1+=ty; y2+=ty; y3+=ty; y4+=ty; } void rotate (float angle) { int tempx=x1; x1=xc+(tempx-xc)*cos(angle)-(y1-yc)*sin(angle); y1=yc+(tempx-xc)*sin(angle)+(y1-yc)*cos(angle); tempx=x2; x2=xc+(tempx-xc)*cos(angle)-(y2-yc)*sin(angle); y2=yc+(tempx-xc)*sin(angle)+(y2-yc)*cos(angle); tempx=x3; x3=xc+(tempx-xc)*cos(angle)-(y3-yc)*sin(angle); y3=yc+(tempx-xc)*sin(angle)+(y3-yc)*cos(angle); tempx=x4; x4=xc+(tempx-xc)*cos(angle)-(y4-yc)*sin(angle); y4=yc+(tempx-xc)*sin(angle)+(y4-yc)*cos(angle); } void scale(int sx, int sy) { x1=xc+(x1-xc)*sx; x2=xc+(x2-xc)*sx; x3=xc+(x3-xc)*sx; x4=xc+(x4-xc)*sx; y1=yc+(y1-yc)*sy; y2=yc+(y2-yc)*sy; y3=yc+(y3-yc)*sy; y4=yc+(y4-yc)*sy; legLength*=sy; } 130 Copyright Virtual University of Pakistan

13-Drawing Example

VU

void draw() { line (x1, y1, x2, y2); line (x2, y2, x3, y3); line (x3, y3, x4, y4); line (x4, y4, x1, y1); line (x1, y1, x1, y1+legLength); line (x2, y2, x2, y2+legLength); line (x3, y3, x3, y3+legLength); line (x4, y4, x4, y4+legLength); } }; void main() { clrscr(); int gdriver = DETECT, gmode, errorcode; initgraph(&gdriver, &gmode, "c:\\tc\\bgi"); Table table; table.draw(); setcolor(CYAN); table.translate(15, 25); table.draw(); table.translate(50, 75); table.scale(3,2); table.draw(); table.translate(-100, 75); table.rotate(3.14/4); table.draw(); getch(); closegraph(); } -------------------------------------------------------------------------------------------------------------------------------Second Method Well now we will design our table by considering pivot point. That is we will decide our pivot point and next all points will be taken according to that pivot point. Similarly you will see that this consideration will do a little effect on drawing portion of the code; otherwise all other things will become simpler.

Table Design So let us start with designing the table, that is to calculate parameters of the table. That is 4 corners plus length of legs. First of all we assume that our pivot point is lying on the center of the screen and initially that is (0, 0). So having pivot point we will calculate other points with respect to that point.
131 Copyright Virtual University of Pakistan

13-Drawing Example

VU

So having length 20 units, left edge of the table will be ten digits away from the pivot point and away on the left side; therefore; x1 and x4 will be 10; and similarly right edge of the table will be ten digits away from the pivot point on the right side. Therefore, x2 and x3 will be 10 (yes, positive ten). Now consider the top and lower edges of the table they will be 7 points away from the pivot point in each direction; therefore value of y on the upper edge will be 7 and on the lower edge it will be 7 (yes, positive seven). Now finally y1, y2 will be 7 and y3, y4 will be 7. Well, length of the leg will be simple ten. Therefore, now take a look at the parameters in this design: xc=320, yc=240;//Center of the figure xp=0; yp=0;//Pivot point for this figure x1=-10, x2=10, x3=10, x4=-10; y1=-7, y2=-7, y3=7, y4=7; legLength=10;

Table Drawing So, points x1, x2, x3, and x4 are not having the value at which they will appear on the screen rather they are at the relative distance from the pivot point. Here, we are also using xc, yc that is center on the screen that will keep pivot point align. Now having vertices defined in this fashion our drawing method will be differ from the older one. That is while drawing lines we will add center of the screen and pivot point in each vertex. That will take us to the exact position of the screen. Let us look at the drawing code: int xc=this->xc+xp; int yc=this->yc+yp; line (xc+x1, yc+y1, xc+x2, yc+y2); line (xc+x2, yc+y2, xc+x3, yc+y3); line (xc+x3, yc+y3, xc+x4, yc+y4); line (xc+x4, yc+y4, xc+x1, yc+y1); line (xc+x1, yc+y1, xc+x1, yc+y1+legLength); line (xc+x2, yc+y2, xc+x2, yc+y2+legLength); line (xc+x3, yc+y3, xc+x3, yc+y3+legLength); line (xc+x4, yc+y4, xc+x4, yc+y4+legLength);
So, in the above code first we have added xp to xc in order to reduce some of the calculations required in each line drawing command. Next, we have added that calculated figure to all line drawing commands in order to draw them exactly at the position where it should be appear in the screen. Now having a bit difficulty while drawing there are many more facilities that we will enjoy in especially transformation.

Table Transformation So, first of all consider Translation. In this technique translation is quite simple that is simply add translation vector in the pivot point. All other points will be calculated accordingly. Now look at the very simple code of translation: void translate(int tx, int ty) { xp+=tx; 132 Copyright Virtual University of Pakistan

13-Drawing Example

VU

yp+=ty; } So how simple add tx to xp and ty to yp. Similarly next consider rotation that is again very simple; no need of translation. Let us look at the code: void rotate (float angle) { int tempx=x1; x1=tempx*cos(angle)-y1*sin(angle); y1=tempx*sin(angle)+y1*cos(angle); tempx=x2; x2=tempx*cos(angle)-y2*sin(angle); y2=tempx*sin(angle)+y2*cos(angle); tempx=x3; x3=tempx*cos(angle)-y3*sin(angle); y3=tempx*sin(angle)+y3*cos(angle); tempx=x4; x4=tempx*cos(angle)-y4*sin(angle); y4=tempx*sin(angle)+y4*cos(angle); } So, here you can check that there is no extra calculation simply rotated points are calculated using formula that is used to rotate a point around the origin. Now similarly given below you can see calculations of scaling. void scale(int sx, int sy) { x1=x1*sx; x2=x2*sx; x3=x3*sx; x4=x4*sx; y1=y1*sy; y2=y2*sy; y3=y3*sy; y4=y4*sy; legLength=legLength*sy; } So very simple calculation is done again that is to multiply scaling factor with old vertices and new vertices will be obtained. So, now we look at the class Table in which table is designed considering pivot point and taking all other points accordingly. /******************************************************************** Table is designed with considering pivot point and taking all other points with respect to that pivot point. Therefore, translation of table involves translation of only pivot point, all other points will change respectively. 133 Copyright Virtual University of Pakistan

13-Drawing Example

VU

Scaling and Rotation will be done directly no translation or other transformation is required. ********************************************************************/ #include <graphics.h> #include <iostream.h> #include <conio.h> #include <math.h> float round(float x) { return x+0.5; } class Table { private: int xc, yc;//Center of the figure int xp, yp;//Pivot point for this figure int x1, x2, x3, x4; int y1, y2, y3, y4; int legLength; int sfx, sfy; //Scaling factor public: Table() { xc=320, yc=240;//Center of the figure xp=0; yp=0;//Pivot point for this figure x1=-10, x2=10, x3=10, x4=-10; y1=-7, y2=-7, y3=7, y4=7; legLength=10; sfx=1, sfy=1; } void translate(int tx, int ty) { xp+=tx; yp+=ty; } void rotate (float angle) { int tempx=x1; x1=tempx*cos(angle)-y1*sin(angle); y1=tempx*sin(angle)+y1*cos(angle); tempx=x2; x2=tempx*cos(angle)-y2*sin(angle); y2=tempx*sin(angle)+y2*cos(angle); tempx=x3; 134 Copyright Virtual University of Pakistan

13-Drawing Example

VU

x3=tempx*cos(angle)-y3*sin(angle); y3=tempx*sin(angle)+y3*cos(angle); tempx=x4; x4=tempx*cos(angle)-y4*sin(angle); y4=tempx*sin(angle)+y4*cos(angle); } void scale(int sx, int sy) { x1=x1*sx; x2=x2*sx; x3=x3*sx; x4=x4*sx; y1=y1*sy; y2=y2*sy; y3=y3*sy; y4=y4*sy; legLength=legLength*sy; } void draw() { int xc=this->xc+xp; int yc=this->yc+yp; line (xc+x1, yc+y1, xc+x2, yc+y2); line (xc+x2, yc+y2, xc+x3, yc+y3); line (xc+x3, yc+y3, xc+x4, yc+y4); line (xc+x4, yc+y4, xc+x1, yc+y1); line (xc+x1, yc+y1, xc+x1, yc+y1+legLength); line (xc+x2, yc+y2, xc+x2, yc+y2+legLength); line (xc+x3, yc+y3, xc+x3, yc+y3+legLength); line (xc+x4, yc+y4, xc+x4, yc+y4+legLength); }}; void main() { clrscr(); int gdriver = DETECT, gmode, errorcode; initgraph(&gdriver, &gmode, "c:\\tc\\bgi"); Table table; table.draw(); setcolor(CYAN); table.translate(15, 25); table.draw(); table.translate(50, 0); table.scale(3,2); table.draw(); table.translate(-100, 0); table.rotate(3.14/4); table.draw(); getch(); closegraph(); } 135 Copyright Virtual University of Pakistan

14-Clipping-I

VU

Lecture No.14
Concept

Clipping-I

It is desirable to restrict the effect of graphics primitives to a sub-region of the canvas, to protect other portions of the canvas. All primitives are clipped to the boundaries of this clipping rectangle; that is, primitives lying outside the clip rectangle are not drawn. The default clipping rectangle is the full canvas (the screen), and it is obvious that we cannot see any graphics primitives outside the screen. A simple example of line clipping can illustrate this idea: This is a simple example of line clipping: the display window is the canvas and also the default clipping rectangle, thus all line segments inside the canvas are drawn. The red box is the clipping rectangle we will use later, and the dotted line is the extension of the four edges of the clipping rectangle.

Point Clipping Assuming a rectangular clip window, point clipping is easy. we save the point if:
136 Copyright Virtual University of Pakistan

14-Clipping-I

VU

xmin <= x <=xmax ymin <= y <= ymax

Line Clipping This section treats clipping of lines against rectangles. Although there are specialized algorithms for rectangle and polygon clipping, it is important to note that other graphic primitives can be clipped by repeated application of the line clipper. Clipping Individual Points Before we discuss clipping lines, let's look at the simpler problem of clipping individual points. If the x coordinate boundaries of the clipping rectangle are Xmin and Xmax, and the y coordinate boundaries are Ymin and Ymax, then the following inequalities must be satisfied for a point at (X, Y) to be inside the clipping rectangle: Xmin < X < Xmax
and Ymin < Y < Ymax If any of the four inequalities does not hold, the point is outside the clipping rectangle. Trivial Accept - save a line with both endpoints inside all clipping boundaries. Trivial Reject - discard a line with both endpoints outside the clipping boundaries. For all other lines - compute intersections of line with clipping boundaries. Parametric representation of a line: x = x1 + u (x2 - x1) y = y1 + u (y2 - y1), and 0 <= u <= 1. If the value of u for an intersection with a clipping edge is outside the range 0 to 1, then the line does not enter the interior of the window at that boundary. If the value of u is within this range, then the line does enter the interior of the window at that boundary.

Solve Simultaneous Equations To clip a line, we need to consider only its endpoints, not its infinitely many interior points. If both endpoints of a line lie inside the clip rectangle (eg AB, refer to the first example ), the entire line lies inside the clip rectangle and can be trivially accepted. If one endpoint lies inside and one outside(eg CD), the line intersects the clip rectangle and we must compute the intersection point. If both endpoints are outside the clip rectaangle, the line may or may not intersect with the clip rectangle (EF, GH, and IJ), and we need to perform further calculations to determine whether there are any intersections. The brute-force approach to clipping a line that cannot be trivially accepted is to intersect that line with each of the four clip-rectangle edges to see whether any intersection points lie on those edges; if so, the line cuts the clip rectangle and is partially inside. For each line and clip-rectangle edge, we therefore take the two mathematically infinite lines that contain them and intersect them. Next, we test whether this intersection point is "interior" -- that is, whether it lies within both the clip rectangle edge and the line; if so, there is an intersection with the clip rectangle. In the first example, intersection points G' and H' are interior, but I' and J' are not. 137 Copyright Virtual University of Pakistan

14-Clipping-I

VU

The Cohen-Sutherland Line-Clipping Algorithm The more efficient Cohen-Sutherland Algorithm performs initial tests on a line to determine whether intersection calculations can be avoided. Steps for Cohen-Sutherland algorithm End-points pairs of the line are checked for trivial acceptance or trivial reject using outcode. If not trivial-acceptance or trivial-reject, the line is divided into two segments at a clip edge. Line is iteratively clipped by testing trivial-acceptance or trivial-rejected, and divided into two segments until completely inside or trivial-rejected. Trivial acceptance/reject test To perform trivial accept and reject tests, we extend the edges of the clip rectangle to divide the plane of the clip rectangle into nine regions. Each region is assigned a 4bit code determined by where the region lies with respect to the outside halfplanes of the clip-rectangle edges. Each bit in the outcode is set to either 1 (true) or 0 (false); the 4 bits in the code correspond to the following conditions: Bit 1: outside halfplane of top edge, above top edge Y > Ymax Bit 2: outside halfplane of bottom edge, below bottom edge Y < Ymin Bit 3: outside halfplane of right edge, to the right of right edge X > Xmax Bit 4: outside halfplane of left edge, to the left of left edge X < Xmin Conclusion In summary, the Cohen-Sutherland algorithm is efficient when out-code testing can be done cheaply (for example, by doing bit-wise operations in assembly language) and trivial acceptance or rejection is applicable to the majority of line segments. (For example, large windows - everything is inside, or small windows - everything is outside). Liang-Barsky Algorithm Faster line clippers have been developed that are based on analysis of the parametric equation of a line segment, which we can write in the form:
x = x1 + u x y = y1 + u y, where 0 <= u <= 1 Where x = x2 - x1 and y = y2 - y1. Using these parametric equations, Cryus and Beck developed an algorithm that is generally more efficient than the Cohen-Sutherland algorithm. Later, Liang and Barsky independently devised an even faster parametric lineclipping algorithm. Following the Liang-Barsky approach, we first write the pointclipping in a parametric way: xmin <= x1 + u x <= xmax ymin <= y1 + u y <= ymax of these four inequalities can be expressed as 138 Copyright Virtual University of Pakistan

14-Clipping-I

VU

u * pk <= qk, for k = 1, 2, 3, 4 Where parameters p and q are defined as: p1 = -x, p2 = -x, p3 = -y, p4 = -y, q1 = x1 - xmin q2 = xmax - x1 q3 = y1 - ymin q4 = ymax - y1

Any line that is parallel to one of the clipping boundaries has pk = 0 for the value of k corresponding to that boundary (k = 1, 2, 3, 4 correspond to the left, bottom, and top boundaries, respectively). If, for that value of k, we also find qk >= 0, the line is inside the parallel clipping boundary. When pk < 0, the infinite extension of the line proceeds from the outside to the inside of the infitite extension of the particular clipping boundary. If pk > 0, the line proceeds from the inside to the outside. For a nonzero value of pk = 0, we can calculate the value of u that corresponds to the point where the infinitely extended line intersects the extension of boundary k as: u = qk / pk For each line, we can calculate values for parameters u1 and u2 that defines that part of the line that lies within the clip rectangle. The value of u1 is determined by looking at the rectangle edges for which the line proceeds from the outer side to the inner side. (p < 0). For these edges we calculate rk = qk / pk. The value of u1 is taken as the largest of the set consisting of o and the various values of r. Conversely, the value of u2 is determined by examining the boundaries for which the line proceeds from inside to outside (p > o). A value of rk is calculated for each of these boundaries and the value of u2 is the minimum of the set consisting of 1 and the calculated r values. If u1 > u2, the line is completely outside the clip window and it can be rejected. Otherwise, the end points of the clipped line are calculated from the two values of parameter u. This algorithm is presented in the following procedure. Line intersection parameters are initialized to the values u1 = 0 and u2 = 1. For each clipping boundary, the appropriate values for p and q are calculated and used by the function clipTest to determine whether the line can be rejected or whether the intersection parameters are to be adjusted. When p < 0, the parameter r is used to update u1; when p < 0, the parameter r is used to update u2. If updating u1 or u2 results in u1 > u2, we reject the line. Otherwise, we update the appropriate u parameter only if the new value results in a shortening of the line.

139 Copyright Virtual University of Pakistan

14-Clipping-I

VU

When p = 0 and q < 0, we can discard the line since it is parallel to and outside of this boundary. If the line has not been rejected after all four values of p and q have been tested, the endpoints of the clipped line are determined from values of u1 and u2.

Conclusion
In general, the Liang-Barsky algorithm is more efficient than the Cohen Sutherland algorithm, since intersection calculations are reduced. Each update of parameters u1 and u2 requires only one division; and window intersections of the line are computed only once, when the final values of u1 and u2 have computed. In contrast, the CohenSutherland algorithm can repeatedly calculate intersections along a line path, even though the line may be completely outside the clip window, and, each intersection calculation requires both a division and a multiplication. Both the Cohen Sutherland and the Liang Barsky algorithms can be extended to three-dimensional clipping.

140 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Lecture No.15

Clipping-II

Polygon Clipping A polygon is usually defined by a sequence of vertices and edges. If the polygons are unfilled, line-clipping techniques are sufficient however, if the polygons are filled, the process in more complicated. A polygon may be fragmented into several polygons in the clipping process, and the original colour associated with each one. The SutherlandHodgeman clipping algorithm clips any polygon against a convex clip polygon. The Weiler-Atherton clipping algorithm will clip any polygon against any clip polygon. The polygons may even have holes. An algorithm that clips a polygon must deal with many different cases. The case is particularly note worthy in that the concave polygon is clipped into two separate polygons. All in all, the task of clipping seems rather complex. Each edge of the polygon must be tested against each edge of the clip rectangle; new edges must be added, and existing edges must be discarded, retained, or divided. Multiple polygons may result from clipping a single polygon. We need an organized way to deal with all these cases. The following example illustrates a simple case of polygon clipping.

Given below are some examples to elaborate further.

141 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Polygon clipping - disjoint polygons.

Polygon clipping - disjoint polygons.

142 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Polygon clipping - open polygons.

Polygon clipping - open polygons.

143 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Polygon clipping - open polygons.

Sutherland and Hodgman's polygon-clipping algorithm Sutherland and Hodgman's polygon-clipping algorithm uses a divide-and-conquer strategy: It solves a series of simple and identical problems that, when combined, solve the overall problem. The simple problem is to clip a polygon against a single infinite clip edge. Four clip edges, each defining one boundary of the clip rectangle, successively clip a polygon against a clip rectangle.
Note the difference between this strategy for a polygon and the Cohen-Sutherland algorithm for clipping a line: The polygon clipper clips against four edges in succession, whereas the line clipper tests outcode to see which edge is crossed, and clips only when necessary.

Steps of Sutherland-Hodgman's polygon-clipping algorithm Polygons can be clipped against each edge of the window one at a time. Windows/edge intersections, if any, are easy to find since the X or Y coordinates are already known. Vertices which are kept after clipping against one window edge are saved for clipping against the remaining edges. Note that the number of vertices usually changes and will often increase. We are using the Divide and Conquer approach. Here is a STEP-BY-STEP example of polygon clipping. Four Cases of polygon clipping against one edge The clip boundary determines a visible and invisible region. The edges from vertex i to vertex i+1 can be one of four types: Case 1 : Wholly inside visible region - save endpoint Case 2 : Exit visible region - save the intersection Case 3 : Wholly outside visible region - save nothing Case 4 : Enter visible region - save intersection and endpoint 144 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Because clipping against one edge is independent of all others, it is possible to arrange the clipping stages in a pipeline. The input polygon is clipped against one edge and any points that are kept are passed on as input to the next stage of the pipeline. In this way four polygons can be at different stages of the clipping process simultaneously. This is often implemented in hardware. Example No # 1 Clipping a Polygon

Original polygon

145 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Clip Left

Clip Right

146 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Clip Bottom

Clip Top Example No #2 Clipping a Rectangle If Clipping Rectangle is denoted by dashed lines and Line is defined by using points P1 and P2 Case i

147 Copyright Virtual University of Pakistan

15-Clipping-II

VU

For each boundary b in [ L(Left), R(Right), T(Top), B(Bottom) ] If P1 outside and P2 inside. Output: intersection Point (P1) Point P2 Case ii

For each boundary b in [ L(Left), R(Right), T(Top), B(Bottom) ] If P1 inside and P2 inside Output: Point P2

148 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Case iii

For each boundary b in [ L(Left), R(Right), T(Top), B(Bottom) ] If P1 outside and P2 outside Do nothing Case iv

For each boundary b in [ L(Left), R(Right), T(Top), B(Bottom) ] If P1 inside and P2 outside (We are going from P1 to P2) Output: Point of intersection (P2) only. 149 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Pipeline Clipping Approach An array, s records the most recent point that was clipped for each clip-window boundary. The main routine passes each vertex p to the clipPoint routine for clipping against the first window boundary. If the line defined by endpoints p and s (boundary) crosses this window boundary, the intersection is calculated and passed to the next clipping stage. If p is inside the window, it is passed to the next clipping stage. Any point that survives clipping against all window boundaries is then entered into the output array of points. The array firstPoint stores for each window boundary the first point flipped against that boundary. After all polygon vertices have been processed, a closing routine clips lines defined by the first and last points clipped against each boundary. Shortcoming of Sutherlands -Hodgeman Algorithm Convex polygons are correctly clipped by the Sutherland-Hodegeman algorithm, but concave polygons may be displayed with extraneous lines. This occurs when the clipped polygon should have two or more separate sections. But since there is only one output vertex list, the last vertex in the list is always joined to the first vertex. There are several things we could do to correct display concave polygons. For one, we could split the concave polygon into two or more convex polygons and process each convex polygon separately.
Another approach to check the final vertex list for multiple vertex points along any clip window boundary and correctly join pairs of vertices. Finally, we could use a more general polygon clipper, such as wither the Weiler-Atherton algorithm or the Weiler algorithm described in the next section.

Weiler-Atherton Polygon Clipping

In this technique, the vertex-processing procedures for window boundaries are modified so that concave polygons are displayed correctly. This clipping procedure was developed as a method for identifying visible surfaces, and so it can be applied with arbitrary polygon-clipping regions. The basic idea in this algorithm is that instead of always proceeding around the polygon edges as vertices are processed, we sometimes want to follow the window boundaries. Which path we follow depends on the polygon-processing direction(clockwise or counterclockwise) and whether the pair of polygon vertices currently being processed represents an outside-to-inside pair or an inside-to-outside pair. For clockwise processing of polygon vertices, we use the following rules:

For an outside-top inside pair of vertices, follow the polygon boundary For an inside-to-outside pair of vertices, follow the window boundary in a clockwise direction

In following figure, the processing direction in the Wieler-Atherton algorithm and the resulting clipped polygon is shown for a rectangular clipping window.

150 Copyright Virtual University of Pakistan

15-Clipping-II

VU

Inside

go up 1

5 8 go up 6

151 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Lecture No.16

3D Concepts

Welcome! You are about to embark on a journey into the wondrous world of threedimensional computer graphics. Before we take the plunge into esoteric 3D jargon and mathematical principles (as we will in the next lectures), lets have a look at what the buzzword 3D actually means. We have heard the term 3D applied to everything from games to the World Wide Web to Microsofts new look for Windows XP. The term 3D is often confusing because games (and other applications) which claim to be 3D, are not really 3D. In a 3D medium, each of our eyes views the scene from slightly different angles. This is the way we perceive the real world. Obviously, the flat monitors most of us use when playing 3D games 3D applications cant do this. However, some Virtual Reality (VR) glasses have this capability by using a separate TV-like screen for each eye. These VR glasses may become common place some years from now, but today, they are not the norm. Thus, for present-day usage, we can define 3D to mean something using a three-dimensional coordinate system. A three-dimensional coordinate system is just a fancy term for a system that measures objects with width, height, and depth (just like the real world). Similarly, 2-dimensional coordinate systems measure objects with width and height --- ignoring depth properties (so unlike the real world).

152 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Shadow of a 3D object on paper 16.1 Coordinate Systems Coordinate systems are the measured frames of reference within which geometry is defined, manipulated and viewed. In this system, you have a well-known point that serves as the origin (reference point), and three lines(axes) that pass through this point and are orthogonal to each other ( at right angles 90 degrees). With the Cartesian coordinate system, you can define any point in space by saying how far along each of the three axes you need to travel in order to reach the point if you start at the origin. Following are three types of the coordinate systems. a) 1-D Coordinate Systems:

This system has the following characteristics:

Direction and magnitude along a single axis, with reference to an origin Locations are defined by a single coordinate Can define points, segments, lines, rays Can have multiple origins (frames of reference) and transform coordinates among them

153 Copyright Virtual University of Pakistan

16-3D Concepts

VU

b) 2-D Coordinate Systems:

Direction and magnitude along two axes, with reference to an origin Locations are defined by x, y coordinate pairs Can define points, segments, lines, rays, curves, polygons, (any planar geometry) Can have multiple origins (frames of reference and transform coordinates among them

c) 3-D Coordinate Systems:

3D Cartesian coordinate systems Direction and magnitude along three axes, with reference to an origin Locations are defined by x, y, z triples Can define cubes, cones, spheres, etc., (volumes in space) in addition to all oneand two-dimensional entities Can have multiple origins (frames of reference) and transform coordinates among them

154 Copyright Virtual University of Pakistan

16-3D Concepts

VU

16.2

Left-handed versus Right-handed

Determines orientation of axes and direction of rotations Thumb = pos x, Index up = pos y, Middle out = pos z Most world and object axes tend to be right handed Left handed axes often are used for cameras

a) Right Handed Rule:

Right Hand Rule for rotations: grasp axis with right hand with thumb oriented in positive direction, fingers will then curl in direction of positive rotation for that axis.

Right handed Cartesian coordinate system describes the relationship of the X,Y, and Z in the following manner: 155 Copyright Virtual University of Pakistan

16-3D Concepts

VU

X is positive to the right of the origin, and negative to the left. Y is positive above the origin, and negative below it. Z is negative beyond the origin, and positive behind it.

Origin

+Y North -Z

+X
West

East

+Z Sout

b) Left Handed Rule: Origin +Y North +Z

+X
West

East

-Z Sout

156 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Left handed Cartesian coordinate system describes the relationship of the X, Y and Z in the following manner: X is positive to the right of the origin, and negative to the left. Y is positive above the origin, and negative below it. Z is positive beyond the origin, and negative behind it.

Defining 3D points in mathematical notations

3D points can be described using simple mathematical notations P = (X, Y, Z) Thus the origin of the Coordinate system is located at point (0,0,0), while five units to the right of that position might be located at point (5,0,0).

Y-up versus Z-up:

z-up typically used by designers y-up typically used by animators orientation by profession supposedly derives from past work habits often handled differently when moving from application to application

157 Copyright Virtual University of Pakistan

16-3D Concepts

VU

16.3

Global and Local Coordinate Systems:

Local coordinate systems can be defined with respect to global coordinate system Locations can be relative to any of these coordinate systems Locations can be translated or "transformed" from one coordinate system to another.

16.4

Multiple Frames of Reference in a 3-D Scene:

In fact, there usually are multiple coordinate systems within any 3-D screen

158 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Application data will be transformed among the various coordinate systems, depending on what's to be accomplished during program execution Individual coordinate systems often are hierarchically linked within the scene

16.5

Defining points in C language structure

You can now define any point in the 3D by saying how far east, up, and north it is from your origin. The center of your computer screen ? it would be at a point such as 1.5 feet east, 4.0 feet up, 7.2 feet north. Obviously, you will want a data structure to represent these points. An example of such a structure is shown in this code snippet: typedef struct _POINT3D { float x; float y; float z; }POINT3D; POINT3D screenCenter = {1.5, 4.0, 7.2};

16.6

The Polar Coordinate System

Cartesian systems are not the only ones we can use. We could have also described the object position in this way: starting at the origin, looking east, rotate 38 degrees northward, 65 degrees upward, and travel 7.47 feet along this line. As you can see, this is less intuitive in a real world setting. And if you try to work out the math, it is harder to manipulate (when we get to the sections that move points around). Because such polar coordinates are difficult to control, they are generally not used in 3D graphics.

16.7

Using Multiple Coordinate Systems

As we start working with 3D objects, you may find that it is more efficient to work with groups of points instead of individual single points. For example, if you want to model your computer, you may want to store it in a structure such as that shown in this code snippet: typedef struct _CPU{ POINT3D center; // the center of the CPU, in World coordinates POINT3D coord[8]; // the 8 corners of the CPU box relative to the center point }CPU; In next lectures we will learn how we can show 3D point on 2D computer screen. 16.8 Defining Geometry in 3-D Here are some definitions of the technical names that will be used in 3D lectures.

Modeling: is the process of describing an object or scene so that we can construct an image of it. Points & Polygons:

159 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Points: three-dimensional locations (or coordinate triples)

Vectors: - have direction and magnitude; can also be thought of as displacement

Polygons: - sequences of correctly co-planar points; or an initial point and a sequence of vectors

Primitives Primitives are the fundamental geometric entities within a given data structure. 160 Copyright Virtual University of Pakistan

16-3D Concepts

VU

We have already touched on point, vector and polygon primitives

Regular Polygon Primitives - square, triangle, circle, n-polygon, etc.

Polygon strips or meshes Meshes provide a more economical description than multiple individual polygons For example, 100 individual triangles, each requiring 3 vertices, would require 100 x 3 or 300 vertex definitions to be stored in the 3-D database. By contrast, triangle strips require n + 2 vertex definitions for any n number or triangles in the strip. Hence, a 100 triangle strip requires only 102 unique vertex definitions.

Meshes also provide continuity across surfaces which is important for shading calculations

3D primitives in a polygonal database

161 Copyright Virtual University of Pakistan

16-3D Concepts

VU

3D shapes are represented by polygonal meshes that define or approximate geometric surfaces.

With curved surfaces, the accuracy of the approximation is directly proportional to the number of polygons used in the representation. More polygons (when well used) yield a better approximation. But more polygons also exact greater computational overhead, thereby degrading interactive performance, increasing render times, etc.

Rendering - The process of computing a two dimensional image using a combination of a three-dimensional database, scene characteristics, and viewing transformations. Various algorithms can be employed for rendering, depending on the needs of the application. Tessellation - The subdivision of an entity or surface into one or more non-overlapping primitives. Typically, renderers decompose surfaces into triangles as part of the rendering process. Sampling - The process of selecting a representative but finite number of values along a continuous function sufficient to render a reasonable approximation of the function for the task at hand. Level of Detail (LOD) - To improve rendering efficiency when dynamically viewing a scene, more or less detailed versions of a model may be swapped in and out of the scene database depending on the importance (usually determined by image size) of the object in the current view.

162 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Polygons and rendering

Clockwise versus counterclockwise

Surface normal - a vector that is perpendicular to a surface and outward facing

Surface normals are used to determine visibility and in the calculation of shading values (among other things)

Convex versus concave A shape is convex if any two points within the shape can be connected with a straight line that never goes out of the shape. If not, the shape is concave. Concave polygons can cause problems during rendering (e.g. tears, etc., in apparent surface).

163 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Polygon meshes and shared vertices

Polygons consisting of non-co-planar vertices can cause problems when rendering (e.g. visible tearing of the surface, etc.) With quad meshes, for example, vertices within polygons can be inadvertently transformed into non-co-planer positions during modeling or animation transformations. With triangle meshes, all polygons are triangles and therefore all vertices within any given polygon will be coplanar.

With polygonal databases:

Explicit, low-level descriptions of geometry tend to be employed Object database files can become very large relative to more economical, higher order descriptions. Organic forms or free-form surfaces can be difficult to model.

16.9

Surface models
164 Copyright Virtual University of Pakistan

Here is brief over view of surface models:

16-3D Concepts

VU

Surfaces can be constructed from mathematical descriptions Resolution independent - surfaces can be tessellated at rendering with an appropriate level of approximation for current display devices and/or viewing parameters Tessellation can be adaptive to the local degree of curvature of a surface.

Primitives

Free-form surfaces can be built from curves Construction history, while also used in polygonal modeling, can be particularly useful with curve and surface modeling techniques.

165 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Parameterization

Curve direction and surface construction

Surface parameterization (u, v, w) are used

o o

For placing texture maps, etc. For locating trimming curves, etc. 166 Copyright Virtual University of Pakistan

16-3D Concepts

VU

Metaballs (blobby surfaces)

Potential functions (usually radially symmetric Gaussian functions) are used to define surfaces surrounding points

Lighting Effects

Texture Mapping: The texture mapping is of the following types that we will be studying in our coming lectures on 3D: 1. 2. 3. 4. 5. 6. 7. Perfect Mapping: Affine Mapping Area Subdivision Scan-line Subdivision Parabolic Mapping Hyperbolic Mapping Constant-Z Mapping 167 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

Lecture No.17
Definition of a 3D Point

3D Transformations I

A point is similar to its 2D counterpart; we simply add an extra component, Z, for the 3rd axis:

Points are now represented with 3 numbers: <x, y, z>. This particular method of representing 3D space is the "left-handed" coordinate system. In the left-handed system the x axis increases going to the right, the y axis increases going up, and the z axis increases going into the page/screen. The right-handed system is the same but with the zaxis pointing in the opposite direction.

Distance between Two 3D Points

The distance between two points <Ax,Ay,Az> and <Bx,By,Bz> can be found by again using the Pythagoras theorem:

dx = Ax-Bx dy = Ay-By dz = Az-Bz distance = sqrt(dx*dx + dy*dy + dz*dz) Definition of a 3D Vector

Like it's 2D counterpart, a vector can be thought of in two ways: either a point at <x,y,z> or a line going from the origin <0,0,0> to the point <x,y,z>. 3D Vector addition and subtraction is virtually identical to the 2D case. You can add a 3D vector <vx,vy,vz> to a 3D point <x,y,z> to get the new point <x',y',z'> like so:

x' = x + vx y' = y + vy z' = z + vz

Vectors themselves can be added by adding each of their components, or they can be multiplied (scaled) by multiplying each component by some constant k (where k <> 0). Scaling a vector by 2 (say) will still cause the vector to point in the same direction, but it will now be twice as long. Of course you can also divide the vector by k (where k <> 0) to get a similar result. To calculate the length of a vector we simply calculate the distance between the origin and the point at <x, y, z>: 168 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

Length = | <x,y,z> - <0,0,0> | = sqrt( (x-0)*(x-0) + (y-0)*(y-0) + (z-0)*(z-0) ) = sqrt(x*x + y*y + z*z) Unit Vector
Often in 3D computer graphics you need to convert a vector to a unit vector, ie a vector that points in the same direction but has a length of 1. This is done by simply dividing each component by the length:

Let <x,y,z> be our vector, length = sqrt(x*x + y*y + z*z) Unit vector = <x,y,z> = | x , y , z | length | length length length |
(Where length = |<x,y,z>|) Note that if the vector is already a unit vector then the length will be 1, and the new values will be the same as the old.

Definition of a Line
As in 2D, we can represent a line by it's endpoints (P1 and P2) or by the parametric equation:

P = P1 + k * (P2-P1)
Where k is some scalar value between 0 and 1

Transformations:
A static set of 3D points or other geometric shapes on screen is not very interesting. You could just use a paint program to produce one of these. To make your program interesting, you will want a dynamic landscape on the screen. You want the points to move in the world coordinate system, and you even want the point-of-view (POV) to move. In short, you want to model the real world. The process of moving points in space is called transformation, and can be divided into translation, rotation and other kind of transformations.

Translation Translation is used to move a point, or a set of points, linearly in space, for example, you may want to move a point 3 meters east, -2 meters up, and 4 meters north. Looking at this textual description, you might think that this looks very much like a Point3D, and you would be close. But the above does not require one critical piece of information: it does not reference the origin. The above only encapsulates direction and distance, not an absolute point in space. This called a vector and can be represented in a structure identical to Point3D:
struct Vector3D float x; float y; float z; end struct distance along x axes distance along y axes distance along z axes 169 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

Vector Addition
You translate a point by adding a vector to it; you add points and vectors by adding the components piecewise: Point3D point = {0, 0, 0} Vector3D vector = {10, -3, 2.5 } Adding vector to point point.x = point.x + vector.x; point.y = point.y + vector.y; point.z = point.z + vector.z; Point will be now at the absolute point < 10,-3 2.5>. you could move it again: point.x = point.x + vector.x; point.y = point.y + vector.y; point.z = point.z + vector.z; And point would now be at the absolute point <20, -6, 5>. In pure mathematical sense, you cannot add two points together such an operation makes no sense (what is Lahore plus Karachi?). However, you can subtract a point from another in order to uncover the vector that would have to be added to the first to translate it into the second: Point3D p1,p2 Vector3D v; Set p1 and p2 to the desired points v.x = p2.x p1.x v.y = p2.y p1.y v.z = p2.z p1.z Now you can add v to p1, you would translate it into the point p2. The following lists the operations you can do between points and vectors: point point => vector point + point = point - ( - point) => vector vector vector => vector vector + vector => vector point vector = point + (-vector) => point point + vector => point

170 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

Multiplying: Scalar Multiplication

Multiplying a vector by a scalar ( a number with no units), and could be coded with: Vector.x = Vector.y = Vector.z = Vector.x * scalarValue Vector.y * scalarValue Vector.z * scalarValue

If you had a vector with a length of 4 and multiplied it by 2.5, you would end up with a vector of length 10 that points in the same direction the original vector pointed. If you multiplied by -2.5 instead, you would still end up with a vector of length 10; but now it would be pointing in the opposite direction of the original vector.

Multiplying: Vector Multiplication

You can multiply with vectors two other ways; both involve multiplying a vector by a vector.

Dot Product
The dot product of two vectors is defined by the formula: Vector A, B

A * B = A.x * B.x + A.y * B.y + A.z * B.z The result of a dot product is a number and has units of As units times Bs units. Thus, if you calculate the dot product for two vectors that both use feet for units, your answer will be in square feet. However, in 3D graphics we usually ignore the units and just treat it like a scalar. Consider the following definition of the dot product that is used by physicists (instead of mathematicians): A * B = |A| * |B| * cos(theta) Where theta is the angle between the two vectors Remember that |v| represents the length of vector V and is a non-negative number; we can replace the vector lengths above and end up with: K = |A| * |B| (therefore k > = 0) A * B = K * cos (theta) Therefore: 171 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

A * B => cos(theta) Where => means directly correlates to. Now, if you remember, the cos(theta) function has the following properties: cos(theta) > 0 iff theta is less than 90 degrees or greater than 270 degrees cos(theta) < iff theta is greater than 90 degrees and less than 270 degrees cos(theta) = 0 iff theta is 90 degrees or 270 degrees We can extend this to the dot product of two vectors, since it directly correlates to the angle between the two vectors: A*B > 0 iff the angle between them is less than 90 or greater than 270 degrees A*B < 0 iff the angle between them is greater than 90 and less than 270 degrees A*B = 0 iff the angle between them is 90 or 270 degrees (they are orthogonal).

Use of Dot Product

Assume you have a point of view at < px,py,pz>. It is looking along the vector <vx,vy,vz>, and you have a point in space <x,y,z> you want to know if the pointof-view can possible see the point, of if the point is behind the POV, as shown in figure.
Direction of View <vx,vy,vz>

Point of View <px,py,pz> Test vector <tx, ty, tz> Point <x, y, z>

Point3D pov; Vector3D povDir; Point3D test; Vector3D vTest float dotProduct; vTest.x = pov.x test.x; vTest.y = pov.y test.y; vTest.z = pov.z test.z; 172 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

dotProduct == vTest.x*povDir.x + vTest.y*povDir.y + vTest.z * povDir.z; if(dotProduct > 0) point is in front of POV else if (dotProduct < 0) point is behind POV else point is orthogonal to the POV direction

Cross Product
Another kind of multiplication that you can do with vectors is called the cross product this is defined as: Vector A, B A X B = < A.y * B.z A.z * B.y, A.z * B.x A.x * B.z, A.x * B.y A.y * B.x > For physicists: |A x B| = |A| * |B| sin(theta) Where theta is the angle between the two vectors. The above formula for A x B came from the determinate of order 3 of the matrix: | X |A.x |B.x Y A.y B.y Z| A.z| B.z|

Transformations
The process of moving points in space is called transformation.

Types of Transformation
There are various types of transformations as we have seen in case of 2D transformations. These include: a) Translation b) Rotation c) Scaling d) Reflection e) Shearing

Translation
Translation is used to move a point, or a set of points, linearly in space. Since now we are talking about 3D, therefore each point has 3 coordinates i.e. x, y and z. similarly, the translation distances can also be specified in any of the 3 dimensions. These Translation Distances are given by tx, ty and tz. For any point P(x,y,z) after translation we have P(x,y,z) where x = x + tx , y = y + ty , 173 Copyright Virtual University of Pakistan

17-3D Transformations I

VU

z = z + tz and (tx, ty , tz) is Translation vector Now this can be expressed as a single matrix equation: P = P + T Where: x P= y z x ' P'= y ' z ' t x T = t y t z

3D Translation Example We may want to move a point 3 meters east, -2 meters up, and 4 meters north. What would be done in such event? Steps for Translation Given a point in 3D and a translation vector, it can be translated as follows:
Point3D point = (0, 0, 0) Vector3D vector = (10, -3, 2.5) Adding vector to point point.x = point.x + vector.x; point.y = point.y + vector.y; point.z = point.z + vector.z; And finally we have translated point.

Homogeneous Coordinates
Analogous to their 2D Counterpart, the homogeneous coordinates for 3D translation can be expressed as : x ' 1 y ' 0 = z ' 0 1 0 0 0 tx x 1 0 ty . y 0 1 tz z 0 0 1 1

Abbreviated as: P = T (tx, ty, tz). P On solving the RHS of the matrix equation, we get: x ' x + tx y ' y + t y = z ' z + tz 1 1 Which shows that each of the 3 coordinates gets translated by the corresponding translation distance. 174 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Lecture No.18

3D Transformations II

Rotation Rotation is the process of moving a point in space in a non-linear manner. More particularly, it involves moving the point from one position on a sphere whose center is at the origin to another position on the sphere. Why would you want to do something like this? As we will show in later section, allowing the point of view to move around is only an illusion projection requires that the POV be at the origin. When the user thinks the POV is moving, you are actually translating all your points in the opposite direction; and when the user thinks the POV is looking down a new vector, you are actually rotating all the points in the opposite direction; and when the user thinks the POV is looking down a new vector, you are actually rotating all the points in the opposite direction. Normalization: Note that this process of moving your points so that your POV is at the origin looking down the +Z axis is called normalization.
Rotation a point requires that you know the coordinates for the point, and That you know the rotation angles. You need to know three different angles: how far to rotate around the X axis( YZ rotation, or pitch); how far to rotate around the Y axis (XZ plane, or yaw); and how far to rotate around the Z axis (XY rotation, or roll). Conceptually, you do the three rotations separately. First, you rotate around one axis, followed by another, then the last. The order of rotations is important when you cascade rotations; we will rotate first around the Z axis, then around the X axis, and finally around the Y axis. To show how the rotation formulas are derived, lets rotate the point <x,y,z> around the Z axis with an angle of degrees. ROLL:-

If you look closely, you should note that when we rotate around the Z axis, the Z element of the point does not change. In fact, we can just ignore the Z we already know what it will be after the rotation. If we ignore the Z element, then we have the same case as if we were rotating the two-dimensional point <x,y> through the angle . 175 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

This is the way to rotate a 2-D point. For simplicity, consider the pivot at origin and rotate point P (x,y) where x = r cos and y = r sin If rotated by then: x = r cos( + ) = r cos cos r sin sin and y = r sin( + ) = r cos sin + r sin cos

Replacing r cos with x and r sin with y, we have: x = x cos y sin and y = x sin + y cos and z = z (as it does not change when rotating around z-axis)

Now for rotation around other axes, cyclic permutation helps form the equations for yaw and pitch as well: X Y In the above equations replacing x with y and y with z gives equations for rotation around x-axis. Now in the modified equations if we replace y with z and z with x then we get the equations for rotation around y-axis. 176 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Rotation about x-axis (i.e. in yz plane): x =x y = y cos z sin z = y sin + z cos Rotation about y-axis (i.e. in xz plane): x = z sin + x cos y =y z = z cos x sin

Using Matrices to create 3D A matrix is usually defined as a two-dimensional array of numbers. However, I think you will find it much more useful to think of a matrix as an array of vectors. When we talk about vectors, what it really mean is an ordered set of numbers ( a tuple in mathematics terms). We can use 3D graphics vectors and points interchangeably for this, since they are both 3-tuples ( or triples).
In general we work with square matrices. This means that the number of vectors in the matrix is the same as the number of elements in the vectors that comprise it. Mathematically, we show a matrix as a 2-D array of numbers surrounded by vertical lines. For example: |x1 |x2 |x3 y1 y2 y3 z1| z2| z3|

we designate this as a 3*3 matrix ( the first 3 is the number of rows, and the second 3 is the number of columns). The rows of the matrix are the horizontal vectors that make it up; in this case, <x1, y1,z1>, <x2,y2,z2>, and <x3,y3,z3>. In mathematics, we call the vertical vectors columns. In this case they are < x1,x2,x3>, <y1,y2,y3> and <z1,z2,z3>. The most important thing we do with a matrix is to multiply it by a vector or another matrix. We follow one simple rule when multiplying something by a matrix: multiply each column by a multiplicand and store this as an element in the result. Now as I said 177 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

earlier, you can consider each column to be a vector, so when we multiply by a matrix, we are just doing a bunch of vector multiplies. So which vector multiply do you use-the dot product, or the crosss product? You use the dot product. We also follow on simple rule when multiplying a matrix by something: mubliply each ro by the multiplier. Again, rows are just vectors, and the type of ultiplicaiton is the dot product. Lets look at some examples. First, lets assume that I have a matrix M, and I want to multiply it by a point < x,y,z>, the first ting I know is that the vector rows of the matrix must contain three elements (in other words, three columns). Why ? because I have to multiply those rows by my point using a dot product, and to do that, the two vectors must have the same number of element. Since I am going to get dot product for each row in M, I will end up with a tuple that has one element for each row in M. as I stated earlier, we work almost exclusively with square matrices, since I must have three columns, M will also have three rows. Lets see: |1 0 0| < x,y,z> * |0 1 0| ={<x,y,z>*< 1,0,0> ,<x,y,z><0,1,0>,<x,y,z> *<0,0,1>}={ x,y,z} |0 0 1|

Using Matrices for Rotation

Roll (rotate about the Z axis): x' cos y ' sin = z ' 0 1 0 sin cos 0 0 0 0 x 0 0 . y 1 0 z 0 1 1 0 x 0 . y 0 z 1 1

Pitch (rotate about the X axis): 0 x' 1 y ' 0 cos = z ' 0 sin 0 1 0 0 sin cos 0

Yaw (rotate about the Y axis):

x' cos y ' 0 = z ' sin 1 0 0 sin 1 0 0 cos 0 0 0 x 0 . y 0 z 1 1

178 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Example: To show this happening, let's manually rotate the point <2,0,0> 45 degrees clockwise about the z axis.

Now you can take an object and apply a sequence of transformations to it to make it do whatever you want. All you need to do is figure out the sequence of transformations needed and then apply the sequence to each of the points in the model. As an example, let's say you want to rotate an object sitting at a certain point p around its z axis. You would perform the following sequence of transformations to achieve this:

The first transformation moves a point such that it is situated about the world origin instead of being situated about the point p. The next one rotates it (remember, you can only rotate about the origin, not arbitrary points in space). Finally, after the point is rotated, you want to move it back so that it is situated about p. The final translation accomplishes this. 179 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Rotation w.r.t. Arbitrary Axis: If an object is required to be rotated with respect to a line acting as an axis of rotation, arbitrarily, then the problem is addressed using multiple transformations. Let us assume that such an arbitrary axis is parallel to one of the coordinate axes, say x-axis.

The first step in such case would be to translate the object such that the arbitrary axis coincides with the x-axis.

180 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

The next step would be to rotate the object w.r.t. x-axis through angle .

Then the object is translated such that the arbitrary axis gets back to its original position.

And thus the job is done. An interesting usage of compound transformations:181 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Now, if the arbitrary axis is not parallel to any of the coordinate axes, then the problem is slightly more difficult. It only adds to the number of steps required to get the job done. Let P1, P2 be the line arbitrary axis.

In the first step, the translation takes place that coincides the point P1 to the origin. Points after this step are P1 and P2.

182 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Now the arbitrary axis is rotated such that the point P2 rotates to become P2 and lies on the z-axis.

183 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

In the next step the object of interest is rotated around z-axis.

Now the object of interest is rotated about origin such that the arbitrary axis is poised like in above figure. Point P2 gets back to its previous position P2.

184 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Finally the translation takes place to position the arbitrary axis back to its original position.

Scaling Coordinate transformations for scaling relative to the origin are X` = X . Sx Y` = Y. Sy Z` = Z. Sz Scaling an object with transformation changes the size of the object and reposition the object relative to the coordinate origin. If the transformation parameters are not all equal, relative dimensions in the object are changed. Uniform Scaling : We preserve the original shape of an object with a uniform scaling ( Sx = Sy = Sz) Differential Scaling : We do not preserve the original shape of an object with a differential scaling ( Sx <> Sy <> Sz) Scaling relative to the coordinate Origin:
Scaling transformation of a position P = (x, y, z) relative to the coordinate origin can be written as

S x 0 0 0

0 Sy 0 0

0 0 Sz 0

0 0 0 1

Scaling with respect to a selected fixed position:

185 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

Scaling with respect to a selected fixed position (Xf,Yf,Zf) can be represented with the following transformation sequence: Translate the fixed point to the origin. Scale the object relative to the coordinate origin Translate the fixed point back to its original position For these three transformations we can have composite transformation matrix by multiplying three matrices into one

1 0 0 0

0 0 1 0 0 1 0 0

X f S x Yf . 0 Zf 0 1 0

0 Sy 0 0

0 0 Sz 0

0 1 0 .0 0 0 1 0

0 0 Xf 1 0 Yf 0 1 Zf 0 0 1

S x 0 0 0

0 Sy 0 0

0 0 Sz 0

(1 S x ) X f (1 S y )Y f (1 S z ) Z f 1

Reflection A three-dimensional reflection can be performed relative to a selected reflection axis or with respect to a selected reflection plane. In general, three-dimensional reflection matrices are set up similarly to those for two dimensions. Reflections relative to a given axis are equivalent to 180 degree rotations.
The matrix representation for this reflection of points relative to the X axis

1 0 0 1 0 0 0 0

0 0 0 0 1 0 0 1
186 Copyright Virtual University of Pakistan

18-3D Transformations II

VU

The matrix representation for this reflection of points relative to the Y axis

1 0 0 0

0 0 0 1 0 0 0 1 0 0 0 1

The matrix representation for this reflection of points relative to the xy plane is

1 0 0 0

0 1

0 0 0 1 0 0 0 1 0 0

Shears Shearing transformations can be used to modify object shapes. As an example of three-dimensional shearing, the following transformation produces a z-axis shear:

1 0 0 0

0 a 0 1 b 0 0 1 0 0 0 1

Parameters a and b can be assigned and real values. The effect of this transformation matrix is to alter x and y- coordinate values by an amount that is proportional to the z value, while leaving the z coordinate unchanged. y-axis Shear

1 0 0 0 1 b c 0

a 0 0 1 0 0 c 1 0 0 0 1 0 1 0 0 0 0 1 0 0 0 0 1
187 Copyright Virtual University of Pakistan

x-axis Shear

19-Projections

VU

Lecture No.19

Projections

For centuries, artists, engineers, designers, drafters, and architects have been facing difficulties and constraints imposed by the problem of representing a three-dimensional object or scene in a two-dimensional medium -- the problem of projection. The implementers of a computer graphics system face the same challenge. Projection can be defined as a mapping of point P(x,y,z) onto its image P`(x`,y`,z`) in the projection plane or view plane, which constitutes the display surface. The mapping is determined by a projection line called the projector that passes through P and intersects the view plane.

There are two basic methods of projection Parallel Projection Perspective Projection These methods are used to solve the basic problems of pictorial representations We characterize each method and introduce the mathematical description of the projection process respectively. Taxonomy of Projection We can construct different projections according to the view that is desired. Following figure provides taxonomy of the families of perspective and parallel projections. Some projections have names cavalier, cabinet, isometric, and so on. Other projections qualify the main type of projection one principal vanishingpoint perspective and so forth. 188 Copyright Virtual University of Pakistan

19-Projections

VU

Parallel Projection Parallel projection methods are used by drafters and engineers to create working drawings of an object which preserves its scale and shape. The complete representation of these details often requires two or more views (projections) of the object onto different view planes. In parallel projection, image points are found as the intersection of the view plane with a projector drawn from the object point and having a fixed direction. The direction of projection is the prescribed direction for all projections. Orthographic projections are characterized by the fact that the direction of projection is perpendicular to the view plane. When the direction of projection is parallel to any of the principal axes, this produces the front, top, and side views of mechanical drawings (also referred to as multi view drawings). Axonometric projections are orthographic projections in which the direction of projection is not parallel to any of the three principal axes. Non orthographic parallel projections are called oblique parallel projection. Mathematical Description of a Parallel Projection Projection rays (projectors) emanate from a Center of Projection (COP) and intersect Projection Plane (PP). The COP for parallel projectors is at infinity. The length of a line on the projection plane is the same as the "true Length".

189 Copyright Virtual University of Pakistan

19-Projections

VU

There are two different types of parallel projections: Orthographic Oblique 1) Orthographic Projection If the direction of projection is perpendicular to the projection plane then it is an orthographic projection.

Look at the parallel projection of a point (x, y, z). (Note the left handed coordinate system). The projection plane is at z = 0. x, y are the orthographic projection values and xp, yp are the oblique projection values (at angle a with the projection plane)

190 Copyright Virtual University of Pakistan

19-Projections

VU

Look at orthographic projection: it is simple, just discard the z coordinates. Engineering drawings frequently use front, side, top orthographic views of an object. Axonometric orthographic projection Orthographic projections that show more than one side of an object are called axonometric orthographic projections. Here are three orthographic views of an object.

There are three axonometric projections isometric Dimetric Trimetric 1) Isometric The most common axonometric projection is an isometric projection where the projection plane intersects each coordinate axis in the model coordinate system at an equal distance or the direction of projection makes equal angles with all of the three principal axes 191 Copyright Virtual University of Pakistan

19-Projections

VU

The projection plane intersects the x, y, z axes at equal distances and the projection plane Normal makes an equal angle with the three axes. To form an orthographic projection xp = x, yp= y , zp = 0. To form different types e.g., Isometric, just manipulate object with 3D transformations. 2) Dimetric The direction of projection makes equal angles with exactly two of the principal axes 3) Trimetric The direction of projection makes unequal angles with the three principal axes Oblique Projection If the direction of projection is not perpendicular to the projection plane then it is an oblique projection.

The projectors are not perpendicular to the projection plane but are parallel from the object to the projection plane. Transformation equations for an orthographic parallel projection are straightforward. If the view plane is placed at position Zvp along the Z axis, Then any point (x,y,z) in viewing coordinates is transformed to projection coordinates as: Xp = x Yp = y 192 Copyright Virtual University of Pakistan

19-Projections

VU

Where the original Z-coordinate value is preserved for the depth information needed in depth cueing and visible-surface determination procedures. An oblique projection is obtained by projecting points along parallel lines that are not perpendicular to the projection plane. In some applications packages, an oblique projection vector is specified with two angles, alpha and phi, as shown in the figure. Point (x,y,z) is projected to position(Xp,Yp) on the view plane. Orthographic projection coordinates on the plane are (x,y). The oblique projection line from (x,y,z) to (Xp,Yp) makes an angle alpha with the line on the projection plane that joins (Xp,Yp) and (x, y). This line, of length L, is at an angle phi with the horizontal direction in the projection plane. We can express the projection coordinates in terms of x, y, L, and phi as

cos(phi) = Xp x / L sin(phi) = Yp y / L Xp = x + L cos(phi) Yp = y + L sin(phi) Length L depends on the angle alpha and the z coordinate of the point to be projected: tan (alpha) = z / L Thus, L = z * 1/ tan (alpha) L = z * L1 Where L1 is the inverse of tan(alpha), which is also the value of L when z = 1, we can then write the oblique projection equations. Xp = x + z (L1 cos(phi) ) Yp = y + z (L1 sin(phi) ) 193 Copyright Virtual University of Pakistan

19-Projections

VU

The transformation matrix for producing any parallel projection onto the xy plane can be written as

Now if Alpha = 90 (projection line is perpendicular to Projection Plane) then tan (Alpha) = infinity => L1 = 0, so have an orthographic projection. Two special cases of oblique projection Cavalier Cabinet 1) Cavalier Alpha = 45, tan (Alpha) = 1 => L1 = 1 this is a Cavalier projection such that all lines perpendicular to the projection plane are projected with no change in length.

2) Cabinet tan (Alpha) = 2, Alpha= 63.40, L1 = 1 / 2 Lines which are perpendicular to the projection plane are projected at 1 / 2 length. This is a Cabinet projection

194 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

Lecture No.20

Perspective Projection

Now that you have a structure that can store a three dimensional point (Point3D), how do you calculate the corresponding screen pixel? First, lets look at what you are modeling. Following figure shows how it would look.

This is a mathematical task as pictured. However, we can make it much simpler if we impose the following requirements: the point of View (POV) must lie on the Z axis, and the screen plane must be parallel to the X-Y plane, with the left and right edges of the screen parallel to the Y axis, and the top and bottom edges of the screen parallel to the X axis, for your view to come out correctly, you will also want the Z axis to pass through the middle of the screen. Why? The POV represents the viewers eye, and we presume that the viewer will be behind the center of the screen. Note: We will use Left hand rule to describe 3D coordinate system. Two common approaches are used with this; 1. The first approach is where the POV is at some point (0, 0, -z) and the screen lies on the X-Y plane, graphically, this looks like figure given below:

195 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

2. The second approach is where the POV lies at the origin, and the screen lies on a plane at some +z coordinate, as shown in figure given below:

As we will see later, this second approach is much more convenient when we add features making it possible for the POV to move around the 3D world or for objects to move around in the world. Calculating the screen pixel that correlates to a 3D point is now a matter of simple geometry. From a viewpoint above the screen and POV (looking at the X-Z plane), the geometry appears like the one shown in figure below:

196 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

In geometric terms, we say that the triangle from A to B to S is similar to the triangle from A to C to P because the three angles that make up the triangles are the same: the angle from AB to AS is the same as the angle from AC to AP, the two right angles are both 90 degrees, and therefore the remaining two angles are the same ( the sum of the angles in a triangle is always 180 degrees). What also holds true from similar triangles is that the ratio of two sides holds between the similar triangles; this means that the ratio of BS to AB is the same as the ratio of CP to AC. But we know what AB is-it is Screen.z! and we know what AC is-it is point.z! and we know what CP is-it is point.x! Therefore: |BS| / |AB| = |CP| / |AC| |BS| = |AB| * |CP| / |AC| |BS| = Screen.z * point.x / point.z Screen.z is the distance d from the point of view at origin or the scaling factor. Notice that |BS| is the length of the line segment that goes from B to S in world units. But we normally address the screen with the point (0,0) at the top left, with +X pixels moving to the right, and +Y pixels moving downand not from the middle of the screen. And we draw to the screen in pixel units not our world units (unless, of course, 1.0 in your world represents one pixel). There is a final transformation that the points must go through in the transformation process. This transformation maps 3D points defined with respect to the view origin (in view space) and turns them into 2D points that can be drawn on the display. After transforming and clipping the polygons that make up the scene such that they are visible on the screen, the final step is to move them into 2D coordinates, since in order to actually draw things on the screen you need to have absolute x, y coordinates on the screen to draw. The way this used to be done was without matrices, just as an explicit projection calculation. The point (x,y,z) would be mapped to (x, y) using the following equations

197 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

Where xCenter and yCenter were half of the width and height of the screen respectively, these days more complex equations are used, especially since there is now the need to make provisions for z-buffering. While you want x and y to still behave the same way, you don't want to use a value as arbitrary as scale. Instead, a better value to use in the calculation of the projection matrix is the horizontal field of view (fov). The horizontal fov will be hard coded, and the code chooses a vertical field of view that will keep the aspect ratio of the screen. This makes sense: You couldn't get away with using the same field of view for both horizontal and vertical directions unless the screen was square; it would end up looking vertically squashed. Finally, you also want to scale the z values appropriately. In future, We'll teach you about z-buffering, but for right now just make note of an important feature: They let you clip out certain values of z-range. Given the two variables znear and zfar, nothing in front of znear will be drawn, nor will anything behind zfar. To make the z-buffer work swimmingly on all ranges of znear and zfar, you need to scale the valid z values to the range of 0.0 to 1.0.

The Perspective Projection Matrix

The aspect ratio of screen, width and height is calculated as

With these parameters, the following projection matrix can be made:

Just for a sanity check, check out the result of this matrix multiplication:

198 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

This is almost the result wanted, but there is more work to be done. Remember that in order to extract the Cartesian (x,y,z) coordinates from the vector, the homogenous w component must be 1.0. Since, after the multiplication, it's set to z (which can be any value), all four components need to be divided by z to normalize it (and have the homogeneity factor equal 1). This gives the following Cartesian coordinate:

As you can see, this is exactly what was wanted. The width and height are still scaled by values as in the above equation and they are still divided by z. The visible x and y pixels are mapped to [1,1], so before rasterization Application multiplies and adds the number by xCenter or yCenter. This, in essence, maps the coordinates from [1,1] to [0,width] and [0,height].

With this last piece of the puzzle, it is now possible to create the entire transformation pipeline. When you want to render a scene, you set up a world matrix (to transform an object's local coordinate points into world space), a view matrix (to transform world coordinate points into a space relative to the viewer), and a projection matrix (to take those viewer-relative points and project them onto a 2D surface so that they can be drawn on the screen). You then multiply the world, view, and projection matrices together (in that order) to get a total matrix that transforms points from object space to screen space.

To draw a triangle, for example, you would take its local space points defining its three corners and multiply them by the transformation matrix. Then you have to remember to divide through by the w component. The points are now in screen space and can be filled in using a 2D raster algorithm. Drawing multiple objects is a snap, too. For each object in the scene all you need to do is change the world matrix and reconstruct the total transformation matrix.

The Perspective Projection Matrix Used by Microsoft Direct3D

The projection matrix is typically a scale and perspective projection. The projection transformation converts the viewing frustum into a cuboid shape. Because the near end of the viewing frustum is smaller than the far end, this has the effect of expanding objects that are near to the camera; this is how perspective is applied to the scene.

The Viewing Frustum

A viewing frustum is 3-D volume in a scene positioned relative to the viewport's camera. The shape of the volume affects how models are projected from camera space onto the screen. The most common type of projection, a perspective projection, is responsible for making objects near the camera appear bigger than objects in the distance. For perspective viewing, the viewing frustum can be visualized as a pyramid, with the camera 199 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

positioned at the tip. This pyramid is intersected by a front and back clipping plane. The volume within the pyramid between the front and back clipping planes is the viewing frustum. Objects are visible only when they are in this volume.

If you imagine that you are standing in a dark room and looking through a square window, you are visualizing a viewing frustum. In this analogy, the near clipping plane is the window, and the back clipping plane is whatever finally interrupts your viewthe skyscraper across the street, the mountains in the distance, or nothing at all. You can see everything inside the truncated pyramid that starts at the window and ends with whatever interrupts your view, and you can see nothing else. The viewing frustum is defined by fov (field of view) and by the distances of the front and back clipping planes, specified in z-coordinates.

In this illustration, the variable D is the distance from the camera to the origin of the space that was defined in the last part of the geometry pipelinethe viewing transformation. This is the space around which you arrange the limits of your viewing frustum. For information about how this D variable is used to build the projection matrix The Matrix In the viewing frustum, the distance between the camera and the origin of the viewing transformation space is defined arbitrarily as D, so the projection matrix looks like:

200 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

The viewing matrix translates the camera to the origin by translating in the z direction by - D. The translation matrix is as follows:

Multiplying the translation matrix by the projection matrix (T*P) gives the composite projection matrix. It looks like:

The following illustration shows how the perspective transformation converts a viewing frustum into a new coordinate space. Notice that the frustum becomes cuboid and also that the origin moves from the upper-right corner of the scene to the center.

In the perspective transformation, the limits of the x- and y-directions are -1 and 1. The limits of the z-direction are 0 for the front plane and 1 for the back plane. This matrix translates and scales objects based on a specified distance from the camera to the near clipping plane, but it doesn't consider the field of view (fov), and the z-values that it produces for objects in the distance can be nearly identical, making depth comparisons difficult. The following matrix addresses these issues, and it adjusts vertices to account for the aspect ratio of the viewport, making it a good choice for the perspective projection.

201 Copyright Virtual University of Pakistan

20-Perspective Projections

VU

In this matrix, Zn is the z-value of the near clipping plane. The variables w, h, and Q have the following meanings. Note that fovw and fovh represent the viewport's horizontal and vertical fields of view, in radians.

For your application, using field-of-view angles to define the x- and y-scaling coefficients might not be as convenient as using the viewport's horizontal and vertical dimensions (in camera space). As the math works out, the following two formulas for w and h use the viewport's dimensions, and are equivalent to the preceding formulas.

In these formulas, Zn represents the position of the near clipping plane, and the Vw and Vh variables represent the width and height of the viewport, in camera space.

202 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

Lecture No.21
Triangles

Triangles and Planes

Triangles are to 3D graphics what pixels are to 2D graphics. Every PC hardware accelerator under the sun uses triangles as the fundamental drawing primitive (well scan line aligned trapezoids actually, but that's a hardware implementation issue). When you draw a polygon, hardware devices really draw a fan of triangles. Triangles "flesh out" a 3D object, connecting them together to form a skin or mesh that defines the boundary surface of an object. Triangles, like polygons, generally have an orientation associated with them, to help in normal calculations. The ordering of the vertices goes clockwise around the triangle. Figure below shows what a clockwise ordered triangle would look like.

Figure: Three points in space, and the triangle connecting them When defining a mesh of triangles that define the boundary of a solid, you set it up so that all of the triangles along the skin are ordered clockwise when viewed from the outside. It is impossible to see triangles that face away from you. (You can find this out by computing the triangle's plane normal and performing a dot product with a vector from the camera location to a location on the plane.) Now let's move on to the code. To help facilitate using the multiple types, I'll implement triangles structure. I only define constructors and keep the access public. struct tri { type v[3]; // Array access useful for loops tri() { // nothing } tri( type v0, type v1, type v2 ) 203 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

{ v[0] = v0; v[1] = v1; v[2] = v2; } }; Strips and Fans Lists of triangles are generally represented in one of three ways. The first is an explicit list or array of triangles, where every three elements represent a new triangle. However, there are two additional representations, designed to save bandwidth while sending triangles to dedicated hardware to draw them. They are called triangle strips and triangle fans. Triangle fans, conceptually, look like the folding fans you see in Asian souvenir shops. They are a list of triangles that all share a common point. The first three elements indicate the first triangle. Then each new element is combined with the first element and the current last element to form a new triangle. Note that an N-sided polygon can be represented efficiently using a triangle fan Figure below illustrates what I'm talking about.

Figure: A list of points composing a triangle fan Triangles in a triangle strip, instead of sharing a common element with all other triangles like a fan, only share elements with the triangle immediately preceding them. The first three elements define the first triangle. Then each subsequent element is combined with the two elements before it, in clockwise order, to create a new triangle. See Figure below for an explanation of strips.

204 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

Figure: A list of points composing a triangle strip Planes The next primitive to discuss is the plane. Planes are to 3D what lines are in 2D; they're n1 dimensional hyperplanes that can help you accomplish various tasks. Planes are defined as infinitely large, infinitely thin slices of space, like big pieces of paper. Triangles that make up your model each exist in their own plane. When you have a plane that represents a slice of 3D space, you can perform operations like classification of points and polygons and clipping. So how do you represent planes? Well it is best to build a structure from the equation that defines a plane in 3D. The implicit equation for a plane is: What do these numbers represent? The triplet <a,b,c> represents what is called the normal of the plane. A normal is a unit vector that, conceptually speaking, sticks directly out of a plane. A stronger mathematical definition would be that the normal is a vector that is perpendicular to all of the points that lie in the plane. The d component in the equation represents the distance from the plane to the origin. The distance is computed by tracing a line towards the plane until you hit it. Finally the triplet <x,y,z> is any point that satisfies the equation. The set of all points <x,y,z> that solve the equation is exactly all the points that lie in the plane. All of the pictures I'm showing you will be of the top-down variety, and the 3D planes will be on edge, appearing as 2D lines. This makes figure drawing much easier. Following are two examples of planes. The first has the normal pointing away from the origin, which causes d to be negative (try some sample values for yourself if this doesn't make sense). The second has the normal pointing towards the origin, so d is positive. Of course, if the plane goes through the origin, d is zero (the distance from the plane to the origin is zero). Figures 1 and Figure 2 provide some insight into this relation.

205 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

Figure 1: d is negative when the normal faces away from the origin

Figure 2: d is positive when it faces towards the origin It's important to notice that technically the normal <a,b,c> does not have to be unit-length for it to have a valid plane equation. But since things end up nicer if the normal is unitlength. Constructing a plane given three points that lie in the plane is a simple task. You just perform a cross product between the two vectors made up by the three points <point2 point0> <point1 point0> and find a normal for the plane. After generating the normal and making it unit length, finding the d value for the plane is just a matter of storing the negative dot product of the normal with any of the points. This holds because it essentially solves the plane equation above for d. Of course plugging a point in the plane equation will make it equal 0, and this constructor has three of them. Following has the code to construct a plane from three points. To calculate a plane from 3 given points we first calculate the normal. If we imagine the 3 points form three edges in the plane then we can take two of the edges and calculate the 206 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

cross-product between them. The resulting directional vector will be the normal, and then we can plug any of the 3 known points into the plane equation to solve for k. For points p1,p2 and p3 we get: normal = (p1-p2) x (p3-p2) k = normal * p1

Note that it is extremely important to keep track of which direction your points are stored in. Let's take 3 points stored in clockwise direction in the x/y plane:

The normal to the plane these 3 points define is: normal = (p1-p2) x (p3-p2) = (0,-1,0) x (1,-1,0) = <(-1)*0 - 0*(-1), 0*1 - 0*0, 0*(-1) - (-1)*1> = <0,0,1> ie the z axis. If we were to store the points counter-clockwise the normal calculated would be <0,0,-1>, which is still the z axis but in the "opposite" direction. It's important to keep track of these things since we often need plane equations to be correct in order to determine which side of a polygon an object (such as the view point) is on. Constructing a plane from three points on the plane: Normal vector = n n = cross product ( (b-a),(c-a) ) Normalize(n); find a unit vector d = - dot product(n,a) If you already have a normal and also have a point on the plane, the first step can be skipped. 207 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

Constructing a plane from a normal and a point on the plane: Normalize(n); find a unit vector d = - dot product(n,a) This brings up an important point. If you have an n-sided polygon, nothing discussed up to this point is forcing all of the points to be coplanar. However, problems can crop up if some of the points in the polygon aren't coplanar. For example, when I discuss back-face culling in a moment, you may misidentify what is actually behind the polygon, since there won't be a plane that clearly defines what is in front of and what is behind the plane. That is one of the advantages of using triangles to represent geometrythree points define a plane exactly. Defining Locality with Relation to a Plane One of the most important operations planes let you perform is defining the location of a point with respect to a plane. If you drop a point into the equation, it can be classified into three cases: in front of the plane, in back of the plane, or coplanar with the plane. Front is defined as the side of the plane the normal sticks out of. Here, once again, precision will rear its ugly head. Instead of doing things the theoretical way, having the planes infinitely thin, I'm going to give them a certain thickness of (you guessed it) epsilon. How do you orient a point in relation to a plane? Well, simply plug x, y, and z into the equation, and see what you get on the right side. If you get zero (or a number close enough to zero by plus or minus epsilon), then the point satisfied the equation and lies on the plane. Points like this can be called coplanar. If the number is greater than zero, then you know that you would have to travel farther along the origin following the path of the normal than you would need to go to reach the plane, so the point must be in front of the plane. If the number is negative, it must be behind the plane. Note that the first three terms of the equation simplify to the dot product of the input vector and the plane normal. Figure below has a visual representation of this operation.

Figure: Classifying points with respect to a plane Once you have code to classify a point, classifying other primitives, like polygons, becomes pretty trivial, as shown below. The one issue is there are now four possible definition states when the element being tested isn't infinitesimally small. The element may be entirely in front of the plane, entirely in back, or perfectly coplanar. It may also be 208 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

partially in front and partially in back. I'll refer to this state as splitting the plane. It's just a term; the element isn't actually splitting anything. Back-face Culling Now that you know how to define a point with respect to a plane, you can perform backface culling, one of the most fundamental optimization techniques of 3D graphics. Let's suppose you have a triangle whose elements are ordered in such a fashion that when viewing the triangle from the front, the elements appear in clockwise order. Back-face culling allows you to take triangles defined with this method and use the plane equation to discard triangles that are facing away. Conceptually, any closed mesh, a cube for example, will have some triangles facing you and some facing away. You know for a fact that you'll never be able to see a polygon that faces away from you; they are always hidden by triangles facing towards you. This, of course, doesn't hold if you're allowed to view the cube from its inside, but this shouldn't be allowed to happen if you want to really optimize your engine. Rather than perform the work necessary to draw all of the triangles on the screen, you can use the plane equation to find out if a triangle is facing towards the camera, and discard it if it is not. How is this achieved? Given the three points of the triangle, you can define a plane that the triangle sits in. Since you know the elements of the triangle are listed in clockwise order, you also know that if you pass the elements in order to the plane constructor, the normal to the plane will be on the front side of the triangle. If you then think of the location of the camera as a point, all you need to do is perform a point-plane test. If the point of the camera is in front of the plane, then the triangle is visible and should be drawn. There's an optimization to be had. Since you know three points that lie in the plane (the three points of the triangle) you only need to hold onto the normal of the plane, not the entire plane equation. To perform the back-face cull, just subtract one of the triangle's points from the camera location and perform a dot product with the resultant vector and the normal. If the result of the dot product is greater than zero, then the view point was in front of the triangle. Figure below can help explain the point.

Figure: A visual example of back-face culling In practice, 3D accelerators can actually perform back-face culling by themselves, so as the triangle rates of cards increase, the amount of manual back-face culling that is performed has steadily decreased. However, the information is useful for custom 3D engines that don't plan on using the facilities of direct hardware acceleration. 209 Copyright Virtual University of Pakistan

21-Triangles and Planes

VU

Intersection between a Line and a Plane This occurs at the point which satisfies both the line and the plane equations. Line equation: p = org + u * dir (1) Plane equation: p * normal - k = 0. (2) Substituting (1) into (2) and rearranging we get: (org + u * dir) * normal - k = 0 ie u * dir * normal = k - org * normal ie u = (k - org * normal) / (dir * normal) If (d * normal) = 0 then the line runs parrallel to the plane and no intersection occurs. The exact point at which intersection does occur can be found by plugging u back into the line equation in (1). Clipping Lines One thing that you'll need is the ability to take two points (a and b) that are on different sides of a plane defining a line segment, and find the point making the intersection of the line with the plane. This is easy enough to do. Think of this parametrically. Point a can be thought of as the point at time 0 and point b as the point at time 1, and the point of intersection you want to find is somewhere between those two. Take the dot product of a and b. Using them and the inverse of the plane's d parameter, you can find the scale value (which is a value between 0 and 1 that defines the parametric location of the particle when it intersects the plane). Armed with that, you just use the scale value, plugging it into the linear parametric equation to find the intersection location. Figure 5.17 shows this happening visually, and Listing 5.20 has the code.

Figure 5.17: Finding the intersection of a plane and a line Listing 5.20: plane3::Split inline const point3 plane3::Split( const point3 &a, const point3 &b ) const { float aDot = (a * n); float bDot = (b * n); float scale = ( -d - aDot) / ( bDot - aDot ); return a + (scale * (b - a)); } 210 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

Lecture No.22
Introduction

Triangle Rasterization

High performance triangle rasterization is a very important topic in Computer Graphics in todays world. Triangles are the foundation of modern real time graphics, and are by far the most popular rendering primitive. Most computer games released in the last few years are almost completely dependent on triangle rasterization performance. Recently the focus of graphics performance optimization is beginning to shift to bandwidth requirements as well as transformation and lighting. Nevertheless, rasterization performance is still a factor, and this lecture will provide most of the basics of high performance triangle rasterization. Also, it will go into detail about two often neglected rendering quality improvements, sub-pixel and sub-texel accuracy. Also, smooth shading and texture mapping techniques will be described. Solid Fill Triangle Rendering The first step in triangle rasterization is to be able to render a solid filled triangle. All triangle drawing routines should fill the same pixels on the screen so it makes sense to start with the simplest example and work up. The goal is to draw a filled triangle by plotting pixels on the screen given three vertex points. The first step is to sort the triangle vertices by y. Label the top vertex (x0 ,y0), the middle vertex (x1 ,y1), and the bottom vertex (x2 ,y2). Now the triangle fill can be thought of as two separate routines, filling the top half (the region between y0 and y1) and filling the bottom half (the region between y1 and y2). Each of the fill routines consists of filling the triangle region one scanline at a time, using the DDA algorithm to find the x values of the beginning and the end of each pixel span to draw. The top half will uses DDA to find the x values on edge01 and edge02. The bottom half uses DDA to find the x values on edge12 and edge02.

Sub-pixel Accuracy The afore mentioned rasterization technique is works well when vertex coordinates are integers, but there are some subtle changes that should be made to the DDA algorithm when the vertex coordinates do not fall on integer bounds. In essence, sub-pixel accuracy is a way of accounting for the fractional components of the vertex positions in the triangle rasterizer. The changes that need to be made are mainly used to prevent jumpiness when there are amounts of motion that are smaller than a pixel. The edges of the triangle reflect the fractional change, and without sub-pixel accuracy, the entire triangle would jump down a pixel at a time. Also, the calculations performed for sub-pixel accuracy allow for quicker edge anti-aliasing. The idea of sub-pixel accuracy is to pre-step the x coordinate of each of the edge DDAs an amount corresponding to the fractional component of the y position of the vertex. For

211 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

notation sake we denote the upper vertex of an edge as xa,ya and the lower vertex as xb,yb. For each edge the starting x coordinate for the DDA algorithm xs is obtained by a prestep amount xprestep to the original x coordinate of the vertex:

The pre step amount xf for x, is calculated by multiplying x/y by the fix up distance, yf . This fix up distance is just the distance of ya to the next lowest scan line. For clarification purposes, here is some pseudo code for the sub-pixel accurate DDA, which can be used to find endpoints for the pixel spans in our triangle rasterizer. This technique can be used to draw sub-pixel accurate lines also. SubPixDDA( float xa, float ya, float xb, float yb) { int yai,ybi, /*scanline range to draw*/ y; /*iterator for scanline*/ float xp, /*prestep in x */ x, /*current x position*/ yf, /*fractional distance in y */ dxdy; /*amount to change x by per scanline (1/slope) */ dxdy=(xb-xa)/(yb-ya); /*perform slope calculation using true */ /*vertex positions*/ yai=ceiling(ya); ybi=ceiling(yb); yf=yai-ya; xp=(dxdy*yf); x=xa+xp; for(y=yai;y<ybi;y++) /*iterate over scanlines*/ { /*do scanline stuff using x,y */ x+=dxdy; } } Smooth Shaded Triangle Rasterization So far we have described a method for solid filled triangle rasterization, but there are a variety of other fill types that are use. Smooth shaded triangles can be used to approximate the effects of lighting over a surface. They can be used for light falloff, or can be used to give the appearance to a curved surface. The idea behind smooth shading is to linearly interpolate the vertex colors over the triangle being drawn. Luckily for us, we already have the tool to do this, DDA. In fact, drawing smooth shaded polygons is not much more difficult than drawing solid filled ones. The vertex colors must be interpolated along each edge of the triangle using DDA. This gives us a separate pair of colors for the beginning and end of each pixel span for 212 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

each scan line. The last step that needs to be performed is to use DDA to interpolate the colors across each pixel span. To do smooth shading with RGB color, you must use separate DDA interpolation routines for the red, green, and blue components of the color. Also note that inside the smooth shading routine r, g, and b must be represented as a type with a fractional component (type float is a good choice). To avoid visual artifacts it is recommended that you use the sub-texel accuracy technique that is described further on in these notes. Texture Mapped Triangle Rasterization

Another common triangle fill method is called texture mapping. Texture mapping is a technique for interpolating an image over the triangle being rasterized. The image being interpolated is known as a texture map, and each pixel in the texture map is known as a texel. Because it allows for the use of images to represent a surface on an object, it has the potential to greatly reduce the number of triangles needed to represent an object. In addition to this, texture mapping can also be used to simulate the effects of complex lighting conditions on an object. This section describes bilinear texture mapping, which is the simplest technique to implement. In order to perform bilinear texture mapping, each vertex contains a u, v texture coordinate. This specifies the location in the texture map that this vertex corresponds to. Given these texture coordinates, texture mapping isnt much more difficult than smooth shading. The texture coordinates u, v are interpolated over the triangle using DDA just like the r, g, and b values are in smooth shading. The difference is that the resulting u, v location for every pixel is used to lookup a color value in the texture map image for the pixel to be drawn. However, there is the problem how to deal with the fractional component of the u, v values; looking up color values in the texture map requires integer coordinates. One technique is to round the u, v values to the nearest integer. This is the quickest approach, but produces a blocky looking triangle image when the texture map is small in comparison to the triangle size. Most software based texture mapping routines used in computer games take use this approach because of the speed advantages. However, the majority of hardware based texture mapping routines has an option to do bilinear sampling. Bilinear sampling uses the fractional component of the u, v coordinate to perform a weighted average of 4 adjacent texel colors. The fractional components of u, v are used to find the distance of u, v from the texels themselves. This distance is used as the weighting, and the formula for the pixel color to be drawn cbilin_samp is: Here is some pseudocode to do this: color BilinearSampling(float u, float v, color texMap[256][256]) 213 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

{ color c00,c01,c10,c11; int u0,u1,v0,v1; float ufrac,vfrac; u0=floor(u); u1=ceiling(u); v0=floor(u); v1=ceiling(u); ufrac=u-u0; vfrac=v=v0; c00=texMap[u0][v0]; c01=texMap[u0][v1]; c10=texMap[u1][v0]; c11=texMap[u1][v1]; return( vfrac * (ufrac*c11 + (1-ufrac)*c01) + (1-vfrac) * (ufrac*c10 + (1-ufrac)*c00) ); }

Sub-Texel Accuracy The disparity between integer screen pixel locations and the mathematical equations for the triangle also causes problems for texture mapping, and smooth shading. Any value that is interpolated over the triangle such as r, g, and b for smooth shading and u and v for texture mapping must take into account the fractional component of the vertex information. Taking these fractional quantities into account is called sub-texel accuracy, for the reason that it is used most commonly with texture mapping. In actuality, sub-texel accuracy can be applied to any quantity interpolated over the triangle. Without sub-texel accuracy, the texture will visibly jump around by a pixel when the triangle undergoes small amounts of motion. The sub-texel accurate DDA interpolators for texture mapping are very similar to the subpixel accurate DDA routine presented earlier. For each edge of the triangle, the sub-texel DDA for the interpolated values is identical to the sub-pixel DDA, when u or v is substituted for x. However, for each scan line, the beginning and end x locations of the pixel span have fractional components which need to be accounted for. To interpolate the texels coordinates correctly over the pixel span for each scan line, a subtexel accurate pixel span DDA is required. Luckily for us, this formulation is also virtually identical to the sub-pixel DDA. All that need to be done is to substitute u or v for x in the original, and substitute x for y in the original.

214 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

Flat Filling Triangles

Drawing triangle (or in general convex polygon, but as we discussed we will use only triangles) is very simple. The basic idea of the line triangle drawing algorithm is as follows. For each scan line (horizontal line on the screen), find the points of intersection with the edges of the triangle. Then, draw a horizontal line between intersections and do this for all scan lines. But how can we find these points quickly? Using linear interpolation! We have 3 vertices and we want to find coordinates of all points belonging to segments determined by these vertices. Assume we have segment given by points: (xa,ya) and (xb,yb). Our task is to find points: (xc,ya+1), (xd,ya+2), ... , (xm,yb-1), (xn,yb). Notice that xa changes to xb in (yb-ya) steps. We also have: xa=xa+0*(xb-xa)/(yb-ya), Xb = xa + ( yb-ya ) * ( xb-xa ) / ( yb-ya ) and, in general, xi = xa + ( yi ya ) * delta , where delta=(xb-xa)/(yb-ya). The general function for linear interpolation is: f(X) = A + X * ( (B-A) / steps) where we slide from A to B in steps steps Here is pseudo code for a triangle filling algorithm. The coordinates of vertices are (A. x,A. y), (B. x,B. y), (C. x,C. y); we assume that A. y <= B. y <= C. y (you should sort them first) dx1,dx2,dx3 are deltas used in interpolation Horizontal line draws horizontal segment with coordinates (S. x, Y), (E. x, Y) S. x, E. x are left and right x-coordinates of the segment we have to draw S = A means that S. x = A. x; S. y = A. y; 215 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

if (B.y - A.y > 0) dx1 = (B.x-A.x) / (B.y-A.y) else dx1=0; if (C.y - A.y > 0) dx2=(C.x-A.x)/(C.y-A.y) else dx2=0; if (C.y-B.y > 0) dx3=(C.x-B.x)/(C.y-B.y) else dx3=0; S=E=A if(dx1 > dx2) { for( ; S.y<=B.y; S.y++, E.y++, S.x+=dx2, E.x += dx1) horizontal line (S.x,E.x,S.y, E.x,color); E=B; for(;S.y<=C.y;S.y++,E.y++,S.x+=dx2,E.x+=dx3) horizontal line(S.x,E.x,S.y, E.x,color); } else { for(;S.y<=B.y;S.y++,E.y++,S.x+=dx1,E.x+=dx2) horizontal line(S.x,E.x,S.y, E.x,color); S=B; for(;S.y<=C.y;S.y++,E.y++,S.x+=dx3,E.x+=dx2) horizontal line(S.x,E.x,S.y, E.x,color); } I ought to explain what is the comparision dx1 > dx2 for. It's optimization trick: in the horizontal line routine, we don't need to compare the x's (S.x is always less than or equal to E.x).

216 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

Gouraud Shading

The idea of gouraud and flat triangle is nearly the same. Gouraud takes only three parameters more (the color value of each of the vertices), and the routine just interpolates among them drawing a beautiful, shaded triangle. You can use 256-colors mode, in which vertices' colors are simply indices to palette or hicolor mode (recommended). Flat triangle interpolated only one value (x in connection with y), 256 colors gouraud needs three (x related to y, color related to y, and color related to x), hi-color gouraud needs seven (x related to y, red, green and blue components of color related to y, and color related to x (also three components)) Drawing a gouraud triangle, we add only two parts to the flat triangle routine. The horizline routine gets a bit more complicated due to the interpolation of the color value related to x but the main routine itself remains nearly the same. We'll give you a full gouraud routine because good pseudo code is better than the best description: the coordinates of vertices are (A.x,A.y), (B.x,B.y), (C.x,C.y) we assume that A.y<=B.y<=C.y (you should sort them first) vertex A has color (A.r,A.g,A.b), B (B.r,B.g,B.b), C (C.r,C.g,C.b), where X.r is color's red component, X.g is color's green component and X.b is color's blue component dx1,dx2,dx3 are deltas used in interpolation of x-coordinate dr1,dr2,dr3, dg1,dg2,dg3, db1,db2,db3 are deltas used in interpolation of color's components putpixel(P) plots a pixel with coordinates (P.x,P.y) and color (P.r,P.g,P.b) S=A means that S.x=A.x; S.y=A.y; S.r=A.r; S.g=A.g; S.b=A.b; Drawing triangle: if (B.y-A.y > 0) { dx1=(B.x-A.x)/(B.y-A.y); dr1=(B.r-A.r)/(B.y-A.y); dg1=(B.g-A.g)/(B.y-A.y); db1=(B.b-A.b)/(B.y-A.y); } else dx1=dr1=dg1=db1=0; if (C.y-A.y > 0) { dx2=(C.x-A.x)/(C.y-A.y); dr2=(C.r-A.r)/(C.y-A.y); dg2=(C.g-A.g)/(C.y-A.y); 217 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

db2=(C.b-A.b)/(C.y-A.y); } else dx2=dr2=dg2=db2=0; if (C.y-B.y > 0) { dx3=(C.x-B.x)/(C.y-B.y); dr3=(C.r-B.r)/(C.y-B.y); dg3=(C.g-B.g)/(C.y-B.y); db3=(C.b-B.b)/(C.y-B.y); } else dx3=dr3=dg3=db3=0; S=E=A; if(dx1 > dx2) { for(;S.y<=B.y;S.y++,E.y++) { if(E.x-S.x > 0) { dr=(E.r-S.r)/(E.x-S.x); dg=(E.g-S.g)/(E.x-S.x); db=(E.b-S.b)/(E.x-S.x); } else dr=dg=db=0; P=S; for(;P.x < E.x;P.x++) { putpixel(P); P.r+=dr; P.g+=dg; P.b+=db; } S.x+=dx2; S.r+=dr2; S.g+=dg2; S.b+=db2; E.x+=dx1; E.r+=dr1; E.g+=dg1; E.b+=db1; } E=B; for(;S.y<=C.y;S.y++,E.y++) { if(E.x-S.x > 0) { dr=(E.r-S.r)/(E.x-S.x); dg=(E.g-S.g)/(E.x-S.x); db=(E.b-S.b)/(E.x-S.x); } else dr=dg=db=0; P=S; for(;P.x < E.x;P.x++) { putpixel(P); P.r+=dr; P.g+=dg; P.b+=db; } S.x+=dx2; S.r+=dr2; S.g+=dg2; S.b+=db2; E.x+=dx3; E.r+=dr3; E.g+=dg3; E.b+=db3; } } else { for(;S.y<=B.y;S.y++,E.y++) { if(E.x-S.x > 0) { dr=(E.r-S.r)/(E.x-S.x); 218 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

dg=(E.g-S.g)/(E.x-S.x); db=(E.b-S.b)/(E.x-S.x); } else dr=dg=db=0; P=S; for(;P.x < E.x;P.x++) { putpixel(P); P.r+=dr; P.g+=dg; P.b+=db; } S.x+=dx1; S.r+=dr1; S.g+=dg1; S.b+=db1; E.x+=dx2; E.r+=dr2; E.g+=dg2; E.b+=db2; } S=B; for(;S.y<=C.y;S.y++,E.y++) { if(E.x-S.x > 0) { dr=(E.r-S.r)/(E.x-S.x); dg=(E.g-S.g)/(E.x-S.x); db=(E.b-S.b)/(E.x-S.x); } else dr=dg=db=0; P=S; for(;P.x < E.x;P.x++) { putpixel(P); P.r+=dr; P.g+=dg; P.b+=db; } S.x+=dx3; S.r+=dr3; S.g+=dg3; S.b+=db3; E.x+=dx2; E.r+=dr2; E.g+=dg2; E.b+=db2; } } I hope you are familiar with idea of interpolation now. Textured Triangles

We can also apply any bitmap on triangle for filling it. I'll show you the idea of linear (or 'classical') texture mapping (without perspective correction). Linear mapping works pretty well (read: fast) in some scenes, but perspective correction is in some way needed in most 3D systems. Again we're using the idea of interpolation: now we'll code a texture triangle filler. And again the idea is perfectly the same, only two more values to interpolate, that is five values total. In texture mapping, we interpolate x, u, and v related to y, and u and v related to x (u and v are coordinates in the 2D bitmap space). The situation is maybe easier to understand by looking at the following picture: 219 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

The left triangle is the triangle which is drawn onto the screen. There's a single scanline (one call to the horizline routine) pointed out as an example. The triangle on the right is the same triangle in the bitmap space, and there's the same scanline drawn from another point of view into it, too. So we need just to interpolate, interpolate, and once more interpolate in texture filler - an easy job if you've understood the idea of gouraud filler. An optimization trick: the color deltas in gouraud and (u,v) coordinate deltas in texture remain constant, so we need to calculate them only once per polygon. Let's take the u delta in linear texturing as an example. Assume, that dx2<=dx3 (we are using the same symbols like in flat and gouraud filler). As we know, we need to interpolate S.u to E.u in the horizline routine in (S.x-E.x) steps. We are in the need of a u delta (du) which would be the same for the whole polygon. So instead of calculating in each scanline this: du = (E.u-S.u) / (E.x-S.x), we do like this in the setup part of the polygon routine: We know that S.x = Ax + (B.y - A.y) * dx1, S.u = A.u + (B.y-A.y) * du1, E.x = B.x = A.x + (B.y-A.y) * dx2, E.u = B.u = A.u + (B.y-A.y) * du2, When y = B.y (when y is the y-coordinate of the second vertex). When we place the values of the variables S.u,E.u,S.x and E.x (above) to the u delta statement, du = (E.u-S.u) / (E.x-S.x), we get the following statement as a result: [A.u+(B.y-A.y)*du2] - [A.u+(B.y-A.y)*du1] du = ----------------------------------------[A.x+(B.y-A.y)*dx2] - [A.x+(B.y-A.y)*dx1] (B.y-A.y)*(A.u-A.u+du2-du1) du = --------------------------(B.y-A.y)*(A.x-A.x+dx2-dx1) du2-du1 du = ------dx2-dx1

220 Copyright Virtual University of Pakistan

22-Triangle Rasterization

VU

outerUdelta2-outerUdelta1 innerUdelta = -------------------------------outerXdelta2-outerXdelta1 Nice! But what if dx2 = dx1? This of course means that the polygon is just one line, so du doesn't need any specific value; zero does the job very well. Note! I find it hard to get good results using fixed point math because of inadequate precision. Environmental Mapping

As I said in 'shading' part, the way demos do environment mapping is very simple. Take the X and Y components of your pseudo-normal vectors (perpendicular to vertices), and use them to index your texture map! Your formulae would be: U = N.x * 128 + 127 V = N.y * 128 + 127 (assuming 256x256 texture maps). Or in general U = N.x * (width / 2) + (width / 2) - 1 V = N.y * (height / 2) + (height / 2) - 1

Using texturing and shading at the same time is quite straightforward to implement: the basic idea being that we just interpolate the values of both texture and shade and blend them in a suitable ratio (alpha-blending).

221 Copyright Virtual University of Pakistan

23-Lighting I

VU

Lecture No.23

Lighting I

MATHEMATICS OF COLOR IN COMPUTER GRAPHICS It is important to understand how color is represented in computer graphics so that we can manipulate it effectively. A color is usually represented in the graphics pipeline by a three-element vector representing the intensities of the red, green, and blue components, or for a more complex object, by a four-element vector containing an additional value called the alpha component that represents the opacity of the color. Thus we can talk about rgb or rgba colors and mean a color that's made up of either three or four elements. There are many different ways of representing the intensity of a particular color element. Colors can also be represented as floating point values in the range [0,1]. Nowadays every PC we can buy has hardware that can render images with thousands or millions of individual colors. Rather than have an array with thousands of color entries, the images instead contain explicit color values for each pixel. A 16-bit display is named since each pixel in a 16-bit image is taken up by 16 bits (2 bytes): 5 bits of red information, 6 bits of green information, and 5 bits of blue information. Incidentally, the extra bit (and therefore twice as much color resolution) is given to green because our eyes are more sensitive to green. A 24-bit display, of course, uses 24 bits, or 3 bytes per pixel, for color information. This gives 1 byte, or 256 distinct values each, for red, green, and blue. This is generally called true color, because 2563 (16.7 million) colors is about as much as your eyes can discern, so more color resolution really isn't necessary, at least for computer monitors. Finally, there is 32-bit color, something seen on most new graphics cards. Many 3D accelerators keep 8 extra bits per pixel around to store transparency information, which is generally referred to as the alpha channel, and therefore take up 4 bytes, or 32 bits, of storage per pixel. Rather than re-implementing the display logic on 2D displays that don't need alpha information, these 8 bits are usually just wasted. Representing Color Before we can go about giving color to anything in a scene, we need to know how to represent color! Usually we use the same red, green, and blue (rgb) channels discussed above, but for this there will also be a fourth component called alpha. The alpha component stores transparency information about a surface. Practically we will use two structures to ease the color duties: color3 and color4. They both use floating-point values for their components; color3 has red, green, and blue, while color4 has the additional fourth component of alpha in there. Colors aren't like pointsthey have a fixed range. Each component can be anywhere between 0.0 and 1.0 (zero contribution of the channel or complete contribution). If performing operations on colors, such as adding them together, the components may rise above 1.0 or below 0.0.

222 Copyright Virtual University of Pakistan

23-Lighting I

VU

The code for color4 appears below. The color4 structure struct color4 { union { struct { float r, g, b, a; // Red, Green, and Blue color data }; float c[4]; }; color4(){} color4( float inR, float inG, float inB, float inA ) : r( inR ), g( inG ), b( inB ), a( inA ) { } color4( const color3& in, float alpha = 1.f ) { r = in.r; g = in.g; b = in.b; a = alpha; } color4( unsigned long color ) { b = (float)(color&255) / 255.f; color >>= 8; g = (float)(color&255) / 255.f; color >>= 8; r = (float)(color&255) / 255.f; color >>= 8; a = (float)(color&255) / 255.f; } unsigned long MakeDWord() { unsigned long iA = (int)(a * 255.f ) << 24; unsigned long iR = (int)(r * 255.f ) << 16; unsigned long iG = (int)(g * 255.f ) << 8; unsigned long iB = (int)(b * 255.f ); return iA | iR | iG | iB; } // if any of the values are >1, cap them. 223 Copyright Virtual University of Pakistan

23-Lighting I

VU

void Saturation() { if( r > 1 ) r = 1.f; if( g > 1 ) g = 1.f; if( b > 1 ) b = 1.f; if( a > 1 ) a = 1.f; if( r < 0.f ) r = 0.f; if( g < 0.f ) g = 0.f; if( b < 0.f ) b = 0.f; if( a < 0.f ) a = 0.f; } }; We should also point out that when dealing with colors, particularly with some of the subtleties that we'll be getting into with lights and shades, we should understand the gamut of the target device. This is where our beautiful clean mathematics meets the real world. The gamut of a device is simply the physical range of colors the device can display. Typically, a high-quality display has a better gamut than a cheap one. A good printer has a gamut that is significantly different from that of a monitor. If we are interested in getting some color images for printing, we shall have to do some manipulation on the color values to make the printed image look like the one our program generated on the screen. We should also be aware that there are color spaces other than the RGB color space. HSV (hue, saturation, and value) is one that's typically used by printers, for example. WHY WE MIGHT WANT 128-BIT COLOR In one of the early magazines articles of Mike Abrash [ABRASH 1992], he tells a story about going from a 256-color palette to hardware that supported 256 levels for each RGB color16 million colors! What would we do with all those colors? He goes on to tell of a story by Sheldon Linker at the eighth Annual Computer Graphics Show on how the folks at the Jet Propulsion Lab back in the 1970s had a printer that could print over 50 million distinct colors. As a test, they printed out words on paper where the background color was only one color index from the word's color. To their surprise, it was easy to discern the wordsthe human eye is very sensitive to color graduations and edge detection. The JPL team then did the same tests on color monitors and discovered that only about 16 million colors could be distinguished. It seems that the eye is (not too surprisingly) better at perceiving detail from reflected light (such as from a printed page) than from emissive light (such as from a CRT). The moral is that the eye is a lot more perceptive than you 224 Copyright Virtual University of Pakistan

23-Lighting I

VU

might think. Twenty four-bits of color really is not that much range, particularly if we are performing multiple passes. Round-off error can and will show up if we aren't careful! An example of the various gamuts is shown in the figure below. The CIE diagrams are the traditional way of displaying perceived color space, which, we should note, is very different from the linear color space used by today's graphics hardware. The colored area is the gamut of the human eye. The gamut of printers and monitors are subsets of this gamut.

Figure 1: The 1931 CIE diagram shows the gamut of the eye and the lesser gamut of output devices.

Multiplying Color Values First we need to be aware of how to treat colors. The calculation of the color of a particular pixel depends, for example, on the surface's material properties that we've programmed in, the color of the ambient light (lighting model), the color of any light shining on the surface (perhaps of the angle of the light to the surface), the angle of the surface to the viewpoint, the color of any fog or other scattering material that's between the surface and the viewpoint, etc. No matter how you are calculating the color of the pixel, it all comes down to color calculations, at least on current hardware, on rgb or rgba vectors where the individual color elements are limited to the [0,1] range. Operations on colors are done piecewisethat is, even though we represent colors as rgb vectors, they aren't really vectors in the mathematical sense. Vector multiplication is different from the operation we perform to multiply color. We'll use the symbol to indicate such piecewise multiplication. Colors are multiplied to describe the interaction between a surface and a light source. The colors of each are multiplied together to estimate the reflected light colorthis is the color of the light that this particular light reflects off this surface. The problem with the 225 Copyright Virtual University of Pakistan

23-Lighting I

VU

standard rgb model is just that we're simulating the entire visible spectrum by three colors with a limited range. Let's start with a simple example of using reflected colors. Later on we will discuss on lighting, we'll discover how to calculate the intensity of a light source, but for now, just assume that we've calculated the intensity of a light, and it's a value called id. This intensity of our light is represented by, say, a nice lime green color. Thus

Let's say we shine this light on a nice magenta surface given by cs.

So, to calculate the color contribution of this surface from this particular light, we perform a piecewise multiplication of the color values.

Note: Piecewise multiplication is denoted by that is element-by-element multiplication, used in color operations, where the vector just represents a convenient notation for an array of scalars that are operated on simultaneously but independently. This gives us the dark plum color shown in figure below. We should note that since the surface has no green component, that no matter what value we used for the light color, there would never be any green component from the resulting calculation. Thus a pure green light would provide no contribution to the intensity of a surface if that surface contained a zero value for its green intensity. Thus it's possible to illuminate a surface with a bright light and get little or no illumination from that light. We should also note that using anything other than a full-bright white light [1,1,1] will involve multiplication of values less than one, which means that using a single light source will only illuminate a surface to a maximum intensity of its color value, never more. This same problem also happens when a texture is modulated by a surface color. The color of the surface will be multiplied by the colors in the texture. If the surface color is anything other than full white, the texture will become darker. Multiple texture passes can make a surface very dark very quickly.

226 Copyright Virtual University of Pakistan

23-Lighting I

VU

Figure 2: Multiplying (modulating) color values results in a color equal to or less than (darker) the original two. Given that using a colored light in a scene makes the scene darker, how do we make the scene brighter? There are a few ways of doing this. Given that color multiplication will never result in a brighter color, it's offset a bit since we end up summing all the light contributions together, which, as we'll see in the next section, brings with it its own problems. But if we are just interested in increasing the brightness on one particular light or texture, one way is to use the API (Library routines e.g. OpenGL or DirectX) to artificially brighten the sourcethis is typically done with texture preprocessing. Or, we can artificially brighten the source, be it a light or a texture, by adjusting the values after we modulate them. Dealing with Saturated Colors On the other hand, what if we have too much contribution to a color? While the colors of lights are modulated by the color of the surface, each light source that illuminates the surface is added to the final color. All these colors are summed up to calculate the final color. Let's look at such a problem. We'll start with summing the reflected colors off a surface from two lights. The first light is an orange color and has rgb values [1.0,0.49,0.0], and the second light is a nice light green with rgb values [0.0,1.0,0.49]. Summing these two colors yields [1.0, 1.49, 0.49], which we can't display because of the values larger than one figure below shows.

Figure 3: Adding colors can result in colors that are outside the displayable range. So, what can be done when color values exceed the range that the hardware can display? It turns out that there are three common approaches [HALL 1990]. Clamping the color values is implemented in hardware, so for shaders (technology used in today computer graphics for lighting and shading), it's the default, and it just means that we clamp any values outside the [0,1] range. Unfortunately, this results in a shift in the color. The second most common approach is to scale the colors by the largest component. This maintains the color but reduces the overall intensity of the color. The third is to try to maintain the intensity of the color by shifting (or clipping) the color toward pure bright white by reducing the colors that are too bright while increasing the other colors and maintaining the overall intensity. Since we can't see what the actual color for (figure above) is, let's see what color each of these methods yields (figure below).

227 Copyright Virtual University of Pakistan

23-Lighting I

VU

Figure 4: The results of three strategies for dealing with the same oversaturated color. As we can see, we get three very different results. In terms of perceived color, the scaled is probably the closest though it's darker than the actual color values. If we weren't interested in the color but more in terms of saturation, then the clipped color is closer. Finally, the clamped value is what we get by default, and as you can see, the green component is biased down so that we lose a good sense of the "greenness" of the color we were trying to create.

228 Copyright Virtual University of Pakistan

24-Lighting II

VU

Lecture No.24
Clamping Color Values

Lighting II

Now it's perfectly fine to end up with an oversaturated color and pass this result along to the graphics engine. What happens in the pipeline is an implicit clamping of the color values. Any value that's greater than one is clamped to one, and any less than zero are clamped to zero. So this has the benefit of requiring no effort on the part of the shader (technology that is being used today for lighting and shading supported by Graphics hardware) writer. Though this may make the rendering engine happy, it probably isn't what we want. Intuitively, we'd think that shining orange and green lights on a white surface would yield a strong green result. But letting the hardware clamp eradicates any predominant effect from the green light. Clamping is fast, but it tends to lose fidelity in the scene, particularly in areas where we would want and expect subtle changes as the light intensities interact, but end up with those interactions getting eradicated because the differences are all getting clamped out by the graphics hardware. Scaling Color Values by Intensity Instead of clamping, we might want to scale the color by dividing by the largest color value, thus scaling the rgb values into the [0,1] range. In the example from figure 1, the final color values were [1.0,1.49,0.49] meaning our largest color value was the green, at 1.49. Using this approach, we divide each element by 1.49, yielding a scaled color of [0.671,1.0,0.329]. Thus any values greater than one are scaled to one, while any other values are also scaled by the same amount. This maintains the hue and saturation but loses the intensity. This might not be acceptable because the contrast with other colors is lost, since contrast perception is nonlinear and we're applying a linear scaling. By looking at the three results, we can see the large difference between the resulting colors.

Figure 1: Adding colors can result in colors that are outside the displayable range. Shifting Color Values to Maintain Saturation One problem with clamping or scaling colors is that they get darker (lose saturation). An alternative to scaling is to maintain saturation by shifting color values. This technique is called clipping, and it's a bit more complicated than color scaling or clamping. The idea is to create a gray-scale vector that runs along the black-white axis of the color cube that's got the same brightness as the original color and then to draw a ray at right angles to this vector that intersects (i.e., clips) the original color's vector. We need to check to make sure that the grayscale vector is itself within the [0,1] range and then to check the sign of the ray elements to see if the color elements need to be increased or decreased. As we are probably wondering, this can result in adding in a color value that wasn't in the original color, but this is a direct result of wanting to make sure that the overall brightness is the 229 Copyright Virtual University of Pakistan

24-Lighting II

VU

same as the original color. And, of course, everything goes to hell in a handbasket if we've got overly bright colors, which leave we with decisions about how to nudge the gray-scale vector into the [0,1] range, since that means you can't achieve the input color's saturation value. Then we're back to clamping or scaling again. ColorSpace Tool The ColorSpace tool is a handy tool that we can use to interactively add two colors together to see the effects of the various strategies for handling oversaturated colors. We simply use the sliders to select the rgb color values for each color. The four displays in Figure 5 show the composite, unmodified values of the resulting color (with no color square) and the clamped, clipped, and scaled color rgb values along with a color square illustrating those color values.

Figure 2: The ColorSpace tool interface. Negative Colors and Darklights We may be wondering, if we can have color values greater than the range in intermediate calculations, can we have negative values? Yes, we can! They are called "darklights" after their description in an article [GLASSNER 1992] in Graphic Gems III. Since this is all just math until we pass it back to the graphics hardware, we can pretty much do anything we want, which is pretty much the idea behind programmable shaders (technology used by today graphics hardware for lighting and shading)! Darklights are nothing more than lights in which one or more of the color values are negative. Thus instead of contributing to the overall lighting in a scene, we can specify a light that diminishes the overall lighting in a scene. Darklights are used to eliminate bright areas when we're happy with all the lighting in your scene except for an overly bright area. Darklights can also be used to affect a scene if we want to filter out a specific rgb color. If we wanted to get a night vision effect, we could use a darklight with negative red and blue values, for example, which would just leave the green channel. 230 Copyright Virtual University of Pakistan

24-Lighting II

VU

Alpha Blending Up to this point, weve been fairly dismissive of the mysterious alpha component that rides along in all of the color4 structure. Now, we may finally learn its dark secrets. A lot of power is hidden away inside the alpha component. Loosely, the alpha component of the RGBA quad represents the opaqueness of a surface. An alpha value of 0xFF (255) means the color is completely opaque, and an alpha value of 0x00 (0) means the color is completely transparent. Of course, the value of the alpha component is fairly meaningless unless we actually activate the alpha blending step. If we want, we can set things up a different way, such as having 0x00(0) mean that the color is completely opaque. The meaning of alpha is dependent on how we set up the alpha blending step. As you rasterize primitives, each pixel that we wish to change in the frame buffer gets sent through the alpha blending step. That pixel is combined using blending factors to the pixel that is currently in the frame buffer. We can add the two pixels together, multiply them together, linearly combine them using the alpha component, and so forth. The name "alpha blending" comes from the fact that generally the blending factors used are either the alpha or the inverse of the alpha. The Alpha Blending Equation The equation that governs the behavior of the blending is defined as follows:

Final color is the color that goes to the frame buffer after the blending operation. Source is the pixel we are attempting to draw to the frame buffer, generally one of the many pixels in a triangle we have to draw. Destination is the pixel that already exists in the frame buffer before we attempt to draw a new one. The source and destination blend factors are variables that modify how the colors are combined together. The blend factors are the components we have control over in the equation; we cannot modify the positions of any of the terms or modify the operations performed on them. For example, say we want an alpha blending equation to do nothingto just draw the pixel from the triangle and not consider what was already there at all. An equation that would accomplish this would be:

As we can see, the destination-blending factor is 0 and the source-blending factor is 1. This reduces the equation to:

A second example would be if we wanted to multiply the source and destination components together before writing them to the frame buffer. This initially would seem difficult, as in the above equation they are only added together. However, the blending factors defined need not be constants; they can in fact be actual color components (or inverses thereof). The equation setup would be:

231 Copyright Virtual University of Pakistan

24-Lighting II

VU

In this equation, the destination blend factor is set to the source color itself. Also, since the source blend factor is set to zero, the left-hand side of the equation drops away and we are left with:

Code Example //This code will blend one image into the second struct COLOR3{ BYTE b; BYTE g; BYTE r; }; //Assume the two bitmaps have the same size COLOR3 *p1; //pointer to first bitmap; COLOR3 *p2; //pointer to Second bitmap; int k=0; now we compute the blending factor av. float av=(float)(alpha&255)/255;// alphaValue in floating point

compute the new pixel by using the alpha blending formula. for(int i=firstBMP.Height; i>0; i--) { for(int j=0; j<firstBMP.Width; j++) { get the red color p1->r =(BYTE)(( p1->r * (av))+ p2->r*(1.0f-(av)) ); get the green color p1->g =(BYTE)(( p1->g * (av))+ p2->g*(1.0f-(av)) ); get the blue color p1->b =(BYTE)(( p1->b * (av))+ p2->b*(1.0f-(av)) ); p1++; p2++; } } send the new bitmap to display device BlitData(displaydeviceContext, 0,0,firstImage.Width,firstImage.Height); Following images shows the result of the above code. 232 Copyright Virtual University of Pakistan

24-Lighting II

VU

Figure 3: First Image

Figure 4: Second Image

Figure 5: Blended image

233 Copyright Virtual University of Pakistan

25-Mathiematics of Lighting and Shading Part I

VU

Lecture No.25

Mathematics of Lighting and Shading Part I

LIGHTS AND MATERIALS In order to understand how an object's color is determined, we'll need to understand the parts that come into play to create the final color. First, we need a source of illumination, typically in the form of a light source in our scene. A light has the properties of color (an rgb value) and intensity. Typically, these are multiplied to give scaled rgb values. Lights can also have attenuation, which means that their intensity is a function of the distance from the light to the surface. Lights can additionally be given other properties such as a shape (e.g., spotlights) and position (local or directional), but that's more in the implementation rather than the math of lighting effects. Given a source of illumination, we'll need a surface on which the light will shine. Here's where we get interesting effects. Two types of phenomena are important lighting calculations. The first is the interaction of light with the surface boundary, and the second is the effect of light as it gets absorbed, transmitted, and scattered by interacting with the actual material itself. Since we really only have tools for describing surfaces of objects and not the internal material properties, lightsurface boundary interactions are the most common type of calculation we'll see used, though we can do some interesting simulations of the interaction of light with material internals. Materials are typically richer in their descriptions in an effort to mimic the effects seen in real lightmaterial surface interactions. Materials are typically described using two to four separate colors in an effort to catch the nuances of real-world lightmaterial surface interactions. These colors are the ambient, diffuse, specular, and emissive colors, with ambient and specular frequently grouped together, and emissive specified only for objects that generate light themselves. The reason there are different colors is to give different effects arising from different environmental causes. The most common lights are as follows: Ambient lighting: It is the overall color of the object due to the global ambient light level. This is the color of the object when there's no particular light, just the general environmental illumination. That is, the ambient light is an approximation for the global illumination in the environment, and relies upon no light in the scene. It's usually a global value that's added to every object in a scene.

Diffuse lighting: It is the color of the object due to the effect of a particular light. The diffuse light is the light of the surface if the surface were perfectly matte. The diffuse light is reflected in all directions from the surface and depends only on the angle of the light to the surface normal. Specular lighting: It is the color of the highlights on the surface. The specular light mimics the shininess of a surface, and its intensity is a function of the light's reflection angle off the surface. 234 Copyright Virtual University of Pakistan

25-Mathiematics of Lighting and Shading Part I

VU

Emissive lighting: When we need an object to "glow" in a scene, we can do this with an emissive light. This is just an additional color source added to the final light of the object. Because we're simulating an object giving off its own light; we'd still have to add a real "light" to get an effect on objects in a scene. Before we get into exactly what these types of lighting are, let's put it in perspective for our purpose of writing shader code. Shading is simply calculating the color reflected off a surface (which is pretty much what shaders do). When a light reflects off a surface, the light colors are modulated by the surface color (typically, the diffuse or ambient surface color). Modulation means multiplication, and for colors, since we are using rgb values, this means component-by-component multiplication. So for light source l with color (r1,g1,b1 shining on surface s with color (rs,gs,bs, the resulting color r would be:

or, multiplying it out, we get

Where the resulting rgb values of the light and surface are multiplied out to get the final color's rgb values The final step after calculating all the lighting contributions is to add together all the lights to get the final color. So a shader might typically do the following: 1. Calculate the overall ambient light on a surface. 2. For each light in a scene, calculate the diffuse and specular contribution for each light. 3. Calculate any emissive light for a surface. 4. Add all these lights together to calculate the final color value. In the real world, we get some sort of interaction (reflection, etc.) when a photon interacts with a surface boundary. Thus we see the effects not only when we have a transparent opaque boundary (like airplastic), but also a transparenttransparent boundary (like airwater). The key feature here is that we get some visual effect when a photon interacts with some boundary between two different materials. The conductivity of the materials directly affects how the photon is reflected. At the surface of a conductor (metals, etc.), the light is mostly reflected. For dielectrics (nonconductors), there is usually more penetration and transmittance of the light. For both kinds of materials, the dispersion of the light is a function of the roughness of the surface (Figure 1 and 2).

235 Copyright Virtual University of Pakistan

25-Mathiematics of Lighting and Shading Part I

VU

Figure 1: Light reflecting from a rough and smooth surface of a conductor.

Figure 2: Light reflecting from a rough and smooth surface of a dielectric showing some penetration. The simplest model assumes that the roughness of the surface is so fine that light is dispersed equally in all directions as shown in Figure 1, though later we'll look at fixing this assumption. A generalization is that conductors are opaque and dielectrics are transparent. This gets confusing since most of the dielectric surfaces that we are interested in modeling are mixtures and don't fall into the simple models we've described so far. Consider a thick colored lacquer surface. The lacquer itself is transparent, but suspended in the lacquer are reflective pigment off of which light gets reflected, bounced, split, shifted or altered before perhaps reemerging from the surface. This can be seen in Figure 3, where the light rays are not just reflected but bounced around a bit inside the medium before getting retransmitted to the outside.

236 Copyright Virtual University of Pakistan

25-Mathiematics of Lighting and Shading Part I

VU

Figure 3: Subsurface scattering typical of pigment-saturated translucent coatings. Metallic paint, brushed metal, velvet, etc. are all materials for which we need to examine better models to try to represent these surfaces. But with a little creativity in the modeling, it's possible to mimic the effect. Figure 4 shows what you get when you use multiple broad specular terms for multiple base colors combined with a more traditional shiny specular term. There's also a high-frequency normal perturbation that simulates the sparkle from a metallic flake pigment. As we can see, we can get something that looks particularly striking with a fairly simple model.

Figure 4: A simple shader to simulate metallic paint: (a) shows the two-tone paint shading pass; (b) shows the specular sparkle shading pass; (c) shows the environment mapping pass; (d) shows the final composite image The traditional model gives us a specular term and a diffuse term. We have been able to add in texture maps to give our scenes some uniqueness, but the lighting effects have been very simple. Shaders allow us to be much more creative with lighting effects. As Figure 4 shows, with just a few additional specular terms, we can bring forth a very interesting look. But before we go off writing shaders, we'll need to take a look at how it all fits together in the graphics pipeline. And a good place to start is by examining the traditional lighting model.

237 Copyright Virtual University of Pakistan

26-Mathiematics of Lighting and Shading Part II Light Types and Shading Models

VU

Lecture No.26
Light Types

Mathematics of Lighting and Shading Part II Light Types and Shading Models

Now that we have a way to find the light hitting a surface, we're going to need some lights! There are three types of lights we are going to discuss. I. Parallel Lights (or Directional Lights) Parallel lights cheat a little bit. They represent light that comes from an infinitely far away light source. Because of this, all of the light rays that reach the object are parallel (hence the name). The standard use of parallel lights is to simulate the sun. While it's not infinitely far away, 93 million miles is good enough! The great thing about parallel lights is that a lot of the math goes away. The attenuation factor is always 1 (for point/spotlights, it generally involves divisions if not square roots). The incoming light vector for calculation of the diffuse reflection factor is the same for all considered points, whereas point lights and spotlights involve vector subtractions and a normalization per vertex. Typically, lighting is the kind of effect that is sacrificed for processing speed. Parallel light sources are the easiest and therefore fastest to process. If we can't afford to do the nicer point lights or spotlights, falling back to parallel lights can keep our frame rates at reasonable levels. II. Point Lights Point lights are one step better than directional lights. They represent infinitesimally small points that emit light. Light scatters out equally in all directions. Depending on how much effort we're willing to expend on the light, we can have the intensity falloff based on the inverse squared distance from the light, which is how real lights work.

The light direction is different for each surface location (otherwise the point light would look just like a directional light). The equation for it is:

238 Copyright Virtual University of Pakistan

26-Mathiematics of Lighting and Shading Part II Light Types and Shading Models

VU

Figure 1: Point light sources III. Spotlights Spotlights are the most expensive type of light we discuss in this course and should be avoided if possible because it is not for real time environment. We model a spotlight not unlike the type we would see in a theatrical production. They are point lights, but light only leaves the point in a particular direction, spreading out based on the aperture of the light. Spotlights have two angles associated with them. One is the internal cone whose angle is generally referred to as theta (). Points within the internal cone receive all of the light of the spotlight; the attenuation is the same as it would be if point lights were used. There is also an angle that defines the outer cone; the angle is referred to as phi. Points outside the outer cone receive no light. Points outside the inner cone but inside the outer cone receive light, usually a linear falloff based on how close it is to the inner cone.

Figure 2: A spotlight If we think all of this sounds mathematically expensive, we're right. Some library packages like OpenGL and Direct3D implements lighting for us, so we won't need to worry about the implementation of the math behind spotlights, but rest assured that they're extremely expensive and can slow down our graphics application a great deal. 239 Copyright Virtual University of Pakistan

26-Mathiematics of Lighting and Shading Part II Light Types and Shading Models

VU

Then again, they do provide an incredible amount of atmosphere when used correctly, so we will have to figure out a line between performance and aesthetics. Shading Models Once we've found basic lighting information, we need to know how to draw the triangles with the supplied information. There are currently three ways to do this; the third has just become a hardware feature with DirectX 9.0 In our previous lectures we have already studied flat and gouraud shading triangle algorithms. I. Lambert Triangles that use Lambertian shading are painted with one solid color instead of using a gradient. Typically each triangle is lit using that triangle's normal. The resulting object looks very angular and sharp. Lambertian shading was used mostly back when computers weren't fast enough to do Gouraud shading in real time. To light a triangle, you compute the lighting equations using the triangle's normal and any of the three vertices of the triangle.

Figure 3: Flat shaded view of our polygon mesh II. Gouraud Gouraud (pronounced garrow) shading is the current de facto shading standard in accelerated 3D hardware. Instead of specifying one color to use for the entire triangle, each vertex has its own separate color. The color values are linearly interpolated across the triangle, creating a smooth transition between the vertex color values. To calculate the lighting for a vertex, we use the position of the vertex and a vertex normal. Of course, it's a little hard to correctly define a normal for a vertex. What people do instead is average the normals of all the polygons that share a certain vertex, using that as the vertex normal. When the object is drawn, the lighting color is found for each vertex (rather than each polygon), and then the colors are linearly interpolated across the object. This creates a slick and smooth look, like the one in Figure 4.

240 Copyright Virtual University of Pakistan

26-Mathiematics of Lighting and Shading Part II Light Types and Shading Models

VU

Figure 4: Gouraud shaded view of our polygon mesh One problem with Gouraud shading is that the triangles' intensities can never be greater than the intensities at the edges. So if there is a spotlight shining directly into the center of a large triangle, Gouraud shading will interpolate the intensities at the three dark corners, resulting in an incorrectly dark triangle. The internal highlighting problem usually isn't that bad. If there are enough triangles in the model, the interpolation done by Gouraud shading is usually good enough. If we really want internal highlights but only have Gouraud shading, we can subdivide the triangle into smaller pieces. III. Phong Phong shading is the most realistic shading model We are going to talk about, and also the most computationally expensive. It tries to solve several problems that arise when we use Gouraud shading. If we're looking for something more realistic, some authors have also discussed nicer shading models like Tarrence-Sparrow, but they aren't real time (at least not right now). First of all, Gouraud shading uses a linear gradient. Many objects in real life have sharp highlights, such as the shiny spot on an apple. This is difficult to handle with pure Gouraud shading. The way Phong does this is by interpolating the normal across the triangle face, not the color value, and the lighting equation is solved individually for each pixel.

Figure 5: Phong shaded view of a polygon mesh Phong shading isn't technically supported in hardware. But we can now program our own Phong rendering engine, and many other special effects, using shaders, a hot new technology. 241 Copyright Virtual University of Pakistan

27-Review II

VU

Lecture No.27
27.1 CLIPPING - Concept

Review II

It is desirable to restrict the effect of graphics primitives to a sub-region of the canvas, to protect other portions of the canvas. All primitives are clipped to the boundaries of this clipping rectangle; that is, primitives lying outside the clip rectangle are not drawn. The default clipping rectangle is the full canvas (the screen), and it is obvious that we cannot see any graphics primitives outside the screen. A simple example of line clipping can illustrate this idea: This is a simple example of line clipping: the display window is the canvas and also the default clipping rectangle, thus all line segments inside the canvas are drawn. The red box is the clipping rectangle we will use later, and the dotted line is the extension of the four edges of the clipping rectangle.

242 Copyright Virtual University of Pakistan

27-Review II

VU

27.2 Point Clipping Assuming a rectangular clip window, point clipping is easy. we save the point if: xmin <= x <=xmax ymin <= y <= ymax 27.3 Line Clipping

This section treats clipping of lines against rectangles. Although there are specialized algorithms for rectangle and polygon clipping, it is important to note that other graphic primitives can be clipped by repeated application of the line clipper. 27.4 Cohen-Sutherland algorithm - Conclusion

In summary, the Cohen-Sutherland algorithm is efficient when out-code testing can be done cheaply (for example, by doing bit-wise operations in assembly language) and trivial acceptance or rejection is applicable to the majority of line segments. (For example, large windows - everything is inside, or small windows - everything is outside). 27.5 Liang-Barsky Algorithm - Conclusion

In general, the Liang_Barsky algorithm is more efficient than the Cohen_Sutherland algorithm, since intersection calculations are reduced. Each update of parameters u1 and u2 requires only one division; and window intersections of the line are computed only once, when the final values of u1 and u2 have computed. In contrast, the CohenSutherland algorithm can repeatedly calculate intersections along a line path, even though the line may be completely outside the clip window, and, each intersection calculation requires both a division and a multiplication. Both the Cohen_Sutherland and the Liang_Barsky algorithms can be extended to three-dimensional clipping. 27.6 Polygon Clipping A polygon is usually defined by a sequence of vertices and edges. If the polygons are unfilled, line-clipping techniques are sufficient however, if the polygons are filled, the process in more complicated. A polygon may be fragmented into several polygons in the clipping process, and the original colour associated with each one. The SutherlandHodgeman clipping algorithm clips any polygon against a convex clip polygon. The Weiler-Atherton clipping algorithm will clip any polygon against any clip polygon. The polygons may even have holes. The following example illustrates a simple case of polygon clipping.

243 Copyright Virtual University of Pakistan

27-Review II

VU

27.7

Sutherland and Hodgman's polygon-clipping algorithm:-

Sutherland and Hodgman's polygon-clipping algorithm uses a divide-and-conquer strategy: It solves a series of simple and identical problems that, when combined, solve the overall problem. The simple problem is to clip a polygon against a single infinite clip edge. Four clip edges, each defining one boundary of the clip rectangle, successively clip a polygon against a clip rectangle. Note the difference between this strategy for a polygon and the Cohen-Sutherland algorithm for clipping a line: The polygon clipper clips against four edges in succession, whereas the line clipper tests the outcode to see which edge is crossed, and clips only when necessary. 27.8

Steps of Sutherland-Hodgman's polygon-clipping algorithm

Polygons can be clipped against each edge of the window one at a time. Windows/edge intersections, if any, are easy to find since the X or Y coordinates are already known. Vertices which are kept after clipping against one window edge are saved for clipping against the remaining edges. Note that the number of vertices usually changes and will often increase.

We are using the Divide and Conquer approach. 27.9 Shortcoming of Sutherlands -Hodgeman Algorithm Convex polygons are correctly clipped by the Sutherland-Hodegeman algorithm, but concave polygons may be displayed with extraneous lines. This occurs when the clipped polygon should have two or more separate sections. But since there is only one output vertex list, the last vertex in the list is always joined to the first vertex. There are several things we could do to correct display concave polygons. For one, we could split the concave polygon into two or more convex polygons and process each convex polygon separately.

244 Copyright Virtual University of Pakistan

27-Review II

VU

Another approach to check the final vertex list for multiple vertex points along any clip window boundary and correctly join pairs of vertices. Finally, we could use a more general polygon clipper, such as wither the Weiler-Atherton algorithm or the Weiler algorithm described in the next section. 27.10 Weiler-Atherton Polygon Clipping In this technique, the vertex-processing procedures for window boundaries are modified so that concave polygons are displayed correctly. This clipping procedure was developed as a method for identifying visible surfaces, and so it can be applied with arbitrary polygon-clipping regions. The basic idea in this algorithm is that instead of always proceeding around the polygon edges as vertices are processed, we sometimes want to follow the window boundaries. Which path we follow depends on the polygon-processing direction(clockwise or counterclockwise) and whether the pair of polygon vertices currently being processed represents an outside-to-inside pair or an inside-to-outside pair. For clockwise processing of polygon vertices, we use the following rules: For an outside-top inside pair of vertices, follow the polygon boundary For an inside-to-outside pair of vertices, follow the window boundary in a clockwise direction

In following figure, the processing direction in the Wieler-Atherton algorithm and the resulting clipped polygon is shown for a rectangular clipping window.
Inside

go up 1

5 8 go up 6

245 Copyright Virtual University of Pakistan

27-Review II

VU

27.11

3D Concepts

27.12 Coordinate Systems Coordinate systems are the measured frames of reference within which geometry is defined, manipulated and viewed. In this system, you have a well-known point that serves as the origin (reference point), and three lines(axes) that pass through this point and are orthogonal to each other ( at right angles 90 degrees). With the Cartesian coordinate system, you can define any point in space by saying how far along each of the three axes you need to travel in order to reach the point if you start at the origin. Following are three types of the coordinate systems. 27.13 1-D Coordinate Systems:

This system has the following characteristics:

Direction and magnitude along a single axis, with reference to an origin Locations are defined by a single coordinate Can define points, segments, lines, rays Can have multiple origins (frames of reference) and transform coordinates among them

27.14 2-D Coordinate Systems:

Direction and magnitude along two axes, with reference to an origin Locations are defined by x, y coordinate pairs Can define points, segments, lines, rays, curves, polygons, (any planar geometry) Can have multiple origins (frames of reference and transform coordinates among them 246 Copyright Virtual University of Pakistan

27-Review II

VU

27.15 3-D Coordinate Systems:

3D Cartesian coordinate systems Direction and magnitude along three axes, with reference to an origin Locations are defined by x, y, z triples Can define cubes, cones, spheres, etc., (volumes in space) in addition to all oneand two-dimensional entities Can have multiple origins (frames of reference) and transform coordinates among them

27.16 Left-handed versus Right-handed

Determines orientation of axes and direction of rotations Thumb = pos x, Index up = pos y, Middle out = pos z Most world and object axes tend to be right handed 247 Copyright Virtual University of Pakistan

27-Review II

VU

Left handed axes often are used for cameras

27.17 Right Handed Rule: Right Hand Rule for rotations: grasp axis with right hand with thumb oriented in positive direction, fingers will then curl in direction of positive rotation for that

axis. Right handed Cartesian coordinate system describes the relationship of the X,Y, and Z in the following manner: X is positive to the right of the origin, and negative to the left. Y is positive above the origin, and negative below it. Z is negative beyond the origin, and positive behind it.

Origin

+Y North -Z

+X
West

East

+Z Sout 248 Copyright Virtual University of Pakistan

27-Review II

VU

27.18 Left Handed Rule:

Origin

+Y North +Z

+X
West

East

-Z Sout

Left handed Cartesian coordinate system describes the relationship of the X, Y and Z in the following manner: X is positive to the right of the origin, and negative to the left. Y is positive above the origin, and negative below it. Z is positive beyond the origin, and negative behind it.

27.19 Defining 3D points in mathematical notations 3D points can be described using simple mathematical notations P = (X, Y, Z) Thus the origin of the Coordinate system is located at point (0,0,0), while five units to the right of that position might be located at point (5,0,0).

249 Copyright Virtual University of Pakistan

27-Review II

VU

27.20 Y-up versus Z-up:

z-up typically used by designers y-up typically used by animators orientation by profession supposedly derives from past work habits often handled differently when moving from application to application

27.21 Global and Local Coordinate Systems:

Local coordinate systems can be defined with respect to global coordinate system Locations can be relative to any of these coordinate systems Locations can be translated or "transformed" from one coordinate system to another.

250 Copyright Virtual University of Pakistan

27-Review II

VU

27.22 Multiple Frames of Reference in a 3-D Scene:

In fact, there usually are multiple coordinate systems within any 3-D screen Application data will be transformed among the various coordinate systems, depending on what's to be accomplished during program execution Individual coordinate systems often are hierarchically linked within the scene

27.23 The Polar Coordinate System Cartesian systems are not the only ones we can use. We could have also described the object position in this way: starting at the origin, looking east, rotate 38 degrees northward, 65 degrees upward, and travel 7.47 feet along this line. As you can see, this is less intuitive in a real world setting. And if you try to work out the math, it is harder to manipulate (when we get to the sections that move points around). Because such polar coordinates are difficult to control, they are generally not used in 3D graphics. 27.24 Defining Geometry in 3-D Here are some definitions of the technical names that will be used in 3D lectures. Modeling: is the process of describing an object or scene so that we can construct an image of it.

251 Copyright Virtual University of Pakistan

27-Review II

VU

27.25 Points & polygons:

Points: three-dimensional locations (or coordinate triples)

Vectors: - have direction and magnitude; can also be thought of as displacement

Polygons: - sequences of correctly co-planar points; or an initial point and a sequence of vectors

27.26 Primitives Primitives are the fundamental geometric entities within a given data structure. 252 Copyright Virtual University of Pakistan

27-Review II

VU

We have already touched on point, vector and polygon primitives

Regular Polygon Primitives - square, triangle, circle, n-polygon, etc.

Polygon strips or meshes Meshes provide a more economical description than multiple individual polygons For example, 100 individual triangles, each requiring 3 vertices, would require 100 x 3 or 300 vertex definitions to be stored in the 3-D database. By contrast, triangle strips require n + 2 vertex definitions for any n number or triangles in the strip. Hence, a 100 triangle strip requires only 102 unique vertex definitions.

Meshes also provide continuity across surfaces which is important for shading calculations

3D primitives in a polygonal database 253 Copyright Virtual University of Pakistan

27-Review II

VU

3D shapes are represented by polygonal meshes that define or approximate geometric surfaces.

With curved surfaces, the accuracy of the approximation is directly proportional to the number of polygons used in the representation. More polygons (when well used) yield a better approximation. But more polygons also exact greater computational overhead, thereby degrading interactive performance, increasing render times, etc.

27.27 Rendering

The process of computing a two dimensional image using a combination of a threedimensional database, scene characteristics, and viewing transformations. Various algorithms can be employed for rendering, depending on the needs of the application.

27.28 Tessellation

The subdivision of an entity or surface into one or more non-overlapping primitives. Typically, renderers decompose surfaces into triangles as part of the rendering process.

27.29 Sampling

The process of selecting a representative but finite number of values along a continuous function sufficient to render a reasonable approximation of the function for the task at hand.

27.30 Level of Detail (LOD)

To improve rendering efficiency when dynamically viewing a scene, more or less detailed versions of a model may be swapped in and out of the scene database depending on the importance (usually determined by image size) of the object in the current view.

27.31 Transformations The process of moving points in space is called transformation. 254 Copyright Virtual University of Pakistan

27-Review II

VU

27.32 Types of Transformation There are various types of transformations as we have seen in case of 2D transformations. These include: a. Translation b. Rotation c. Scaling d. Reflection e. Shearing a) Translation Translation is used to move a point, or a set of points, linearly in space. Since now we are talking about 3D, therefore each point has 3 coordinates i.e. x, y and z. similarly, the translation distances can also be specified in any of the 3 dimensions. These Translation Distances are given by tx, ty and tz. For any point P(x,y,z) after translation we have P(x,y,z) where x = x + tx , y = y + ty , z = z + tz and (tx, ty , tz) is Translation vector Now this can be expressed as a single matrix equation: P = P + T Where: x P= y z
x ' P'= y ' z ' t x T = t y t z

3D Translation Example We may want to move a point 3 meters east, -2 meters up, and 4 meters north. What would be done in such event? Steps for Translation Given a point in 3D and a translation vector, it can be translated as follows: Point3D point = (0, 0, 0) Vector3D vector = (10, -3, 2.5) Adding vector to point point.x = point.x + vector.x; point.y = point.y + vector.y; point.z = point.z + vector.z; And finally we have translated point. Homogeneous Coordinates Analogous to their 2D Counterpart, the homogeneous coordinates for 3D translation can be expressed as :

x' 1 0 0 t x x y ' 0 1 0 t y y = . z ' 0 0 1 t z z 1 0 0 0 1 1 Copyright Virtual University of Pakistan

255

27-Review II

VU

Abbreviated as: P = T (tx, ty , tz) . P On solving the RHS of the matrix equation, we get:

Which shows that each of the 3 coordinates gets translated by the corresponding translation distance.

x' x + t x y ' y + t y = z ' z + t z 1 1

Rotation Rotation is the process of moving a point in space in a non-linear manner

We need to know three different angles: How far to rotate around the X axis(YZ rotation, or pitch) How far to rotate around the Y axis (XZ rotation, or yaw) How far to rotate around the Z axis (XY rotation, or roll)
Column vector representation: P = R . P Where
x' P' = y ' cos R= sin sin cos x P= y

Rotation: Homogeneous Coordinates The rotation can now be expressed using homogeneous coordinates as:
x' cos y ' = sin 0 1 sin cos 0 0 x 0 . y 1 1

Abbreviated as:

P = R () . P Now in 3D Rotation can be about any of the three axes: About z-axis (i.e. in xy plane) About x-axis (i.e. in yz plane) About y-axis (i.e. in xz plane) Roll : around z-axis Pitch: around x-axis Yaw: around y-axis 256 Copyright Virtual University of Pakistan

27-Review II

VU

Rotation about z-axis (i.e. in xy plane): x = x cos y sin y = x sin + y cos z =z

by Cyclic permutation Rotation about x-axis (i.e. in yz plane): x =x y = y cos z sin z = y sin + z cos and Rotation about y-axis (i.e. in xz plane): x = z sin + x cos y =y z = z cos x sin

b) SCALING:-

Coordinate transformations for scaling relative to the origin are X` = X . Sx Y` = Y . Sy Z` = Z . Sz

Uniform Scaling

We preserve the original shape of an object with a uniform scaling ( Sx = Sy = Sz)
Differential Scaling

We do not preserve the original shape of an object with a differential scaling ( Sx <> Sy <> Sz)
S x 0 0 0 0 Sy 0 0 0 0 Sz 0 0 0 0 1

Scaling w.r.t. Origin

27.33 PROJECTION Projection can be defined as a mapping of point P(x,y,z) onto its image P`(x`,y`,z`) in the projection plane or view plane, which constitutes the display surface

257 Copyright Virtual University of Pakistan

27-Review II

VU

Methods of Projection

Parallel Projection Perspective Projection

Parallel Projection is divided into Orthographic and Oblique transformations.

Orthographic
258 Copyright Virtual University of Pakistan

27-Review II

VU

Oblique

Orthographic projection has two types Multiview Axonometric Multiview: There are three orthographic views of an object.

Axonometric projections: There are three axonometric projections:

Isometric Dimetric Trimetric

259 Copyright Virtual University of Pakistan

27-Review II

VU

1. Isometric The projection plane intersects each coordinate axis in the model coordinate system at an equal distance or the direction of projection makes equal angles with all of the three principal axes 2. Dimetric The direction of projection makes equal angles with exactly two of the principal axes 3. Trimetric The direction of projection makes unequal angles with the three principal axes

Xp Yp

= x + z ( L1 cos () ) = y + z ( L1 sin () )

Where L1 = L/z

260 Copyright Virtual University of Pakistan

28-Review III

VU

Lecture No.28
28.1

Review III

Perspective Projection

As opposed to parallel projection, perspective projection gives a more realistic view of the objects in the scene. The objects away from the POV are projected smaller whereas those nearby are projected to appear proportionately larger. The idea is that of looking at the scene through the projection plane / screen. The POV represents the viewers eye, and we presume that the viewer will be behind the center of the screen. Note: We will use Left hand rule to describe 3D coordinate system. Two common approaches are used with this; 1. The first approach is where the POV is at some point (0, 0, -z) and the screen lies on the X-Y plane, graphically, this looks like figure given below: 2. The second approach is where the POV lies at the origin, and the screen lies on a plane at some +z coordinate, as shown in figure given below: As we will see later, this second approach is much more convenient when we add features making it possible for the POV to move around the 3D world or for objects to move around in the world. Calculating the screen pixel that correlates to a 3D point is now a matter of simple geometry. From a viewpoint above the screen and POV (looking at the X-Z plane), the geometry appears like the one shown in figure below: In geometric terms, we say that the triangle from A to B to S is similar to the triangle from A to C to P because the three angles that make up the triangles are the same: the angle from AB to AS is the same as the angle from AC to AP, the two right angles are both 90 degrees, and therefore the remaining two angles are the same (the sum of the angles in a triangle is always 180 degrees). What also holds true from similar triangles is that the ratio of two sides holds between the similar triangles; this means that the ratio of BS to AB is the same as the ratio of CP to AC. But we know what AB is-it is Screen.z ! and we 261 Copyright Virtual University of Pakistan

28-Review III

VU

know what AC is-it is point.z ! and we know what CP is-it is point.x ! therefore: |BS| / |AB| = |CP| / |AC| |BS| = |AB| * |CP| / |AC| |BS| = Screen.z * point.x / point.z Screen.z is the distance d from the point of view at origin or the scaling factor. Notice that |BS| is the length of the line segment that goes from B to S in world units. But we normally address the screen with the point (0,0) at the top left, with +X pixels moving to the right, and +Y pixels moving downand not from the middle of the screen. And we draw to the screen in pixel units not our world units (unless, of course, 1.0 in your world represents one pixel).

28.2

Triangles

Triangles are to 3D graphics what pixels are to 2D graphics. Every PC hardware accelerator under the sun uses triangles as the fundamental drawing primitive (well scan line aligned trapezoids actually, but that's a hardware implementation issue). When you draw a polygon, hardware devices really draw a fan of triangles. Triangles "flesh out" a 3D object, connecting them together to form a skin or mesh that defines the boundary surface of an object. Triangles, like polygons, generally have an orientation associated with them, to help in normal calculations. The ordering of the vertices goes clockwise around the triangle. Figure below shows what a clockwise ordered triangle would look like.

Figure: Three points in space, and the triangle connecting them When defining a mesh of triangles that define the boundary of a solid, you set it up so that all of the triangles along the skin are ordered clockwise when viewed from the outside. 262 Copyright Virtual University of Pakistan

28-Review III

VU

It is impossible to see triangles that face away from you. (You can find this out by computing the triangle's plane normal and performing a dot product with a vector from the camera location to a location on the plane.) Now let's move on to the code. To help facilitate using the multiple types, I'll implement triangles structure. I only define constructors and keep the access public. struct tri { type v[3]; // Array access useful for loops tri() { // nothing } tri( type v0, type v1, type v2 ) { v[0] = v0; v[1] = v1; v[2] = v2; } };

28.3

Strips and Fans

Lists of triangles are generally represented in one of three ways. The first is an explicit list or array of triangles, where every three elements represent a new triangle. However, there are two additional representations, designed to save bandwidth while sending triangles to dedicated hardware to draw them. They are called triangle strips and triangle fans. Triangle fans, conceptually, look like the folding fans you see in Asian souvenir shops. They are a list of triangles that all share a common point. The first three elements indicate the first triangle. Then each new element is combined with the first element and the current last element to form a new triangle. Note that an N-sided polygon can be represented efficiently using a triangle fan, Figure below illustrates what I'm talking about.

263 Copyright Virtual University of Pakistan

28-Review III

VU

Figure: A list of points composing a triangle fan Triangles in a triangle strip, instead of sharing a common element with all other triangles like a fan, only share elements with the triangle immediately preceding them. The first three elements define the first triangle. Then each subsequent element is combined with the two elements before it, in clockwise order, to create a new triangle. See Figure below for an explanation of strips.

Figure: A list of points composing a triangle strip

28.4

Planes

The next primitive to discuss is the plane. Planes are to 3D what lines are in 2D; they're n1 dimensional hyper-planes that can help you accomplish various tasks. Planes are defined as infinitely large, infinitely thin slices of space, like big pieces of paper. Triangles that make up your model each exist in their own plane. When you have a plane that represents a slice of 3D space, you can perform operations like classification of points and polygons and clipping. So how do you represent planes? Well it is best to build a structure from the equation that defines a plane in 3D. The implicit equation for a plane is:

264 Copyright Virtual University of Pakistan

28-Review III

VU

What do these numbers represent? The triplet <a,b,c> represents what is called the normal of the plane. A normal is a unit vector that, conceptually speaking, sticks directly out of a plane. A stronger mathematical definition would be that the normal is a vector that is perpendicular to all of the points that lie in the plane. The d component in the equation represents the distance from the plane to the origin. The distance is computed by tracing a line towards the plane until you hit it. Finally the triplet <x,y,z> is any point that satisfies the equation. The set of all points <x,y,z> that solve the equation is exactly all the points that lie in the plane. All of the pictures I'm showing you will be of the top-down variety, and the 3D planes will be on edge, appearing as 2D lines. This makes figure drawing much easier.

Following are two examples of planes. The first has the normal pointing away from the origin, which causes d to be negative (try some sample values for yourself if this doesn't make sense). The second has the normal pointing towards the origin, so d is positive. Of course, if the plane goes through the origin, d is zero (the distance from the plane to the origin is zero). Figures 1 and Figure 2 provide some insight into this relation.

Figure 1: d is negative when the normal faces away from the origin

Figure 2: d is positive when it faces towards the origin 265 Copyright Virtual University of Pakistan

28-Review III

VU

It's important to notice that technically the normal <a,b,c> does not have to be unit-length for it to have a valid plane equation. But since things end up nicer if the normal is unitlength. Constructing a plane given three points that lie in the plane is a simple task. You just perform a cross product between the two vectors made up by the three points <point2 point0> <point1 point0> and find a normal for the plane. After generating the normal and making it unit length, finding the d value for the plane is just a matter of storing the negative dot product of the normal with any of the points. This holds because it essentially solves the plane equation above for d. Of course plugging a point in the plane equation will make it equal 0, and this constructor has three of them.

28.5

Back-face Culling

Now that you know how to define a point with respect to a plane, you can perform backface culling, one of the most fundamental optimization techniques of 3D graphics. Let's suppose you have a triangle whose elements are ordered in such a fashion that when viewing the triangle from the front, the elements appear in clockwise order. Back-face culling allows you to take triangles defined with this method and use the plane equation to discard triangles that are facing away. Conceptually, any closed mesh, a cube for example, will have some triangles facing you and some facing away. You know for a fact that you'll never be able to see a polygon that faces away from you; they are always hidden by triangles facing towards you. This, of course, doesn't hold if you're allowed to view the cube from its inside, but this shouldn't be allowed to happen if you want to really optimize your engine. Rather than perform the work necessary to draw all of the triangles on the screen, you can use the plane equation to find out if a triangle is facing towards the camera, and discard it if it is not. How is this achieved? Given the three points of the triangle, you can define a plane that the triangle sits in. Since you know the elements of the triangle are listed in clockwise order, you also know that if you pass the elements in order to the plane constructor, the normal to the plane will be on the front side of the triangle. If you then think of the location of the camera as a point, all you need to do is perform a point-plane test. If the point of the camera is in front of the plane, then the triangle is visible and should be drawn. There's an optimization to be had. Since you know three points that lie in the plane (the three points of the triangle) you only need to hold onto the normal of the plane, not the entire plane equation. To perform the back-face cull, just subtract one of the triangle's points from the camera location and perform a dot product with the resultant vector and the normal. If the result of the dot product is greater than zero, then the view point was in front of the triangle. Figure below can help explain the point.

266 Copyright Virtual University of Pakistan

28-Review III

VU

Figure: A visual example of back-face culling In practice, 3D accelerators can actually perform back-face culling by themselves, so as the triangle rates of cards increase, the amount of manual back-face culling that is performed has steadily decreased. However, the information is useful for custom 3D engines that don't plan on using the facilities of direct hardware acceleration.

28.6

Intersection between a Line and a Plane (1) (2)

This occurs at the point which satisfies both the line and the plane equations.

Line equation: p = org + u * dir Plane equation: p * normal - k = 0.

Substituting (1) into (2) and rearranging we get:

(org + u * dir) * normal - k = 0 ie u * dir * normal = k - org * normal ie u = (k - org * normal) / (dir * normal)
If (d * normal) = 0 then the line runs parrallel to the plane and no intersection occurs. The exact point at which intersection does occur can be found by plugging u back into the line equation in (1).

28.7 Triangle Rasterization High performance triangle rasterization is a very important topic in Computer Graphics in todays world. Triangles are the foundation of modern real time graphics, and are by far the most popular rendering primitive. Most computer games released in the last few years are almost completely dependent on triangle rasterization performance. Recently the focus of graphics performance optimization is beginning to shift to bandwidth requirements as well as transformation and lighting. Nevertheless, rasterization performance is still a factor, and this lecture will provide most of the basics of high performance triangle rasterization. Also, it will go into detail about two often neglected rendering quality improvements, sub-pixel and sub-texel accuracy. Also, smooth shading and texture mapping techniques will be described.

267 Copyright Virtual University of Pakistan

28-Review III

VU

28.8

Flat Filling Triangles

Drawing triangle (or in general convex polygon, but as we discussed we will use only triangles) is very simple. The basic idea of the line triangle drawing algorithm is as follows. For each scan line (horizontal line on the screen), find the points of intersection with the edges of the triangle. Then, draw a horizontal line between intersections and do this for all scan lines. But how can we find these points quickly? Using linear interpolation! We have 3 vertices and we want to find coordinates of all points belonging to segments determined by these vertices. Assume we have segment given by points: (xa,ya) and (xb,yb). Our task is to find points: (xc,ya+1), (xd,ya+2), ... , (xm,yb-1), (xn,yb). Notice that xa changes to xb in (yb-ya) steps. We also have: xa=xa+0*(xb-xa)/(yb-ya), Xb = xa + ( yb-ya ) * ( xb-xa ) / ( yb-ya ) and, in general, xi = xa + ( yi ya ) * delta , where delta=(xb-xa)/(yb-ya). The general function for linear interpolation is: f(X) = A + X * ( (B-A) / steps) where we slide from A to B in steps steps

Here is pseudo code for a triangle filling algorithm.

The coordinates of vertices are (A. x,A. y), (B. x,B. y), (C. x,C. y); we assume that A. y <= B. y <= C. y (you should sort them first) dx1,dx2,dx3 are deltas used in interpolation Horizontal line draws horizontal segment with coordinates (S. x, Y), (E. x, Y) S. x, E. x are left and right x-coordinates of the segment we have to draw S = A means that S. x = A. x; S. y = A. y; 268 Copyright Virtual University of Pakistan

28-Review III

VU

if (B.y - A.y > 0) dx1 = (B.x-A.x) / (B.y-A.y) else dx1=0; if (C.y - A.y > 0) dx2=(C.x-A.x)/(C.y-A.y) else dx2=0; if (C.y-B.y > 0) dx3=(C.x-B.x)/(C.y-B.y) else dx3=0; S=E=A if(dx1 > dx2) { for( ; S.y<=B.y; S.y++, E.y++, S.x+=dx2, E.x += dx1) horizontal line (S.x,E.x,S.y, E.x,color); E=B; for(;S.y<=C.y;S.y++,E.y++,S.x+=dx2,E.x+=dx3) horizontal line(S.x,E.x,S.y, E.x,color); } else { for(;S.y<=B.y;S.y++,E.y++,S.x+=dx1,E.x+=dx2) horizontal line(S.x,E.x,S.y, E.x,color); S=B; for(;S.y<=C.y;S.y++,E.y++,S.x+=dx3,E.x+=dx2) horizontal line(S.x,E.x,S.y, E.x,color); } I ought to explain what is the comparison dx1 > dx2 for. It's optimization trick: in the horizontal line routine, we don't need to compare the x's (S.x is always less than or equal to E.x).

269 Copyright Virtual University of Pakistan

28-Review III

VU

28.9

Gouraud Shading

The idea of gouraud and flat triangle is nearly the same. Gouraud takes only three parameters more (the color value of each of the vertices), and the routine just interpolates among them drawing a beautiful, shaded triangle. You can use 256-colors mode, in which vertices' colors are simply indices to palette or hicolor mode (recommended). Flat triangle interpolated only one value (x in connection with y), 256 colors gouraud needs three (x related to y, color related to y, and color related to x), hi-color gouraud needs seven (x related to y, red, green and blue components of color related to y, and color related to x (also three components)) Drawing a gouraud triangle, we add only two parts to the flat triangle routine. The horizline routine gets a bit more complicated due to the interpolation of the color value related to x but the main routine itself remains nearly the same.

28.10 Textured Triangles

We can also apply any bitmap on triangle for filling it. I'll show you the idea of linear (or 'classical') texture mapping (without perspective correction). Linear mapping works pretty well (read: fast) in some scenes, but perspective correction is in some way needed in most 3D systems. Again we're using the idea of interpolation: now we'll code a texture triangle filler. And again the idea is perfectly the same, only two more values to interpolate, that is five values total. In texture mapping, we interpolate x, u, and v related to y, and u and v related to x (u and v are coordinates in the 2D bitmap space). The situation is maybe easier to understand by looking at the following picture:

270 Copyright Virtual University of Pakistan

28-Review III

VU

The left triangle is the triangle which is drawn onto the screen. There's a single scanline (one call to the horizline routine) pointed out as an example. The triangle on the right is the same triangle in the bitmap space, and there's the same scanline drawn from another point of view into it, too. So we need just to interpolate, interpolate, and once more interpolate in texture filler - an easy job if you've understood the idea of gouraud filler.

28.11 COLOR It is important to understand how color is represented in computer graphics so that we can manipulate it effectively. A color is usually represented in the graphics pipeline by a three-element vector representing the intensities of the red, green, and blue components, or for a more complex object, by a four-element vector containing an additional value called the alpha component that represents the opacity of the color. Thus we can talk about rgb or rgba colors and mean a color that's made up of either three or four elements. There are many different ways of representing the intensity of a particular color element. Colors can also be represented as floating point values in the range [0,1].
Nowadays every PC we can buy has hardware that can render images with thousands or millions of individual colors. Rather than have an array with thousands of color entries, the images instead contain explicit color values for each pixel. A 16-bit display is named since each pixel in a 16-bit image is taken up by 16 bits (2 bytes): 5 bits of red information, 6 bits of green information, and 5 bits of blue information. Incidentally, the extra bit (and therefore twice as much color resolution) is given to green because our eyes are more sensitive to green. A 24-bit display, of course, uses 24 bits, or 3 bytes per pixel, for color information. This gives 1 byte, or 256 distinct values each, for red, green, and blue. This is generally called true color, because 2563 (16.7 million) colors is about as much as your eyes can discern, so more color resolution really isn't necessary, at least for computer monitors. Finally, there is 32-bit color, something seen on most new graphics cards. Many 3D accelerators keep 8 extra bits per pixel around to store transparency information, which is generally referred to as the alpha channel, and therefore take up 4 bytes, or 32 bits, of storage per pixel. Rather than reimplement the display logic on 2D displays that don't need alpha information, these 8 bits are usually just wasted.

WHY WE MIGHT WANT 128-BIT COLOR?

In one of the early magazines articles of Mike Abrash [ABRASH 1992], he tells a story about going from a 256-color palette to hardware that supported 256 levels for each RGB color16 million colors! What would we do with all those colors? He goes on to tell of a story by Sheldon Linker at the eighth Annual Computer Graphics Show on how the folks at the Jet Propulsion Lab back in the 1970s had a printer that could print over 50 million distinct colors. As a test, they printed out words on paper where the background color was only one color index from the word's color. To their surprise, it was easy to discern the wordsthe human eye is very sensitive to color graduations and edge detection. The JPL team then did the same tests on color monitors and discovered that only about 16 million colors could be distinguished. It seems that the eye is (not too surprisingly) better at perceiving detail from reflected light (such as from a printed page) than from emissive light (such as from a CRT). The moral is that the eye is a lot more perceptive than you might think. Twenty four-bits of color really is not that much range, particularly if we are performing multiple passes. Round-off error can and will show up if we aren't careful! 271 Copyright Virtual University of Pakistan

28-Review III

VU

An example of the various gamuts is shown in the figure below. The CIE diagrams are the traditional way of displaying perceived color space, which, we should note, is very different from the linear color space used by today's graphics hardware. The colored area is the gamut of the human eye. The gamut of printers and monitors are subsets of this gamut.

Figure 1: The 1931 CIE diagram shows the gamut of the eye and the lesser gamut of output devices. 28.12 Multiplying Color Values
First we need to be aware of how to treat colors. The calculation of the color of a particular pixel depends, for example, on the surface's material properties that we've programmed in, the color of the ambient light (lighting model), the color of any light shining on the surface (perhaps of the angle of the light to the surface), the angle of the surface to the viewpoint, the color of any fog or other scattering material that's between the surface and the viewpoint, etc. No matter how you are calculating the color of the pixel, it all comes down to color calculations, at least on current hardware, on rgb or rgba vectors where the individual color elements are limited to the [0,1] range. Operations on colors are done piecewisethat is, even though we represent colors as rgb vectors, they aren't really vectors in the mathematical sense. Vector multiplication is symbol to different from the operation we perform to multiply color. We'll use the indicate such piecewise multiplication. Colors are multiplied to describe the interaction between a surface and a light source. The colors of each are multiplied together to estimate the reflected light colorthis is the color of the light that this particular light reflects off this surface. The problem with the standard rgb model is just that we're simulating the entire visible spectrum by three colors with a limited range. 272 Copyright Virtual University of Pakistan

28-Review III

VU

Let's start with a simple example of using reflected colors. Later on we will discuss on lighting, we'll discover how to calculate the intensity of a light source, but for now, just assume that we've calculated the intensity of a light, and it's a value called id. This intensity of our light is represented by, say, a nice lime green color. Thus

Let's say we shine this light on a nice magenta surface given by cs.

So, to calculate the color contribution of this surface from this particular light, we perform a piecewise multiplication of the color values.

Note: Piecewise multiplication is denoted by that is element-by-element multiplication. Used in color operations, where the vector just represents a convenient notation for an array of scalars that are operated on simultaneously but independently. This gives us the dark plum color shown in figure below. We should note that since the surface has no green component, that no matter what value we used for the light color, there would never be any green component from the resulting calculation. Thus a pure green light would provide no contribution to the intensity of a surface if that surface contained a zero value for its green intensity. Thus it's possible to illuminate a surface with a bright light and get little or no illumination from that light. We should also note that using anything other than a full-bright white light [1,1,1] will involve multiplication of values less than one, which means that using a single light source will only illuminate a surface to a maximum intensity of its color value, never more. This same problem also happens when a texture is modulated by a surface color. The color of the surface will be multiplied by the colors in the texture. If the surface color is anything other than full white, the texture will become darker. Multiple texture passes can make a surface very dark very quickly.

Figure 2: Multiplying (modulating) color values results in a color equal to or less than (darker) the original two.
273 Copyright Virtual University of Pakistan

28-Review III

VU

Given that using a colored light in a scene makes the scene darker, how do we make the scene brighter? There are a few ways of doing this. Given that color multiplication will never result in a brighter color, it's offset a bit since we end up summing all the light contributions together, which, as we'll see in the next section, brings with it its own problems. But if we are just interested in increasing the brightness on one particular light or texture, one way is to use the API (Library routines e.g. OpenGL or DirectX) to artificially brighten the sourcethis is typically done with texture preprocessing. Or, we can artificially brighten the source, be it a light or a texture, by adjusting the values after we modulate them.

28.13 Dealing with Saturated Colors

On the other hand, what if we have too much contribution to a color? While the colors of lights are modulated by the color of the surface, each light source that illuminates the surface is added to the final color. All these colors are summed up to calculate the final color. Let's look at such a problem. We'll start with summing the reflected colors off a surface from two lights. The first light is an orange color and has rgb values [1.0,0.49,0.0], and the second light is a nice light green with rgb values [0.0,1.0,0.49]. Summing these two colors yields [1.0, 1.49, 0.49], which we can't display because of the values larger than one figure below shows.

Figure 3: Adding colors can result in colors that are outside the displayable range.
So, what can be done when color values exceed the range that the hardware can display? It turns out that there are three common approaches [HALL 1990]. Clamping the color values is implemented in hardware, so for shaders (technology used in today computer graphics for lighting and shading), it's the default, and it just means that we clamp any values outside the [0,1] range. Unfortunately, this results in a shift in the color. The second most common approach is to scale the colors by the largest component. This maintains the color but reduces the overall intensity of the color. The third is to try to maintain the intensity of the color by shifting (or clipping) the color toward pure bright white by reducing the colors that are too bright while increasing the other colors and maintaining the overall intensity. Since we can't see what the actual color for (figure above) is, let's see what color each of these methods yields (figure below).

274 Copyright Virtual University of Pakistan

28-Review III

VU

Figure 4: The results of three strategies for dealing with the same oversaturated color.
As we can see, we get three very different results. In terms of perceived color, the scaled is probably the closest though it's darker than the actual color values. If we weren't interested in the color but more in terms of saturation, then the clipped color is closer. Finally, the clamped value is what we get by default, and as you can see, the green component is biased down so that we lose a good sense of the "greenness" of the color we were trying to create.

275 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

Lecture No.29

Mathematics of Lighting and Shading Part III

Traditional 3D Hardware-Accelerated Lighting Models We will now take a look at the traditional method of calculating lighting in hardwarea method that we'll find is sufficient for most of our needs. The traditional approach in realtime computer graphics has been to calculate lighting at a vertex as a sum of the ambient, diffuse, and specular light. In the simplest form (used by OpenGL and Direct3D), the function is simply the sum of these lighting components (clamped to a maximum color value). Thus we have an ambient term and then a sum of all the light from the light sources.

i total

= k a ia +

(k d id + k s i s )

Where itotal , is the intensity of light (as an rgb value) from the sum of the intensity of the global ambient value and the diffuse and specular components of the light from the light sources. This is called a local lighting model since the only light on a vertex is from a light source, not from other objects. That is, lights are lights, not objects. Objects that are brightly lit don't illuminate or shadow any other objects. Weve included the reflection coefficients for each term, k for completeness since we'll frequently see the lighting equation. The reflection coefficients are in the [0, 1] range and are specified as part of the material property. However, they are strictly empirical and since they simply adjust the overall intensity of the material color, the material color values are usually adjusted so the color intensity varies rather than using a reflection coefficient, so we'll ignore them in our actual color calculations. This is a very simple lighting equation and gives fairly good results. However, it does fail to take into account any gross roughness or anything other than perfect isotropic reflection. That is, the surface is treated as being perfectly smooth and equally reflective in all directions. Thus this equation is really only good at modeling the illumination of objects that don't have any "interesting" surface properties. By this we mean anything other than a smooth surface (like fur or sand) or a surface that doesn't really reflect light uniformly in all directions (like brushed metal, hair, or skin). However, with liberal use of texture maps to add detail, this model has served pretty well and can still be used for a majority of the lighting processing to create a realistic environment in real time. Let's take a look at the individual parts of the traditional lighting pipeline.

Ambient Light
Ambient light is the light that comes from all directionsthus all surfaces are illuminated equally regardless of orientation. However, this is a big hack in traditional lighting calculations since "real" ambient light really comes from the light reflected from the "environment." This would take a long time to calculate and would require ray tracing or the use of radiosity methods, so traditionally, we just say that there's x amount of global ambient light and leave it at that. This makes ambient light a little different from the other lighting components since it doesn't depend on a light source. However, we typically do want ambient light in our scene because having a certain amount of ambient light makes the scene look natural. One large problem with the simplified lighting model is that there is no illumination of an object with reflected lightthe calculations required are enormous for a scene of any complexity (every object can potentially reflect some light and provide some illumination for every other object in a scene) and are too time consuming to be considered for real-time graphics. So, like most things in computer 276 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

graphics, we take a look at the real world, decide it's too complicated, and fudge up something that kind a works. Thus the ambient light term is the "fudge factor" that accounts for our simple lighting model's lack of an inter-object reflectance term. The ambient light equation is given by

ia = ma sa
Where ia is the ambient light intensity, ma is the ambient material color, and sa is the light source ambient color. Typically, the ambient light is some amount of white (i.e., equal rgb values) light, but we can achieve some nice effects using colored ambient light. Though it's very useful in a scene, ambient light doesn't help differentiate objects in a scene since objects rendered with the same value of ambient tend to blend since the resulting color is the same. Figure 1 shows a scene with just ambient illumination. We can see that it's difficult to make out details or depth information with just ambient light.

Figure 1: Ambient light provides illumination, but no surface details.

Ambient lighting is our friend. With it we make our scene seem more realistic than it is. A world without ambient light is one filled with sharp edges, of bright objects surrounded by sharp, dark, harsh shadows. A world with too much ambient light looks washed out and dull. Since the number of actual light sources supported by hardware FFP is limited (typically to eight simultaneous), we'll be better off to apply the lights to add detail to the area that our user is focused on and let ambient light fill in the rest. Before we point out that talking about the hardware limitation of the number of lights has no meaning on shaders, where we do the lighting calculations, we'll point out that eight lights were typically the maximum that the hardware engineers created for their hardware. It was a performance consideration. There's nothing stopping us (except buffer size) from writing a shader that calculates the effects from a hundred simultaneous lights. But we think that we'll find that it runs much too slowly to be used to render our entire scene. But the nice thing about shaders is we can.

Diffuse Light
Diffuse light is the light that is absorbed by a surface and is reflected in all directions. In the traditional model, this is ideal diffuse reflectiongood for rough surfaces where the reflected intensity is constant across the surface and is independent of viewpoint but 277 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

depends only upon the direction of the light source to the surface. This means that regardless of the direction from which we view an object with a stationary diffuse light source on it, the brightness of any point on the surface will remain the same. Thus, unlike ambient light, the intensity of diffuse light is directional and is a function of the angle of the incoming light and the surface. This type of shading is called Lambertian shading after Lambert's cosine law, which states that the intensity of the light reflected from an ideal diffuse surface is proportional to the cosine of the direction of the light to the vertex normal. Since we're dealing with vertices here and not surfaces, each vertex has a normal associated with it. We might hear talk of per-vertex normals vs. per-polygon normals. The difference being that per polygon has one normal shared for all vertices in a polygon, whereas per vertex has a normal for each vertex. OpenGL has the ability to specify perpolygon normals, and Direct3D does not. Since vertex shaders can't share information between vertices (unless we explicitly copy the data our self). We'll focus on per-vertex lighting. Figure 2 shows the intensity of reflected light as a function of the angle between the vertex normal and the light direction.

Figure 2: Diffuse light decreases as the angle between the light vector and the surface normal increases.
The equation for calculating diffuse lighting is Which is similar to the ambient light equation, except that the diffuse light term is now multiplied by the dot product of the unit normal of the vertex and the unit direction vector to the light from the vertex (not the direction from the light). Note that the md value is a color vector, so there are rgb or rgba values that will get modulated. , where theta is the angle between vectors, when the angle Since between them is zero, cos(theta ) is 1 and the diffuse light is at its maximum. When the angle is 90, cos (theta) is zero and the diffuse light is zero. One calculation advantage is that when the cos(theta ) value is negative, this means that the light isn't illuminating the vertex at all. However, since we (probably!) don't want the light illuminating sides that it physically can't shine on, we want to clamp the contribution of the diffuse light to contribute only when cos(theta ) is positive. Thus the equation in practice looks more like

Where we've clamped the diffuse value to only positive values, Figure 3 was rendered with just diffuse lighting. Notice how we can tell a lot more detail about the objects and pick up distance cues from the shading.

278 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

Figure 3: Diffuse shading brings out some surface details.

The problem with just diffuse lighting is that it's independent of the viewer's direction. That is, it's strictly a function of the surface normal and the light direction. Thus as we change the viewing angle to a vertex, the vertex's diffuse light value never changes. You have to rotate the object (change the normal direction) or move the light (change the light direction) to get a change in the diffuse lighting of the object. However, when we combine the ambient and diffuse, as in Figure 4, we can see that the two types of light give a much more realistic representation than either does alone. This combination of ambient and diffuse is used for a surprisingly large number of items in rendered scenes since when combined with texture maps to give detail to a surface we get a very convincing shading effect.

Figure 4: When diffuse and ambient terms are combined, you get more detail and a more natural-looking scene. The final color is the combination of the ambient and diffuse colors. Specular Light
Ambient light is the light that comes from the environment (i.e., it's directionless); diffuse light is the light from a light source that is reflected by a surface evenly in all directions (i.e., it's independent of the viewer's position). Specular light is the light from a light source that is reflected by a surface and is reflected in such a manner that it's both a function of the light's vector and the viewer's direction. While ambient light gives the 279 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

object an illuminated matte surface, specular light is what gives the highlights to an object. These highlights are greatest when the viewer is looking directly along the reflection angle from the surface. This is illustrated in Figure 5.

Figure 5: Specular light's intensity follows the reflection vector.

Most discussions of lighting (including this one) start with Phong's lighting equation (which is not the same as Phong's shading equation). In order to start discussing specular lighting, let's look at a diagram of the various vectors that are used in a lighting equation. We have a light source, some point the light is shining on, and a viewpoint. The light direction (from the point to the light) is vector l, the reflection vector of the light vector (as if the surface were a mirror) is r, the direction to the viewpoint from the point is vector v. The point's normal is n.

Phong's Specular Light Equation

Warnock [WARNOCK 1969] and Romney [ROMNEY 1969] were the first to try to simulate highlights using a cos n () term. But it wasn't until Phong Bui-Tong [BUI 1998] reformulated this into a more general model that formalized the power value as a measure of surface roughness that we approach the terms used today for specular highlights. Phong's equation for specular lighting is

Figure 6: The relationship between the normal n, the light vector v, the view direction v, and the reflection vector r. 280 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

It basically says that the more the view direction, v, is aligned with the reflection direction, r, the brighter the specular light will be. The big difference is the introduction of the ms term, which is a power term that attempts to approximate the distribution of specular light reflection. The ms term is typically called the "shininess" value. The larger the ms value, the "tighter" (but not brighter) the specular highlights will be. This can be for values of m ranging from 1 to seen in the Figure 7, which shows values of 128. As we can see, the specular highlights get narrower for higher values, but they don't get any brighter.

Figure 7: Phong's specular term for various values of the "shininess" term. Note that the values never get above 1.
Now, as we can see, this requires some calculations since we can't know r before hand since it's the v vector reflected around the point's normal. To calculate r we can use the following equation:

If l and n are normalized, then the resulting r is normalized and the equation can be simplified.

And just as we did for diffuse lighting, if the dot product is negative, then the term is ignored.

281 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

Figure 8 shows the scene with just specular lighting. As we can see, we get an impression of a very shiny surface.

Figure 8: A specular term just shows the highlights.

When we add the ambient, diffuse, and specular terms together, we get Figure 8A. The three terms all act in concert to give us a fairly good imitation of a nice smooth surface that can have a varying degree of shininess to it. We may have noticed that computing the reflection vector took a fair amount of effort. In the early days of computer graphics, there was a concerted effort to reduce anything that took a lot of computation, and the reflection vector of Phong's equation was one such item.

Figure 8A: A combination of ambient, diffuse, and specular illumination. Blinn's Simplification: OpenGL and DirectX Lighting
Now it's computationally expensive to calculate specular lighting using Phong's equation since computing the reflection vector is expensive. Blinn [BLINN 1977] suggested, instead of using the reflection and view vectors, that we create a "half" vector that lies between the light and view vectors. This is shown as the h vector in Figure 9. Just as Phong's equation maximizes when the reflection vector is coincident with the view vector (thus the viewer is looking directly along the reflection vector), so does Blinn's. When the half vector is coincident with the normal vector, then the angle between the view vector 282 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

and the normal vector is the same as between the light vector and the normal vector. Blinn's version of Phong's equation is:

Figure 9: The half-angle vector is an averaging of the light and view vectors. where the half vector is defined as

The advantage is that no reflection vector is needed; instead, we can use values that are readily available, namely, the view and light vectors. Note that both OpenGL and the DirectX FFP use Blinn's equation for specular light. Besides a speed advantage, there are some other effects to note between Phong's specular equation and Blinn's. If we multiply Blinn's exponent by 4, we approximate the results of Phong's equation. Thus if there's an upper limit on the value of the exponent, Phong's equation can produce sharper highlights. For l v angles greater than 45 (i.e., when the light is behind an object and we're looking at an edge), the highlights are longer along the edge direction for Phong's equation. Blinn's equation produces results closer to those seen in nature. For an in-depth discussion of the differences between the two equations, there's an excellent discussion in [FISHER 1994]. Figure 10 shows the difference between Phong lighting and BlinnPhong lighting.

283 Copyright Virtual University of Pakistan

29-Mathematics of Lighting and Shading Part III

VU

Figure 10: Blinn-Phong specular on the left, Phong specular on the right.

Note: Some of the material, for the preparation of this lecture, is taken from a book Real time shader Programming by Ron fosner

284 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

Lecture No.30

Mathematics of Lighting and Shading Part IV

The Lighting Equation So now that we've computed the various light contributions to our final color value, we can add them up to get the final color value. Note that the final color values will have to be made to fit in the [0,1] range for the final rgb values.

Our final scene with ambient, diffuse, and (Blinn's) specular light contributions (with one white light above and to the left of the viewer) looks like Figure 1.

Figure 1: A combination of ambient, diffuse, and specular illumination.

It may be surprising to discover that there's more than one way to calculate the shading of an object, but that's because the model is empirical, and there's no correct way, just different ways that all have tradeoffs. Until now though, the only lighting equation we've been able to use has been the one we just formulated. Most of the interesting work in computer graphics is tweaking that equation, or in some cases, throwing it out altogether and coming up with something new. The next sections will discuss some refinements and alternative ways of calculating the various coefficients of the lighting equation.

Light Attenuation
Light in the real world loses its intensity as the inverse square of the distance from the light source to the surface being illuminated. However, when put into practice, this seemed to drop off the light intensity in too abrupt a manner and then not to vary too much after the light was far away. An empirical model was developed that seems to give satisfactory results. This is the attenuation model that's used in OpenGL and DirectX. The fatten factor is the attenuation factor. The distance d between the light and the vertex is always positive. The attenuation factor is calculated by the following equation:

285 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

Where the kc, k1, and kq parameters are the constant, linear, and quadratic attenuation constants respectively, to get the "real" attenuation factor, we can set kq to one and the others to zero. The attenuation factor is multiplied by the light diffuse and specular values. Typically, each light will have a set of these parameters for itself. The lighting equation with the attenuation factor looks like this.

Figure 2 shows a sample of what attenuation looks like. This image is the same as the one shown in Figure 1, but with light attenuation added.

Figure 2: A scene with light attenuation. The white sphere is the light position. Schlick's simplification for the Specular Exponential Term Real-time graphics programmers are always looking for simplifications. Weve probably gathered that there's no such thing as the "correct" lighting equation, just a series of hacks to make things look right with as little computational effort as possible. Schlick [SCHLICK 1994] suggested a replacement for the exponential term since that's a fairly expensive operation. If we define part of our specular light term as follows:
where S is either the Phong or Blinn-Phong flavor of the specular lighting equation, then Schlick's simplification is to replace the preceding part of the specular equation with

Which eliminates the need for an exponential term, At first glance, a plot of Schlick's function looks very similar to the exponential equation (Figure 3).

Figure 3: Schlick's term for specular looks very much like the more expensive Phong term. 286 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

If we plot both equations in the same graph (Figure 4), we can see some differences and evaluate just how well Schlick's simplification works. The blue values are Schlick's, and the red are the exponential plot. As the view and light angles get closer (i.e., get closer to zero on the x axis), we can see that the values of the curves are quite close. (For a value of zero, they overlap.) As the angles approach a grazing angle, we can see that the approximation gets worse. This would mean that when there is little influence from a specular light, Schlick's equation would be slightly less sharp for the highlight.

Figure 4: Schlick's vs. Phong's specular terms.

We might notice the green line in Figure 4. Unlike the limit of a value of 128 for the exponential imposed in both OpenGL and DirectX FFP, we can easily make our values in the approximation any value we want. The green line is a value of 1024 in Schlick's equation. We may be thinking that we can make a very sharp specular highlight using Schlick's approximation with very large valuessharper than is possible using the exponential term. Unfortunately, we can't since we really need impractically large values (say, around 100 million) to boost it significantly over the exponential value for 128. But that's just the kind of thinking that's going to get our creative juices flowing when writing our own shaders! If the traditional way doesn't work, figure out something that will.

OrenNayar Diffuse Reflection Though there's been a lot of research on specular reflection models, there's been less research on diffuse reflection models. One of the problems of the standard Lambertian model is that it considers the surface as a smooth diffuse surface. Surfaces that are really rough, like sandpaper, exhibit much more of a backscattering effect, particularly when the light source and the view direction are in the same direction. The classic example of this is a full moon. If we look at the picture of the moon shown in Figure 5, it's pretty obvious that this doesn't follow the Lambertian distributionif it did, the edges of the moon would be in near darkness. In fact, the edges look as bright as the center of the moon. This is because the moon's surface is roughthe surface is made of a jumble of dust and rock with diffuse reflecting surfaces at all anglesthus the quantity of reflecting surfaces is uniform no matter the orientation of the surface; hence no matter the orientation of the surface to the viewer, the amount of light reflecting off the surface is nearly the same.
287 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

Figure 5: The full moon is a good example of something that doesn't show Lambertian diffuse shading.
The effect we're looking at is called backscattering. Backscattering is when a rough surface bounces around a light ray and then reflects the ray in the direction the light originally came from. Note that there is a similar but different effect called retro reflection. Retro reflection is the effect of reflecting light toward the direction from which it came, no matter the orientation of the surface. This is the same effect that we see on bicycle reflectors. However, this is due to the design of the surface features (made up of vshaped or spherical reflectors) rather than a scattering effect. In a similar manner, when the light direction is closer to the view direction, we get the effect of forward scattering. Forward scattering is just backscattering from a different direction. In this case, instead of near uniform illumination though, we get near uniform loss of diffuse lighting. We can get the same effects here on Earth. Figures 6 and 7, shows the same surfaces demonstrating backscattering and forward scattering. Both the dirt field in Figure 6 and the soybean field in Figure 7 can be considered rough diffuse reflecting surfaces.

Figure 6: The same dirt field showing wildly differing reflection properties.
288 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

Figure 7: A soybean field showing differing reflection properties.

Notice how the backscattering image shows a near uniform diffuse illumination, whereas the forward scattering image shows a uniform dull diffuse illumination. Also note that we can see specular highlights and more color variation because of the shadows due to the rough surface whereas the backscattered image washes out the detail. In an effort to better model rough surfaces, Oren and Nayar [OREN 1992] came up with a generalized version of a Lambertian diffuse shading model that tries to account for the roughness of the surface. They applied the TorranceSparrow model for rough surfaces with isotropic roughness and provided parameters to account for the various surface structures found in the TorranceSparrow model. By comparing their model with actual data, they simplified their model to the terms that had the most significant impact. The Oren Nayar diffuse shading model looks like this. Where

Now this may look daunting, but it can be simplified to something we can appreciate if we replace the original notation with the notation we've already been using. / is a surface reflectivity property, which we can replace with our surface diffuse color. E0 is a light input energy term, which we can replace with our light diffuse color. And the i term is just our familiar angle between the vertex normal and the light direction. Making these exchanges gives us

Which looks a lot more like the equations we've used, there are still some parameters to explain. 289 Copyright Virtual University of Pakistan

30-Mathematics of Lighting and Shading Part IV

VU

is the surface roughness parameter. It's the standard deviation in radians of the angle of distribution of the microfacets in the surface roughness model. The larger the value, the rougher the surface. r is the angle between the vertex normal and the view direction. r - i is the circular angle (about the vertex normal) between the light vector and the view vector. is max(i, r). is min (i, r). Note that if the roughness value is zero, the model is the same as the Lambertian diffuse model. Oren and Nayar also note that we can replace the value 0.33 in coefficient A with 0.57 to better account for surface inter-reflection.

290 Copyright Virtual University of Pakistan

31-Mathematics of Lighting and Shading Part V

VU

Lecture No.31

Mathematics of Lighting and Shading Part V

Physically Based Illumination

In order to get a more realistic representation of lighting, we need to move away from the simplistic models that are found hard coded in most graphics pipelines and move to something that is based more in a physical representation of light as a wave with properties of its own that can interact with its environment. To do this, we'll need to understand how light passes through a medium and how hitting the boundary layer at the intersection of two media can affect light's properties. In Figure 1, there's an incident light hitting a surface. At the boundary of the two media (in this case, air and glass), there are two resulting rays of light. The reflected ray is the one that we've already discussed to some extent, and the other ray is the refracted or transmitted ray.

Figure 1: Light being reflected and refracted through a boundary.

In addition to examining the interaction of light with the surface boundary, we need a better description of real surface geometries. Until now, we've been treating our surfaces as perfectly smooth and uniform. Unfortunately, this prevents us from getting some interesting effects. We'll go over trying to model a real surface later, but first let's look at the physics of light interacting at a material boundary.

Reflection
Reflection of a light wave is the change in direction of the light ray when it bounces off the boundary between two media. The reflected light wave turns out to be a simple case since light is reflected at the same angle as the incident wave (when the surface is smooth and uniform, as we'll assume for now). Thus for a light wave reflecting off a perfectly smooth surface Until now, we've treated all of our specular lighting calculations as essentially reflection off a perfect surface, a surface that doesn't interact with the light in any manner other than reflecting light in proportion to the color of the surface itself. Using a lighting model based upon the BlinnPhong model means that we'll always get a uniform specular highlight based upon the color of the reflecting light and material, which means that all reflections based on this model, will be reminiscent of plastic. In order to get a more 291 Copyright Virtual University of Pakistan

31-Mathematics of Lighting and Shading Part V

VU

interesting and realistic lighting model, we need to add in some nonlinear elements to our calculations. First, let's examine what occurs when light is reflected off a surface. For a perfect reflecting surface, the angle of the incoming light (the angle of incidence) is equal to that of the reflected light. Phong's equation just blurs out the highlight a bit in a symmetrical fashion. Until we start dealing with non uniform smooth surfaces in a manner a bit more realistic than Phong's.

Refraction
Refraction happens when a light wave goes from one medium into another. Because of the difference in the speed of light of the media, light bends when it crosses the boundary. Snell's law gives the change in angles. Where the n's are the material's index of refraction, Snell's law states that when light refracts through a surface, the refracted angle is shifted by a function of the ratio of the two material's indices of refraction. The index of refraction of vacuum is 1, and all other material's indices of refraction are greater than 1. What this means is that in order to realistically model refraction, we need to know the indices of refraction of the two materials that the light is traveling through. Let's look at an example (Figure 2) to see what this really means. Let's take a simple case of a ray of light traveling through the air (n air = 1) and intersecting a glass surface (n Glass = 1.5). If the light ray hits the glass surface at 45, at what angle does the refracted ray leave the interface?

Figure 2: The refracted ray's angle is less than the incoming ray's.
The angle of incidence is the angle between the incoming vector and the surface. Rearranging Snell's law, we can solve for the refracted angle.

Inserting the values give us 292 Copyright Virtual University of Pakistan

31-Mathematics of Lighting and Shading Part V

VU

This is a fairly significant change in the angle! If we change things around so that we are following a light ray emerging from water into the air, we can run into another phenomenon. Since the index of refraction is just a measure of the change in speed that light travels in a material, we can observe from Snell's law (and the fact that the index of refraction in a vacuum is 1) that light bends toward the normal when it slows down (i.e., when the material it's intersecting with has a higher index of refraction). Consequently, when we intersect a medium that has a lower index of refraction (e.g., going from glass to air), then the angle will increase. Ah, we must be thinking, we're approaching a singularity here since we can then easily generate numbers that we can't take the inverse sine of! If we use Snell's law for light going from water to air, and plug in 90 for the refracted angle, we get 41.8 for the incident angle. This is called the critical angle at which we observe the phenomenon of total internal reflection. At any angle greater than this, light will not pass though a boundary but will be reflected internally. One place that we get interesting visual properties is in the diamondair interface. The refractive index of a diamond is fairly high, 2.24, which means that it's got a very low critical angle, just 24.4. This means that a good portion of the light entering a diamond will bounce around the inside of the diamond hitting a number of airdiamond boundaries, and as long as the angle is 24.4 or greater, it will keep reflecting internally. This is why diamonds are cut to be relatively flatish on the top but with many faceted sides, so that light entering in one spot will bounce around and exit at another, giving rise to the sparkle normally associated with diamonds. Another place where a small change in the indices of refraction occurs is on a road heated by the sun when viewed from far away (hence a glancing incident angle). The hot air at the road's surface has a slightly smaller index of refraction than the denser, cooler air above it. This is why we get the effect of a road looking as though it were covered with water and reflecting the image above itthe light waves are actually reflected off the warm aircold air interface.

Figure 3: The critical angle.

What makes this really challenging to model is that the index of refraction for most materials is a function of the wavelength of the light. This means that not only is there a 293 Copyright Virtual University of Pakistan

31-Mathematics of Lighting and Shading Part V

VU

shift in the angle of refraction, but that the shift is different for differing wavelengths of light. Figure 4 and 5 show the index of refraction for fused quartz and sapphire plotted against the wavelength. We can see the general trend that shorter wavelength light (bluish) tends to bend more than the longer (reddish) wavelengths.

Figure 4: Index of refraction as a function of wavelength for quartz.

Figure 5: Index of refraction as a function of wavelength for sapphire.

This is the phenomenon that's responsible for the spectrum that can be seen when white light is passed through a prism (Figure 6). It's refraction that will break apart a light source into its component colors, not reflection.

294 Copyright Virtual University of Pakistan

31-Mathematics of Lighting and Shading Part V

VU

Figure 6: The wavelength dependence of the index of refraction in action.

This is one area where our simplistic model of light breaks down since we're not computing an entire spectrum of light waves, but we're limited to three primary colors. For reference, the rgb values can be assigned to a range of wavelengths as follows:

There's a lot more to color science than just determining wavelengths, but that's beyond our scope. While the spectrum spreading effect of refraction is interesting in itself, the rgb nature of computer color representation precludes performing this spreading directlywe can't break up a color value into multiple color values. However, with some work, we can compute the shade of the color for a particular angle of refraction and then use that as the material color to influence the refracted color.

Temperature Correction for Refractive Index

Refractive index is a function of temperature, mostly due to density changes in materials with changes in temperature. A simple correction can be applied in most circumstances to allow us to use a value given at one temperature at another. For example, suppose the index of refraction value we have is given at 25C: 25. To convert the index to another temperature, t, we can use the following equation: Where the actual temperature we want is t, and the 25 is the temperature (both in C) of the actual index we have, 25.

295 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

Lecture No.32

Introduction to OpenGL

As a software interface for graphics hardware, OpenGL renders multidimensional objects into a frame buffer. OpenGL is industry-standard graphics software with which programmers can create high-quality still and animated three-dimensional color images.

Where Applicable:
OpenGL is built for compatibility across hardware and operating systems. This architecture makes it easy to port OpenGL programs from one system to another. While each operating system has unique requirements, the OpenGL code in many programs can be used as is. Developer Audience: Designed for use by C/C++ programmers Run-time Requirements: OpenGL can run on Linux and all versions of 32 bit Microsoft Windows.

Most Widely Adopted Graphics Standard

OpenGL is the premier environment for developing portable, interactive 2D and 3D graphics applications. Since its introduction in 1992, OpenGL has become the industry's most widely used and supported 2D and 3D graphics application programming interface (API), bringing thousands of applications to a wide variety of computer platforms. OpenGL fosters innovation and speeds application development by incorporating a broad set of rendering, texture mapping, special effects, and other powerful visualization functions. Developers can leverage the power of OpenGL across all popular desktop and workstation platforms, ensuring wide application deployment.

High Visual Quality and Performance

Any visual computing application requiring maximum performance-from 3D animation to CAD to visual simulation-can exploit high-quality, high-performance OpenGL capabilities. These capabilities allow developers in diverse markets such as broadcasting, CAD/CAM/CAE, entertainment, medical imaging, and virtual reality to produce and display incredibly compelling 2D and 3D graphics.

Developer-Driven Advantages

Industry standard
An independent consortium, the OpenGL Architecture Review Board, guides the OpenGL specification. With broad industry support, OpenGL is the only truly open, vendor-neutral, multiplatform graphics standard.

Stable
OpenGL implementations have been available for more than seven years on a wide variety of platforms. Additions to the specification are well controlled, and proposed updates are announced in time for developers to adopt changes. 296 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

Backward compatibility requirements ensure that existing applications do not become obsolete.

Reliable and portable

All OpenGL applications produce consistent visual display results on any OpenGL API-compliant hardware, regardless of operating system or windowing system.

Evolving
Because of its thorough and forward-looking design, OpenGL allows new hardware innovations to be accessible through the API via the OpenGL extension mechanism. In this way, innovations appear in the API in a timely fashion, letting application developers and hardware vendors incorporate new features into their normal product release cycles.

Scalable
OpenGL API-based applications can run on systems ranging from consumer electronics to PCs, workstations, and supercomputers. As a result, applications can scale to any class of machine that the developer chooses to target.

Easy to use
OpenGL is well structured with an intuitive design and logical commands. Efficient OpenGL routines typically result in applications with fewer lines of code than those that make up programs generated using other graphics libraries or packages. In addition, OpenGL drivers encapsulate information about the underlying hardware, freeing the application developer from having to design for specific hardware features.

Well-documented
Numerous books have been published about OpenGL, and a great deal of sample code is readily available, making information about OpenGL inexpensive and easy to obtain.

Simplifies Software Development, Speeds Time-to-Market

OpenGL routines simplify the development of graphics softwarefrom rendering a simple geometric point, line, or filled polygon to the creation of the most complex lighted and texture-mapped NURBS curved surface. OpenGL gives software developers access to geometric and image primitives, display lists, modeling transformations, lighting and texturing, anti-aliasing, blending, and many other features. Every conforming OpenGL implementation includes the full complement of OpenGL functions. The well-specified OpenGL standard has language bindings for C, C++, Fortran, Ada, and Java. All licensed OpenGL implementations come from a single specification and language binding document and are required to pass a set of conformance tests. Applications utilizing OpenGL functions are easily portable across a wide array of platforms for maximized programmer productivity and shorter time-tomarket. 297 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

All elements of the OpenGL stateeven the contents of the texture memory and the frame buffercan be obtained by an OpenGL application. OpenGL also supports visualization applications with 2D images treated as types of primitives that can be manipulated just like 3D geometric objects. As shown in the OpenGL visualization programming pipeline diagram above, images and vertices defining geometric primitives are passed through the OpenGL pipeline to the frame buffer.

Available Everywhere
Supported on all UNIX workstations, and shipped standard with every Windows 95/98/2000/NT and MacOS PC, no other graphics API operates on a wider range of hardware platforms and software environments. OpenGL runs on every major operating system including Mac OS, OS/2, UNIX, Windows 95/98, Windows 2000, Windows NT, Linux, OPENStep, and BeOS; it also works with every major windowing system, including Win32, MacOS, Presentation Manager, and X-Window System. OpenGL is callable from Ada, C, C++, Fortran, Python, Perl and Java and offers complete independence from network protocols and topologies.

Architected for Flexibility and Differentiation!

Although the OpenGL specification defines a particular graphics processing pipeline, platform vendors have the freedom to tailor a particular OpenGL implementation to meet unique system cost and performance objectives. Individual calls can be executed on dedicated hardware, run as software routines on the standard system CPU, or implemented as a combination of both dedicated hardware and software routines. This implementation flexibility means that OpenGL hardware acceleration can range from simple rendering to full geometry and is widely available on everything from low-cost PCs to high-end workstations and supercomputers. Application developers are assured consistent display results regardless of the platform implementation of the OpenGL environment. Using the OpenGL extension mechanism, hardware developers can differentiate their products by developing extensions that allow software developers to access additional performance and technological innovations. Main purpose of OpenGL As a software interface for graphics hardware, the main purpose of OpenGL is to render two- and three-dimensional objects into a frame buffer. These objects are described as sequences of vertices (that define geometric objects) or pixels (that define images). OpenGL performs several processes on this data to convert it to pixels to form the final desired image in the frame buffer. The following topics present a global view of how OpenGL works:

Primitives and Commands discusses points, line segments, and polygons as the basic units of drawing; and the processing of commands. OpenGL Graphic Control describes which graphic operations OpenGL controls and which it does not control. Execution Model discusses the client/server model for interpreting OpenGL commands.
298 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

Basic OpenGL Operation gives a high-level description of how OpenGL processes data to produce a corresponding image in the frame buffer.

Primitives and Commands OpenGL draws primitive points, line segments, or polygons subject to several selectable modes. You can control modes independently of one another. That is, setting one mode doesn't affect whether other modes are set (although many modes may interact to determine what eventually ends up in the frame buffer). To specify primitives, set modes, and perform other OpenGL operations, you issue commands in the form of function calls. Primitives are defined by a group of one or more vertices. A vertex defines a point, an endpoint of a line, or a corner of a polygon where two edges meet. Data (consisting of vertex coordinates, colors, normals, texture coordinates, and edge flags) is associated with a vertex, and each vertex and its associated data are processed independently, in order, and in the same way. The only exceptions to this rule are cases in which the group of vertices must be clipped so that a particular primitive fits within a specified region. In this case, vertex data may be modified and new vertices created. The type of clipping depends on which primitive the group of vertices represents. Commands are always processed in the order in which they are received, although there may be an indeterminate delay before a command takes effect. This means that each primitive is drawn completely before any subsequent command takes effect. It also means that state-querying commands return data that is consistent with complete execution of all previously issued OpenGL commands.

OpenGL Graphic Control

OpenGL provides you with fairly direct control over the fundamental operations of twoand three-dimensional graphics. This includes specification of such parameters as transformation matrices, lighting equation coefficients, antialiasing methods, and pixelupdate operators. However, it doesn't provide you with a means for describing or modeling complex geometric objects. Thus, the OpenGL commands you issue specify how a certain result should be produced (what procedure should be followed) rather than what exactly that result should look like. That is, OpenGL is fundamentally procedural rather than descriptive. To fully understand how to use OpenGL, it helps to know the order in which it carries out its operations.

Execution Model
The model for interpretation of OpenGL commands is client/server. Application code (the client) issues commands, which are interpreted and processed by OpenGL (the server). The server may or may not operate on the same computer as the client. In this sense, OpenGL is network-transparent. A server can maintain several OpenGL contexts, each of which is an encapsulated OpenGL state. A client can connect to any one of these contexts. The required network protocol can be implemented by augmenting an already existing protocol (such as that of the X Window System) or by using an independent protocol. No OpenGL commands are provided for obtaining user input. The window system that allocates frame buffer resources ultimately controls the effects of OpenGL commands on the frame buffer. The window system:

299 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

Determines which portions of the frame buffer OpenGL may access at any given time. Communicates to OpenGL how those portions are structured.

Therefore, there are no OpenGL commands to configure the frame buffer or initialize OpenGL. Frame buffer configuration is done outside of OpenGL in conjunction with the window system; OpenGL initialization takes place when the window system allocates a window for OpenGL rendering. Basic OpenGL Operation The following diagram illustrates how OpenGL processes data. As shown, commands enter from the left and proceed through a processing pipeline. Some commands specify geometric objects to be drawn, and others control how the objects are handled during various processing stages.

The processing stages in basic OpenGL operation are as follows:

Display list Rather than having all commands proceed immediately through the pipeline, you can choose to accumulate some of them in a display list for processing later. Evaluator The evaluator stage of processing provides an efficient way to approximate curve and surface geometry by evaluating polynomial commands of input values. Per-vertex operations and primitive assembly OpenGL processes geometric primitivespoints, line segments, and polygonsall of which are described by vertices. Vertices are transformed and lit, and primitives are clipped to the view port in preparation for rasterization. Rasterization The rasterization stage produces a series of frame-buffer addresses and associated values using a two-dimensional description of a point, line segment, or polygon. Each fragment so produced is fed into the last stage, perfragment operations. Per-fragment operations these are the final operations performed on the data before it's stored as pixels in the frame buffer.
Per-fragment operations include conditional updates to the frame buffer based on incoming and previously stored z values (for z buffering) and blending of 300 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

incoming pixel colors with stored colors, as well as masking and other logical operations on pixel values. Data can be input in the form of pixels rather than vertices. Data in the form of pixels, such as might describe an image for use in texture mapping, skips the first stage of processing described above and instead is processed as pixels, in the pixel operations stage. Following pixel operations, the pixel data is either:

Stored as texture memory, for use in the rasterization stage. Rasterized, with the resulting fragments merged into the frame buffer just as if they were generated from geometric data.

OpenGL Processing Pipeline Many OpenGL functions are used specifically for drawing objects such as points, lines, polygons, and bitmaps. Some functions control the way that some of this drawing occurs (such as those that enable antialiasing or texturing). Other functions are specifically concerned with frame buffer manipulation. The topics in this section describe how all of the OpenGL functions work together to create the OpenGL processing pipeline. This section also takes a closer look at the stages in which data is actually processed, and ties these stages to OpenGL functions. The following diagram details the OpenGL processing pipeline. For most of the pipeline, you can see three vertical arrows between the major stages. These arrows represent vertices and the two primary types of data that can be associated with vertices: color values and texture coordinates. Also note that vertices are assembled into primitives, then into fragments, and finally into pixels in the framebuffer.

The OpenGL Visualization Programming Pipeline

OpenGL operates on image data as well as geometric primitives. A detailed view on the next page.

301 Copyright Virtual University of Pakistan

32-Introduction to OpenGL

VU

302 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

Lecture No.33

OpenGL Programming - I

For writing a program, using OpenGL, for any of the operating systems we can use OpenGL Utility Library named, glut. glut is used to initialize OpenGL on any platform e.g. Microsoft Windows or Linux etc because it is platform independent. glut can create window, get keyboard input and run an event handler or message loop in our graphics application. All those functions that start with the prefix gl are the core OpenGL functions and those which start with glu or glut are the OpenGL utility library functions. Lets write a program that uses glut and then uses OpenGL function to create graphics. #include <GL/glut.h> int main() { glutCreateWindow( "first graphics window" ); } glutCreateWindow glutCreateWindow creates a top-level window. Usage int glutCreateWindow(char *name); name: ASCII character string for use as window name. Description: glutCreateWindow creates a top-level window. The name will be provided to the window system as the window's name. The intent is that the window system will label the window with the name. Implicitly, the current window is set to the newly created window. Each created window has a unique associated OpenGL context. State changes to a window's associated OpenGL context can be done immediately after the window is created. The display state of a window is initially for the window to be shown. But the window's display state is not actually acted upon until glutMainLoop is entered. This means until glutMainLoop is called, rendering to a created window is ineffective because the window cannot yet be displayed. The value returned is a unique small integer identifier for the window. The range of allocated identifiers starts at one. This window identifier can be used when calling glutSetWindow. X Implementation Notes The proper X Inter-Client Communication Conventions Manual (ICCCM) top-level properties are established. The WM_COMMAND property that lists the command line used to invoke the GLUT program is only established for the first window created. 303 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

This is the simple program that we have written so far. Now we will use more features of glut library. #include <GL/glut.h> int main() { //Set up the OpenGL rendering context. glutInitDisplayMode( GLUT_RGB | GLUT_DOUBLE | GLUT_DEPTH ); //Create a window and set its width and height. glutCreateWindow( "Deform" ); glutReshapeWindow( 640, 480 ); //The keyboard function gets called whenever we hit a key. glutKeyboardFunc( keyboard ); //The display function gets called whenever the window //requires an update or when we explicitly call glutDisplayFunc( display ); //The reshape function gets called whenever the window changes //shape. glutReshapeFunc( reshape ); // The idle function gets called when there are no window// events to process. glutIdleFunc( idle ); //Get the ball rolling! glutMainLoop(); } In the above program we have used glut functions. Let us discuss them in detail. GlutInitDisplayMode glutInitDisplayMode sets the initial display mode. Usage void glutInitDisplayMode(unsigned int mode); mode Display mode, normally the bitwise OR-ing of GLUT display mode bit masks. See values below: GLUT_RGBA Bit mask to select an RGBA mode window. This is the default if neither GLUT_RGBA nor GLUT_INDEX are specified. GLUT_RGB An alias for GLUT_RGBA. GLUT_INDEX 304 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

Bit mask to select a color index mode window. This overrides GLUT_RGBA if it is also specified. GLUT_SINGLE Bit mask to select a single buffered window. This is the default if neither GLUT_DOUBLE or GLUT_SINGLE are specified. GLUT_DOUBLE Bit mask to select a double buffered window. This overrides GLUT_SINGLE if it is also specified. GLUT_ACCUM Bit mask to select a window with an accumulation buffer. GLUT_ALPHA Bit mask to select a window with an alpha component to the color buffer(s). GLUT_DEPTH Bit mask to select a window with a depth buffer. GLUT_STENCIL Bit mask to select a window with a stencil buffer. GLUT_MULTISAMPLE Bit mask to select a window with multisampling support. If multisampling is not available, a non-multisampling window will automatically be chosen. Note: both the OpenGL client-side and server-side implementations must support the GLX_SAMPLE_SGIS extension for multisampling to be available. GLUT_STEREO Bit mask to select a stereo window. GLUT_LUMINANCE Bit mask to select a window with a ``luminance'' color model. This model provides the functionality of OpenGL's RGBA color model, but the green and blue components are not maintained in the frame buffer. Instead each pixel's red component is converted to an index between zero and glutGet(GLUT_WINDOW_COLORMAP_SIZE)-1 and looked up in a perwindow color map to determine the color of pixels within the window. The initial colormap of GLUT_LUMINANCE windows is initialized to be a linear gray ramp, but can be modified with GLUT's colormap routines. Description The initial display mode is used when creating top-level windows, subwindows, and overlays to determine the OpenGL display mode for the to-be-created window or overlay. Note that GLUT_RGBA selects the RGBA color model, but it does not request any bits of alpha (sometimes called an alpha buffer or destination alpha) be allocated. To request alpha, specify GLUT_ALPHA. The same applies to GLUT_LUMINANCE. GLUT_LUMINANCE Implementation Notes GLUT_LUMINANCE is not supported on most OpenGL platforms. glutReshapeWindow glutReshapeWindow requests a change to the size of the current window. Usage void glutReshapeWindow(int width, int height); width New width of window in pixels. 305 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

height New height of window in pixels. Description glutReshapeWindow requests a change in the size of the current window. The width and height parameters are size extents in pixels. The width and height must be positive values. The requests by glutReshapeWindow are not processed immediately. The request is executed after returning to the main event loop. This allows multiple glutReshapeWindow, glutPositionWindow, and glutFullScreen requests to the same window to be coalesced. In the case of top-level windows, a glutReshapeWindow call is considered only to be a request for sizing the window. The window system is free to apply its own policies to toplevel window sizing. The intent is that top-level windows should be reshaped according to glutReshapeWindow's parameters. Whether a reshape actually takes effect and, if so, the reshaped dimensions are reported to the program by a reshape callback. glutReshapeWindow disables the full screen status of a window if previously enabled. glutKeyboardFunc glutKeyboardFunc sets the keyboard callback for the current window. Usage void glutKeyboardFunc(void (*func)(unsigned char key, int x, int y)); func The new keyboard callback function. Description glutKeyboardFunc sets the keyboard callback for the current window. When a user types into the window, each key press generating an ASCII character will generate a keyboard callback. The key callback parameter is the generated ASCII character. The state of modifier keys such as Shift cannot be determined directly; their only effect will be on the returned ASCII data. The x and y callback parameters indicate the mouse location in window relative coordinates when the key was pressed. When a new window is created, no keyboard callback is initially registered, and ASCII key strokes in the window are ignored. Passing NULL to glutKeyboardFunc disables the generation of keyboard callbacks. During a keyboard callback, glutGetModifiers may be called to determine the state of modifier keys when the keystroke generating the callback occurred. We can also see glutSpecialFunc for a means to detect non-ASCII key strokes. glutDisplayFunc glutDisplayFunc sets the display callback for the current window. Usage void glutDisplayFunc(void (*func)(void)); func The new display callback function. 306 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

Description glutDisplayFunc sets the display callback for the current window. When GLUT determines that the normal plane for the window needs to be redisplayed, the display callback for the window is called. Before the callback, the current window is set to the window needing to be redisplayed and (if no overlay display callback is registered) the layer in use is set to the normal plane. The display callback is called with no parameters. The entire normal plane region should be redisplayed in response to the callback (this includes ancillary buffers if your program depends on their state). GLUT determines when the display callback should be triggered based on the window's redisplay state. The redisplay state for a window can be either set explicitly by calling glutPostRedisplay or implicitly as the result of window damage reported by the window system. Multiple posted redisplays for a window are coalesced by GLUT to minimize the number of display callbacks called. When an overlay is established for a window, but there is no overlay display callback registered, the display callback is used for redisplaying both the overlay and normal plane (that is, it will be called if either the redisplay state or overlay redisplay state is set). In this case, the layer in use is not implicitly changed on entry to the display callback. See glutOverlayDisplayFunc to understand how distinct callbacks for the overlay and normal plane of a window may be established. When a window is created, no display callback exists for the window. It is the responsibility of the programmer to install a display callback for the window before the window is shown. A display callback must be registered for any window that is shown. If a window becomes displayed without a display callback being registered, a fatal error occurs. Passing NULL to glutDisplayFunc is illegal as of GLUT 3.0; there is no way to ``deregister'' a display callback (though another callback routine can always be registered). Upon return from the display callback, the normal damaged state of the window (returned by calling glutLayerGet(GLUT_NORMAL_DAMAGED) is cleared. If there is no overlay display callback registered the overlay damaged state of the window (returned by calling glutLayerGet(GLUT_OVERLAY_DAMAGED) is also cleared. glutReshapeFunc glutReshapeFunc sets the reshape callback for the current window. Usage void glutReshapeFunc(void (*func)(int width, int height)); func The new reshape callback function. Description glutReshapeFunc sets the reshape callback for the current window. The reshape callback is triggered when a window is reshaped. A reshape callback is also triggered immediately before a window's first display callback after a window is created or whenever an overlay for the window is established. The width and height parameters of the callback specify the new window size in pixels. Before the callback, the current window is set to the window that has been reshaped. If a reshape callback is not registered for a window or NULL is passed to glutReshapeFunc (to deregister a previously registered callback), the default reshape 307 Copyright Virtual University of Pakistan

33-OpenGL Programming I

VU

callback is used. This default callback will simply call glViewport(0,0,width,height) on the normal plane (and on the overlay if one exists). If an overlay is established for the window, a single reshape callback is generated. It is the callback's responsibility to update both the normal plane and overlay for the window (changing the layer in use as necessary). When a top-level window is reshaped, subwindows are not reshaped. It is up to the GLUT program to manage the size and positions of subwindows within a top-level window. Still, reshape callbacks will be triggered for subwindows when their size is changed using glutReshapeWindow. glutIdleFunc glutIdleFunc sets the global idle callback. Usage void glutIdleFunc(void (*func)(void)); Description glutIdleFunc sets the global idle callback to be func so a GLUT program can perform background processing tasks or continuous animation when window system events are not being received. If enabled, the idle callback is continuously called when events are not being received. The callback routine has no parameters. The current window and current menu will not be changed before the idle callback. Programs with multiple windows and/or menus should explicitly set the current window and/or current menu and not rely on its current setting. The amount of computation and rendering done in an idle callback should be minimized to avoid affecting the program's interactive response. In general, not more than a single frame of rendering should be done in an idle callback. Passing NULL to glutIdleFunc disables the generation of the idle callback. glutMainLoop glutMainLoop enters the GLUT event processing loop. Usage void glutMainLoop(void); Description glutMainLoop enters the GLUT event processing loop. This routine should be called at most once in a GLUT program. Once called, this routine will never return. It will call as necessary any callbacks that have been registered. glutSwapBuffers glutSwapBuffers swaps the buffers of the current window if double buffered. Usage void glutSwapBuffers(void); Description Performs a buffer swap on the layer in use for the current window. Specifically, glutSwapBuffers promotes the contents of the back buffer of the layer in use of the current window to become the contents of the front buffer. The contents of the back buffer then become undefined. The update typically takes place during the vertical retrace of the monitor, rather than immediately after glutSwapBuffers is called. An implicit glFlush is done by glutSwapBuffers before it returns. Subsequent OpenGL commands can be issued immediately after calling glutSwapBuffers, but are not executed until the buffer exchange is completed. If the layer in use is not double buffered, glutSwapBuffers has no effect. 308 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

Lecture No.34
#include "GL/glut.h" #include <stdlib.h> static void reshape( int w, int h ) { glViewport( 0, 0, w, h ); }

OpenGL Programming - II

static void keyboard( unsigned char key, int x, int y ) { switch( key ) { case 27: //Escape exit(0); break; } } static void display() { } static void idle() { glClearColor(0.0f,0.0f,0.0f,0.0f); glClear(GL_COLOR_BUFFER_BIT|GL_DEPTH_BUFFER_BIT); /////////////////////////////////////////////////////// // glRotatef(0.5f,0.0f,1.0f,0.0); glBegin(GL_TRIANGLES); glVertex3f(0.5f,0.2f,0.0f); glVertex3f(0.5f,0.0f,0.0f); glVertex3f(0.0f,0.0f,0.0f); glEnd(); glutSwapBuffers(); } static void initGL() { float ratio= (float)640 /480 ; glMatrixMode( GL_PROJECTION ); glLoadIdentity(); 309 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

gluPerspective( 45.0,(float) 640 /480 , 1.0, 100.0 ); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); glTranslated(0.0, 0.0, -5.0 ); } int main( int argc, char* argv[] ) { //Initialize the GLUT library. glutInit( &argc, argv ); //Set up the OpenGL rendering context. glutInitDisplayMode( GLUT_RGB | GLUT_DOUBLE | GLUT_DEPTH ); //Create a window and set its width and height. glutCreateWindow( "Deform" ); glutReshapeWindow( 640, 480 ); //Initialize our data structures. initGL(); //The keyboard function gets called whenever we hit a key. glutKeyboardFunc( keyboard ); //The display function gets called whenever the window //requires an update or when we explicitly call //glutPostRedisplay() glutDisplayFunc( display ); //The reshape function gets called whenever the window changes //shape. glutReshapeFunc( reshape ); //The idle function gets called when there are no window //events to process. glutIdleFunc( idle ); //Get the ball rolling! glutMainLoop(); return 0; } //eof In the above program we have first set perspective projection matrix and then rendered a triangle in idle function. 310 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

glMatrixMode The glMatrixMode function specifies which matrix is the current matrix. void glMatrixMode( GLenum mode ); Parameters mode The matrix stack that is the target for subsequent matrix operations. The mode parameter can assume one of three values: Value Meaning GL_MODELVIEW GL_PROJECTION GL_TEXTURE Applies subsequent matrix operations to the modelview matrix stack. Applies subsequent matrix operations to the projection matrix stack. Applies subsequent matrix operations to the texture matrix stack.

Remarks The glMatrixMode function sets the current matrix mode. The following function retrieves information related to glMatrixMode: Error Codes The following are the error codes generated and their conditions. Error code GL_INVALID_ENUM GL_INVALID_OPERATION Condition mode was not an accepted value. glMatrixMode was called between a call to glBegin and the corresponding call to glEnd.

glLoadIdentity The glLoadIdentity function replaces the current matrix with the identity matrix. void glLoadIdentity( void ); Remarks The glLoadIdentity function replaces the current matrix with the identity matrix. It is semantically equivalent to calling glLoadMatrix with the identity matrix

311 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

but in some cases it is more efficient. The following functions retrieve information related to glLoadIdentity: Error Codes The following is the error code and its condition. Error code GL_INVALID_OPERATION Condition glLoadIdentity was called between a call to glBegin and the corresponding call to glEnd.

glTranslated, glTranslatef The glTranslated and glTranslatef functions multiply the current matrix by a translation matrix. void glTranslated( GLdouble x, GLdouble y, GLdouble z ); void glTranslatef( GLfloat x, GLfloat y, GLfloat z ); Parameters x, y, z The x, y, and z coordinates of a translation vector. Remarks The glTranslate function produces the translation specified by (x, y, z). The translation vector is used to compute a 4x4 translation matrix:

The current matrix (see glMatrixMode) is multiplied by this translation matrix, with the product replacing the current matrix. That is, if M is the current matrix and T is the translation matrix, then M is replaced with MT. If the matrix mode is either GL_MODELVIEW or GL_PROJECTION, all objects drawn after glTranslate is called are translated. Use glPushMatrix and glPopMatrix to save and restore the untranslated coordinate system. The following functions retrieve information related to glTranslated and glTranslatef: Error Codes 312 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

The following is the error code and its condition. Error code GL_INVALID_OPERATION Condition glTranslate was called between a call to glBegin and the corresponding call to glEnd.

gluPerspective The gluPerspective is the function of gl utility library. This function sets up a perspective projection matrix. void gluPerspective( GLdouble fovy, GLdouble aspect, GLdouble zNear, GLdouble zFar ); Parameters fovy The field of view angle, in degrees, in the y-direction. aspect The aspect ratio that determines the field of view in the x-direction. The aspect ratio is the ratio of x (width) to y (height). zNear The distance from the viewer to the near clipping plane (always positive). zFar The distance from the viewer to the far clipping plane (always positive). Remarks The gluPerspective function specifies a viewing frustum into the world coordinate system. In general, the aspect ratio in gluPerspective should match the aspect ratio of the associated viewport. For example, aspect = 2.0 means the viewer's angle of view is twice as wide in x as it is in y. If the viewport is twice as wide as it is tall, it displays the image without distortion. The matrix generated by gluPerspective is multiplied by the current matrix, just as if glMultMatrix were called with the generated matrix. To load the perspective matrix onto the current matrix stack instead, precede the call to gluPerspective with a call to glLoadIdentity.

glRotated, glRotatef The glRotated and glRotatef functions multiply the current matrix by a rotation matrix. void glRotated( GLdouble angle, GLdouble x, GLdouble y, GLdouble z ); void glRotatef( GLfloat angle, 313 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

GLfloat x, GLfloat y, GLfloat z ); Parameters angle The angle of rotation, in degrees. x, y, z The x, y, and z coordinates of a vector, respectively. Remarks The glRotate function computes a matrix that performs a counterclockwise rotation of angle degrees about the vector from the origin through the point (x, y, z). The current matrix (see glMatrixMode) is multiplied by this rotation matrix, with the product replacing the current matrix. That is, if M is the current matrix and R is the translation matrix, then M is replaced with MR. If the matrix mode is either GL_MODELVIEW or GL_PROJECTION, all objects drawn after glRotate is called are rotated. Use glPushMatrix and glPopMatrix to save and restore the unrotated coordinate system. Error Codes The following is the error code and its condition. Error code GL_INVALID_OPERATION Condition glRotate was called between a call to glBegin and the corresponding call to glEnd.

glClearColor The glClearColor function specifies clear values for the color buffers. void glClearColor( GLclampf red, GLclampf green, GLclampf blue, GLclampf alpha ); Parameters red, green, blue, alpha The red, green, blue, and alpha values that glClear uses to clear the color buffers. The default values are all zero. Remarks The glClearColor function specifies the red, green, blue, and alpha values used by glClear to clear the color buffers. Values specified by glClearColor are clamped to the range [0,1]. Error Codes The following is the error code generated and its condition. Error code Condition 314 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

GL_INVALID_OPERATION glColor These functions set the current color.

glClearColor was called between a call to glBegin and the corresponding call to glEnd.

glColor3b, glColor3d, glColor3f, glColor3i, glColor3ub, glColor3ui, glColor3us, glColor4b, glColor4f, glColor4i, glColor4s, glColor4ub, glColor4us, glColor3bv, glColor3dv, glColor3fv, glColor3sv, glColor3ubv, glColor3uiv, glColor4bv, glColor4dv, glColor4fv, glColor4iv, glColor4ubv, glColor4uiv, glColor4usv.

glColor3s, glColor4d, glColor4ui, glColor3iv, glColor3usv, glColor4sv,

void glColor3b( GLbyte red, GLbyte green, GLbyte blue ); void glColor3f( GLfloat red, GLfloat green, GLfloat blue ); Consult yourself for the documentation of rest of the functions of this type. Parameters red, green, blue New red, green, and blue values for the current color. alpha A new alpha value for the current color. Included only in the four-argument glColor4 function. Remarks OpenGL stores both a current single-valued color index and a current four-valued RGBA color. The glColor function sets a new four-valued RGBA color. There are two major variants to glColor:

The glColor3 variants specify new red, green, and blue values explicitly, and set the current alpha value to 1.0 implicitly.

The glColor4 variants specify all four color components explicitly.

The glColor3b, glColor4b, glColor3s, glColor4s, glColor3i, and glColor4i functions take three or four signed byte, short, or long integers as arguments. When you append v to the name, the color functions can take a pointer to an array of such values. Current color values are stored in floating-point format, with unspecified mantissa and exponent sizes. Unsigned integer color components, when specified, are linearly mapped to floating-point values such that the largest representable value maps to 1.0 (full intensity), and zero maps to 0.0 (zero intensity). Signed integer color components, when specified, are linearly mapped to floating-point values such that the most positive representable value maps to 1.0, and the most negative representable value maps to 1.0. Floating-point values are mapped directly. 315 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

Neither floating-point nor signed integer values are clamped to the range [0,1] before updating the current color. However, color components are clamped to this range before they are interpolated or written into a color buffer. You can update the current color at any time. In particular, you can call glColor between a call to glBegin and the corresponding call to glEnd. glClear The glClear function clears buffers to preset values. void glClear( GLbitfield mask ); Parameters mask Bitwise OR operators of masks that indicate the buffers to be cleared. The four masks are as follows. Mask Buffer to be Cleared GL_COLOR_BUFFER_BIT GL_DEPTH_BUFFER_BIT GL_ACCUM_BUFFER_BIT GL_STENCIL_BUFFER_BIT Remarks The glClear function sets the bitplane area of the window to values previously selected by glClearColor, glClearIndex, glClearDepth, glClearStencil, and glClearAccum. You can clear multiple color buffers simultaneously by selecting more than one buffer at a time using glDrawBuffer. The pixel-ownership test, the scissor test, dithering, and the buffer writemasks affect the operation of glClear. The scissor box bounds the cleared region. The glClear function ignores the alpha function, blend function, logical operation, stenciling, texture mapping, and z-buffering. The glClear function takes a single argument (mask) that is the bitwise OR of several values indicating which buffer is to be cleared. The value to which each buffer is cleared depends on the setting of the clear value for that buffer. If a buffer is not present, a glClear call directed at that buffer has no effect. Error Codes The buffers currently enabled for color writing. The depth buffer. The accumulation buffer. The stencil buffer.

The following are the error codes generated and their conditions.

316 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

Error code GL_INVALID_VALUE GL_INVALID_OPERATION

Condition Any bit other than the four defined bits was set in mask. glClear was called between a call to glBegin and the corresponding call to glEnd.

glBegin, glEnd The glBegin and glEnd functions delimit the vertices of a primitive or a group of like primitives. void glBegin( GLenum mode ); void glEnd( void ); Parameters mode The primitive or primitives that will be created from vertices presented between glBegin and the subsequent glEnd. The following are accepted symbolic constants and their meanings: GL_POINTS Treats each vertex as a single point. Vertex n defines point n. N points are drawn. GL_LINES Treats each pair of vertices as an independent line segment. Vertices 2n 1 and 2n define line n. N/2 lines are drawn. GL_LINE_STRIP Draws a connected group of line segments from the first vertex to the last. Vertices n and n+1 define line n. N 1 lines are drawn. GL_LINE_LOOP Draws a connected group of line segments from the first vertex to the last, then back to the first. Vertices n and n+1 define line n. The last line, however, is defined by vertices N and 1. N lines are drawn. GL_TRIANGLES Treats each triplet of vertices as an independent triangle. Vertices 3n 2, 3n 1, and 3n define triangle n. N/3 triangles are drawn. GL_TRIANGLE_STRIP Draws a connected group of triangles. One triangle is defined for each vertex presented after the first two vertices. For odd n, vertices n, n + 1, and n + 2 define triangle n. For even n, vertices n + 1, n, and n + 2 define triangle n. N 2 triangles are drawn. GL_TRIANGLE_FAN Draws a connected group of triangles. One triangle is defined for each vertex presented after the first two vertices. Vertices 1, n + 1, and n + 2 define triangle n. N 2 triangles are drawn. GL_QUADS Treats each group of four vertices as an independent quadrilateral. Vertices 4n 3, 4n 2, 4n 1, and 4n define quadrilateral n. N/4 quadrilaterals are drawn. 317 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

GL_QUAD_STRIP Draws a connected group of quadrilaterals. One quadrilateral is defined for each pair of vertices presented after the first pair. Vertices 2n 1, 2n, 2n + 2, and 2n + 1 define quadrilateral n. N quadrilaterals are drawn. Note that the order in which vertices are used to construct a quadrilateral from strip data is different from that used with independent data. GL_POLYGON Draws a single, convex polygon. Vertices 1 through N define this polygon. Remarks The glBegin and glEnd functions delimit the vertices that define a primitive or a group of like primitives. The glBegin function accepts a single argument that specifies which of ten primitives the vertices compose. Taking n as an integer count starting at one, and N as the total number of vertices specified, the interpretations are as follows:

You can use only a subset of OpenGL functions between glBegin and glEnd. The functions you can use are: glVertex glColor glIndex glNormal glMaterial You can also use glCallList or glCallLists to execute display lists that include only the preceding functions. If any other OpenGL function is called between glBegin and glEnd, the error flag is set and the function is ignored.

Regardless of the value chosen for mode in glBegin, there is no limit to the number of vertices you can define between glBegin and glEnd. Lines, triangles, quadrilaterals, and polygons that are incompletely specified are not drawn. Incomplete specification results when either too few vertices are provided to specify even a single primitive or when an incorrect multiple of vertices is specified. The incomplete primitive is ignored; the complete primitives are drawn.

The minimum specification of vertices for each primitive is: Minimum number of vertices Type of primitive 1 2 3 4 3 Point Line Triangle Quadrilateral Polygon 318 Copyright Virtual University of Pakistan

34-OpenGL Programming II

VU

Modes that require a certain multiple of vertices are GL_LINES (2), GL_TRIANGLES (3), GL_QUADS (4), and GL_QUAD_STRIP (2). Error Codes The following are the error codes generated and their conditions. Error code GL_INVALID_ENUM GL_INVALID_OPERATION Condition mode was set to an unaccepted value. A function other than glVertex, glColor, glIndex, glNormal, glTexCoord, glEvalCoord, glEvalPoint, glMaterial, glEdgeFlag, glCallList, or glCallLists was called between glBegin and the corresponding glEnd. The function glEnd was called before the corresponding glBegin was called, or glBegin was called within a glBegin/glEnd sequence.

glVertex These functions specify a vertex. glVertex2d, glVertex2f, glVertex2i, glVertex3f, glVertex3i, glVertex3s, glVertex4i, glVertex4s, glVertex2dv, glVertex2sv, glVertex3dv, glVertex3fv, glVertex4dv, glVertex4fv, glVertex4iv, glVertex4sv glVertex2s, glVertex4d, glVertex2fv, glVertex3iv, glVertex3d, glVertex4f, glVertex2iv, glVertex3sv,

void glVertex3f( GLfloat x, GLfloat y, GLfloat z ); Consult the documentation to yourself for the functions of the same nature. Parameters x, y, z, w The x, y, z, and w coordinates of a vertex. Not all parameters are present in all forms of the command. Remarks The glVertex function commands are used within glBegin/glEnd pairs to specify point, line, and polygon vertices. The current color, normal, and texture coordinates are associated with the vertex when glVertex is called. When only x and y are specified, z defaults to 0.0 and w defaults to 1.0. When x, y, and z are specified, w defaults to 1.0. Invoking glVertex outside of a glBegin/glEnd pair results in undefined behavior.

319 Copyright Virtual University of Pakistan

35-Curves

VU

Lecture No.35

Curves

We all know what a curve is. In this lecture we will explore the mathematical definition of a curve in a form that is very useful to geometric modeling and other computer graphics applications: that definition consists of a set of parametric equations. The mathematics of parametric equations is the basis for Bezier, NURBS (Non Uniform Rational Beta Splines), and Hermite curves. We will discuss both plane curves and space curves here and also discussion of the tangent vector, blending functions, conic curves, re-parameterization, and continuity and composite curves. Parametric Equations of a Curve A parametric curve is one whose defining equations are given in terms of a single, common, independent variable called the parametric variable. We have already encountered parametric variables in earlier discussions of vectors, lines, and planes. Imagine a curve in three-dimensional space, each point on the curve has a unique set of coordinates: a specific x value, y value, and z value. Each coordinate is controlled by a separate parametric equation, whose general form looks like

x = x (u ),

y = y (u ),

z = z (u )

LLL (1)

where x(u) stands for some as yet unspecified function in which u is the independent variable; for example, x(u) = au2 + bu + c, and similarly for y(u) and z(u). It is important to understand that each of these is an independent expression. This will become clear as we discuss specific examples later. The dependent variables are the x,y, and z coordinates themselves, because their values depend on the value of the parametric variable u. Engineers and programmers who do geometric modeling usually prefer these kinds of expressions because the coordinates x, y, and z are independent of each other, and each is defined by its own parametric equation.

Figure 1: point of a curve defined by a vector. Each point on a curve is defined by a vector p (figure 1). The components of this vector are x(u), y(u), and z(u). We express this as

p = p (u )

LLL ( 2)

Which says that the vector p is a function of the parametric variable u.

320 Copyright Virtual University of Pakistan

35-Curves

VU

There is a lot of information in equation 2. When we expand it into component form, it becomes

p (u ) = [ x (u )

y (u )

z (u )]

LLL (3)

The specific functions that define the vector components of p determine the shape of the curve. In fact, this is one way to define a curve by simple choosing or designing these mathematical functions. There re only a few simple rules that we must follow: 1) define each component by a single, common parametric variable, and 2) make sure that each point on the curve corresponds to a unique value of the parametric variable. The last rule can be put another way: each value of the parametric variable must correspond to a unique point on the curve. Plane Curves To define plane curves, we use parametric functions that are second-degree polynomials:

x (u ) = a xu 2 + bxu + c x y (u ) = a y u 2 + b y u + c y z (u ) = a z u 2 + bz u + c z
Where the a, b, and c terms are constant coefficients. We can combine x(u), y(u), z(u), and their respective coefficients into and equivalent, more concise, vector equation:

LLL ( 4)

P (u ) = au 2 + bu + c

LLL (5)

We allow the parametric variable to take on values only in the interval 0 u 1. This ensures that the equation produces a bounded line segment. The coefficients a, b, c, in this equation are vectors, and each has three components; for example, a = [ax ay az ]. This curve has serious limitations. Although it can generate all the conic curves, or a close approximation to them, it cannot generate curve with an inflection point, like an Sshaped curve, no matter what values we select for the coefficients a, b, c. to do this requires a cubic polynomial. How do we define a specific plane curve, one that we can display, with define end points, and a precise orientation in space? first, note in equation 4 or 5 that there are nine coefficients that we must determine: ax, bx, , cz. if we know the two end points and the intermediate point. End points and an intermediate point on the curve, then we now nine quantities that we can express in terms of these coefficients (3 points x 3 coordinates each = 9 known quantities), and we can use these three points to define a unique curve (Figure 2). By applying some simple algebra to these relationships, we can rewrite Equation 5 in terms 321 Copyright Virtual University of Pakistan

35-Curves

VU

of the three points. To one of the end points we assign u = 0, and to the other u = 1. To the intermediate point, we arbitrarily assign u = 0.5. We can write this points as P0 = [x0 y0 P0.5 = [x0.5 y0.5 P1 = [x1 y1 z0] z0.5] z1]

(6)

Where the subscripts indicate the value of the parametric variable at each point. Now we solve equations 4 for the ax, bx, , cz coefficients in terms of these points. Thus, for x at u = 0, u =0.5, and u =1, we have x0 = cx x0.5 = 0.25ax + 0.5bx + cx x1 = ax + bx + cx with similar equations for y, and z.

(7)

Figure 2: A plane curve defined by three points

Next we solve these three equations in three unknowns for ax, bx, and cx, finding ax = 2x0 - 4x0.5 + 2x1 bx = -3x0 + 4x0.5 + x1 cx = x0 Substituting this result in to equation 4 yields x(u) = (2x0 - 4x0.5 + 2x1 ) u2 + (-3x0 + 4x0.5 - x1) u + x0 Again, there are equivalent expressions for y( u) and z(u). We rewrite equation 9 as follows: x(u) = (2u2 3u + 1) x0 + (-4u2 + 4u)x0.5 + (2u2 u) x1 (10) (9) (8)

322 Copyright Virtual University of Pakistan

35-Curves

VU

Using this result and equivalent expressions for y(u) and z(u), we combine them into a single vector equation: P(u) = (2u2 3u + 1) P0 + (-4u2 + 4u)P0.5 + (2u2 u) P1 (11)

Equation 11 produces the same curve as Equation 5. The curve will always lie in a plane no matter what three points we choose, Furthermore, it is interesting to note that the point P0.5 which is on the curve a t u= 0.5, is not necessarily half way along the length of the curve between p0 and p1. We can show this quite convincingly by choosing three points to define a curve such that two of them are relatively close together (figure 3). In fact, if we assign a different value to the parametric variable for the intermediate point, then we obtain different values for the coefficients in equations 8. This, in turn, means that a different curve is produced, although it passes through the same three points. Equation 5 is the algebraic form and equation 11 is the geometric form. Each of these equations can be written more compactly with matrices. Compactness is not the only advantage to matrix notation. Once a curve is defined in matrix form, we can use the full power of matrix algebra to solve many geometry problems.

Figure 3: Curve defined by three non-uniformly spaced points. So now we rewrite equation 5 using the following substitutions:

[u 2

a = au 2 + bu + c u 1] b c

LLL (12)

U = [u 2

u 1]

LLL (13)
LLL (14)

A = [ a b c ]T

323 Copyright Virtual University of Pakistan

35-Curves

VU

And finally we obtain

p (u ) = UA

LLL (15)
az bz cz

Remember that A is really a matrix of vectors, so that

a a x = b A= b x c cx

ay by cy

LLL (16)

The nine terms on the right are called the algebraic coefficients. Next, we convert equation 11 into matrix form. The right-hand side looks like the product of two matrices: [(2u 2 3u + 1) (4u 2 + 4u ) (2u 2 u )] and [ p 0 p 0.5 p1 ] . This means that

p0 p(u) = [(2u2 3u +1) (4u2 + 4u) (2u2 u)] p 0 . 5 p1

Using the following substitutions:

LLL(17)

F = [(2u 2 3u + 1) (4u 2 + 4u ) (2u 2 u )]

and

LLL(18)

p0 x0 = x p= p 0 . 5 0 .5 x1 p1

y0 y 0 .5 y1

z0 z 0 .5 z1

LLL (19)

where P is the control point matrix and the nine terms on the right are its elements or the geometric coefficients, we can now write

p (u ) = FP

LLL ( 20)

This is the matrix version of the geometric form. Because it is the same curve in algebraic form, p(u)=UA, or geometric form, p(u)=FP, we can write

FP = UA

LLL ( 21)
324

The F matrix is itself the product of two other matrices:

Copyright Virtual University of Pakistan

35-Curves

VU

F = [u 2

2 4 2 3 4 1 u 1] 0 0 1

LLL ( 22)

The matrix on the left we recognize as U, and we can denote the other matrix as

2 4 2 3 4 1 M = 0 0 1
This means that

LLL ( 23)

F = UM
Using this we substitute appropriately to find

LLL ( 24)

UMP = UA

LLL ( 25)
LLL ( 26) LLL ( 27 ) LLL ( 28)

Pre-multiplying each side of this equation by U 1 yields

MP = A A = MP
Or

This expresses a simple relationship between the algebraic and geometric coefficients

P = M 1 A

The matrix M is called a basis transformation matrix, and F is called a blending function matrix.

325 Copyright Virtual University of Pakistan

36-Space Curves

VU

Lecture No.36

Space Curves

A space curve is not confined to a plane. It is free to twist through space. To define a space curve we must use parametric functions that are cubic polynomials. For x (u) we write
x (u ) = a x u 3 + bx u 2 + c x u + d x
(1)

With similar expressions for y(u) and z(u). Again the a, b, c and d terms are constant coefficients. As we did with Equation for a plane curve, we combine the x(u), y(u) , and z(u) expressions into a single vector equation :
p (u ) = au 3 + bu 2 + cu + d

(2)

If a = 0, then this equation is identical to Equation discussed in plane curves To define a specific curve in space, we use the same approach as we did for a plane curve. This time, though, there are 12 coefficients to be determined. We specify four points through which we want the curve to pass, which provides all the information we need to determine a, b, c, and d. but which four points? Two are obvious: p (0) and p (1) , the end points at u = 0 and u = 1 . For various reasons beyond our scope, it turns out to be 1 advantageous to use two intermediate points that we assign parametric values of u = 3 1 2 2 and u = , or p and p . So we now have the four points we need: 3 3 3
P1 P1/3 z P0 y x P2/3

Figure 1 Four points define a cubic space curve

1 2 p (0) , p and p and p (1) , which we can rewrite as the more convenient p1, p2, 3 3 p3, and p4 (Figure 1). Substituting each of the values of the parametric variable (u = 0, 1 3, 2 3, 1) into Equation 1, we obtain the following four equations in four unknowns: x1 = d x 1 1 1 x2 = a x + bx + c x + d x 27 9 3 8 4 2 x 3 = a x + bx + c x + d x 27 9 3

(3)

326 Copyright Virtual University of Pakistan

36-Space Curves

VU

x 4 = a x + bx + c x + d x Now we can express ax, bx, cx and dx in terms of x1, x2, x3, and x4. After doing the necessary algebra, we obtain 9 27 27 9 x1 + x2 x3 + x4 2 2 2 2 45 9 bx = 9 x1 x2 + 18 x3 x4 2 2 11 9 c x = x1 + 9 x2 x3 + x4 2 2 d x = x1 ax = We substitute these results into Equation 1, producing
27 27 9 45 9 9 x (u ) = x1 + x2 x3 + x4 u 3 + 9 x1 x2 + 18 x3 x u 2 2 2 2 2 2 2 9 11 + x1 + 9 x2 x3 + x 4 u + x1 2 2

(4)

(5)

All this looks a bit messy right now, but we can put it into a neat, much more compact form. We begin by rewriting Equation 5 as follows:

11 45 27 9 x (u ) = u 3 + 9u 2 u + 1 x1 + u 3 u 2 + 9u x2 2 2 2 2 9 9 27 3 u + 18u 2 9u x3 + u 3 u 2 + u x 4 + 2 2 2

(6)

Using equivalent expressions for y(u) and z(u), we can summarize them as a single vector equation:
11 45 27 9 p (u ) = u 3 + 9u 2 u + 1 p1 + u 3 u 2 + 9u p2 2 2 2 2 9 9 27 3 u + 18u 2 9u p3 + u 3 u 2 + u p4 + 2 2 2

(7)

This means that, given four point assigned successive values of u (in this case at u=0, 1/3, 2/3, 1), equation 7 produces a curve that starts at p1, passes through p2 and p3, and ends at p4. Now lets take one more step toward a more compact notation. Using the four parametric functions appearing in Equation 7, we define a new matrix, G = [G1 G2 G3 G4 ] , where 11 9 G1 = u 3 + 9u 2 u + 1 2 2 27 G3 = u 3 + 18u 2 9u 2 45 27 G 2 = u 3 u 2 + 9u 2 2 9 9 G4 = u 3 u 2 + u 2 2 P2 P3
T

(8)

And then define a matrix P containing the control points, P = [P1

P4 ] , so that

327 Copyright Virtual University of Pakistan

36-Space Curves

VU

P (u ) = GP The matrix G is the product of two other matrices, U and N:

G = UN

(9) (10)

Where U = u 3

u2

u 1 and
27 2 9 2 9 2 0 9 2 1 1 0

27 9 2 2 45 9 N = 2 11 9 2 0 1 Now we let

(11)

(Note that N is another example of a basis transformation matrix.)

a a x b b x A= = c cx d d x P (u ) = UA

ay by cy dy

az bz cz dz

(12)

Using matrices, Equation 2 becomes

(13)

Which looks a lot like Equation for a plane curve, except that we have defined new U and A matrices? In fact, Equation is a special case of the formulation for a space curve. To convert the information in the A matrix into that required for the P matrix, we do some simple matrix algebra, using Equations 9, 10 and 13. First we have
GP = UNP

(14) (15) (16)

And then
UA = UNP

Or more simply
A = NP

328 Copyright Virtual University of Pakistan

37-The Tangent Vector

VU

Lecture No.37

The Tangent Vector

Another way to define a space curve does not use intermediate points. It uses the tangents at each end of the curve, instead. Every point on a curve has a straight line associated with it called the tangent line, which is related to the first derivation of the Parametric functions x(u), y(u), and z(u), such as those given by Equation 2 of previous lecture. Thus

d x (u ), du

d y (u ), du

and

d z (u ) du

(1)

From elementary calculus, we can compute, for example,

dy dy (u ) / du = du dx (u ) / du

(2)

We can treat dx (u ) du, dy (u ) du, and dz (u ) du as components of a vector along the tangent line to the curve. We call this the tangent vector, and define it as

d P u (u ) = x (u )i du
Or more simply as

d y (u ) j du

d z (u )k du

(3)

Pu = xu

yu

zu

(4)

(Here the superscript u indicates the first derivative operation with respect to the independent variable u). This is a very powerful idea, and we will now see how to use it to define a curve. In the last section, we discussed how to define a curve by specifying four points. Now we have another way to define a curve. We will still use the two end points, but instead of two intermediate points, we will use the tangent vectors at each end to supply the information we need to define a curve. By manipulating these tangent vectors, we can , and p1u are called the control the slope at each end. The set of vectors p0 , p1 , p u 0

boundary conditions. This method itself is called the cubic Hermite interpolation, after C. Hermite (1822-1901) the French mathematician who made significant contributions to our understanding of cubic and quadratic polynomials.
We differentiate to obtain the x component of the tangent vector:

d dx (u ) = x u = 3a x u 2 + 2bx u + c x du
Pu 1 P1 Pu 0 P0

(5)

Figure 1 Defining a curve using end points and tangent vectors.

x (u ) = a x u 3 + bx u 2 + c x u + d x
Copyright Virtual University of Pakistan

(1A)
329

37-The Tangent Vector

VU

Evaluating (1A) and Equation 5 at u = 0, u = 1, yields

x0 = d x x1 = a x + bx + c x + d x
u x0 = cx

(6)

x1u = 3a x + 2bx + c x

Using these four equations in four unknowns, we solve for ax , bx , cx and dx in terms of the boundary conditions
u bx = 3( x 0 + x1 ) 2 x0 x1u u c x = x0 u a x = 2( x0 x1 ) + x0 + x1u

(7)

d x = x0

Substituting the result into Equation (1A), yields

u u u )u 2 + x0u + x0 x (u ) = (2 x0 2 x1 + x0 + x1u )u 3 + ( 3x0 + 3x1 2 x0 x1u x0

(8)

Rearranging terms we can rewrite this as

u x (u ) = (2u 3 3u 2 + 1)x0 + ( 2u 3 + 3u 2 )x1 + (u 3 2u 2 + u )x0 + (u 3 u 2 )x1u

(9)

Because y(u) and z(u) have equivalent forms, we can include them by rewriting Equation 9 in vector form:
u p (u ) = (2u 3 3u 2 + 1)p0 + ( 2u 3 + 3u 2 )p1 + (u 3 2u 2 + u )p0 + (u 3 u 2 )p1u

(10)

To express Equation 10 in matrix notation, we first define a blending function matrix F = [F1 F2 F3 F4 ] , where
F1 = 2u 3 3u 2 + 1 F2 = 2u 3 + 3u 2 F3 = u 3 2u 2 + u F4 = u 3 u 2 These matrix elements are the polynomial coefficients of the vectors which we rewrite as
u p (u ) = F1 p0 + F2 p1 + F3 p0 + F4 p1u

(11)

(12)

If we assemble the vectors representing the boundary conditions into a matrix B,

B = p0
Then

p1

u p0

p1u

(13)

p (u ) = FB

(14) (15)
330

Here again we write the matrix F as the product of two matrices, U and M, so that

F = UM
where
Copyright Virtual University of Pakistan

37-The Tangent Vector

VU

U = u3
and

u2

u 1

(16)

1 2 2 1 3 3 2 1 M= 0 1 0 0 0 0 0 1 Rewriting Equation 14 using these substitutions, we obtain

P (u ) = UMB

(17)

(18)

It is easy to show that the relationship between the algebraic and geometric coefficients for a space curve. Since
P (u ) = UA

(19)

the relationship between A and B is, again,

A = MB

(20)

Consider the four vectors that make up the boundary condition matrix. There is nothing extraordinary about the vectors defining the end points, but what about the two tangent vector? A tangent vector certainly defines the slope at one end of the curve, but a vector has characteristics of both direction and magnitude. All we need to specify ht slope is a unit tangent vector at each end, say t0 and t1. But p0, p1, t0, and t1 supply only 10 of the 12 pieces of information needed to completely determine the curve. So the magnitude of the tangent vector is also necessary and contributes to the shape of the curve. In fact, we can u and p1u as: write p0
u p0 = m0 t 0

(21)

And

p1u = m1t1
u Clearly, m0, and m1 are the magnitudes of p0 and p1u .

(22)

Using these relationships, we modify Equation 10 as follows:

p (u ) = (2u 3 3u 2 + 1)p0 + ( 2u 3 + 3u 2 )p1 + (u 3 2u 2 + u )m0 t0 + (u 3 u 2 )m1t1

(23)

Now we can experiment with a curve (Figure 2). Lets hold p0, p1, t0, and t1 constants and see what happens to the shape of the curve as we vary m0 and m1. For simplicity we u and z1u are all equal to will consider a curve in the x, y plane. This means that z0, z1, z 0 zero. The B matrix for the curve drawn with the bold line (and with m0 = m1 = 1) is

331 Copyright Virtual University of Pakistan

37-The Tangent Vector

VU

P0 P B= 1 m0 t 0 m1t1

0 1 0.707 0.707

0 0 0.707 0.707

0 0 0 0

(24)

Carefully consider this array of 12 elements; they uniquely define the curve. By changing either m0 or m1, or both, we can change the shape of the curve. But it is a restricted kind of change because not only do the end points remain fixed, but the end slopes are also unchanged! The three curves drawn with light lines in Figure 3 show the effects of varying m0 and m1. This is a very powerful tool for designing curves, making it possible to join up endto-end many curves in a smooth way and still exert some control over the interior shape of each individual curve. For example, as we increase the value of m0 while holding m1 fixed, the curve seems to be pushed toward p1. Keeping m0 and m1 equal but increasing their value increases the maximum deflection of the curve from the x-axis and increases the curvature at the maximum. (Under some conditions, not necessarily desirable, we can force a loop to form).

332 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

Lecture No.38

Bezier Curves

The Bezier curve is an important part of almost every computer-graphics illustration program and computer-aided design system in use today. It is used in many ways, from designing the curves and surfaces of automobiles to defining the shape of letters in type fonts. And because it is numerically the most stable of all the polynomial-based curves used in these applications, the Bezier curve is the ideal standard for representing the more complex piecewise polynomial curves. In the early 1960s, Peter Bezier began looking for a better way to define curves and surfaces one that would be useful to a design engineer. He was familiar with the work of Ferguson and Coons and their parametric cubic curves and bicubic surfaces. However, these did not offer an intuitive way to alter and control shape. The results of Beziers research led to the curves and surfaces that bear his name and became part of the UNISURF system. The French automobile manufacturer, Renault used UNISURF to design the sculptured surfaces of many of its products.

A Geometric Construction
We can draw a Bezier curve using a simple recursive geometric construction. Lets begin by constructing a second-degree curve. We select three points A, B, C so that line AB is a tangent to the curve at A, and BC is tangent at C. The curve begins at A and ends at C. For any ratio, ui, we construct points D and E so that

Figure 1: Geometric Construction of a second-degree Bezier curve.

AD BE = = ui AB BC

(1)

On DE we construct F so that DF/DE = uj . Point F is on the curve. Repeating this process for other values of ui, we produce a series of points on a Bezier curve. Note that we must be consistent in the order in which we sub-divide AB and BC. To define this curve in a coordinate system, let point A = (xA, yA), B= (xB, yB). Then coordinates of points D and E for some value of ui are x D = x A + ui ( x B x A ) (2) y D = y A + ui ( y B y A ) And 333 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

x E = x B + ui ( x C x B )

y E = y B + ui ( y C y B )

(3)

(4) yF = yD + ui ( yE yD ) To obtain xF and yF in terms of the coordinates of points A, B and C, for any value of ui in the unit interval, we substitute appropriately from Equation 2 and 3 into equations 4 of plane curve. After rearranging terms to simplify, we find 2 x F = (1 ui ) x A + 2ui (1 ui )x B + ui2 xC (5) 2 y F = (1 ui ) y A + 2ui (1 ui ) y B + ui2 y C We generalize this set of equations for any point on the curve using the following substitutions: x (u ) = x F (6) y (u ) = y F And we let x0 = x A x1 = x B x 2 = xC (7) y 0 = y A y1 = y B y 2 = y C

The coordinates of point F for some value of ui are xF = xD + ui (xE xD )

Now we can rewrite Equation 5 as

2 2

x(u ) = (1 u ) x0 + 2u (1 u )x1 + u 2 x2 y (u ) = (1 u ) y0 + 2u (1 u ) y1 + u 2 y2

(8)

This is the set of second-degree equations for the coordinates of points on Bezier curve based on our construction. We express this construction process and Equations 8 in terms of vectors with the following substitutions. Let the vector po represent point A, p1 point B, and p2 point C. Form vector geometry we have D >>>> and E == >>> if we let F = >>> we see that p (u ) = p0 + u ( p1 p0 ) + u[ p1 + u ( p 2 p1 ) p0 + u ( p1 p0 )] (9) We rearrange terms to obtain a more compact vector equation of a second degree Bezier curve: p (u ) = (1 u ) p0 + 2u (1 u ) p1 + u 2 p 2
2

(10)

The ratio u is the parametric variable. Later we will see that this equation is an example of a Bernstein polynomial. Note that the curve will always lie in the plane containing the three control points, but the points do not necessarily lie in the xy plane. Similar constructions apply to Bezier curves of any degree. In fact the degree of a Bezier curve is equal to n-1, where n is the number of control points. Figure 2 shows the construction of point on a cubic Bezier curve, which requires four control points A, B, C, and D to define it. The curve begins at point A tangent to the, AB, and ends at D and tangent to CD. We construct points E, F, and G so that

334 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

AE BF CG = = = ui AB BC CD

(11)

Figure 2: Geometric construction of a cubic Bezier curve On EF and FG we locate H and I, respectively, so that EH FI (12) = = ui EF FG Finally on HI we locate J so that HJ (13) = ui HI We can make no more subdivisions, which means that point J is on the curve. If we continue this process for a sequence of points, then their locus defines the curve. If points A, B, C, and D are represented by the vectors >>>>> respectively, then expressing the construction of the intermediate points E, F, G, H, and I in terms of these vectors to produce point J, or P(u), yields p (u ) = p0 + u ( p1 p0 ) + u[ p1 + u ( p2 p1 ) p0 + u ( p1 p0 )] u ( p1 p0 ) + u[ p1 + u ( p2 p1 ) p0 u ( p1 p0 )]} This awkward expression simplifies nicely to 3 2 p (u ) = (1 u ) p0 + 3u (1 u ) p1 + 3u 2 (1 u ) p 2 + u 3 p0 + u{ p1 + u ( p2 p1 ) + u[ p2 + u ( p3 p2 ) p1 + u ( p2 p0 )] p0 (14)

(15)

Of course this construction, of a cubic curve with its four control points is done in the plane of the paper. However, the cubic polynomial allows a curve that is nonplanar; that is, it can represent a curve that twists in space. The geometric construction of a Bezier curve shows how the control points influence its shape. The curve begins on the first point and ends on the last point. It is tangent to the lines connecting the first two points and the last two points. The curve is always contained within the convex hull of the control points. No one spends time constructing and plotting the points of a Bezier curve by hand, of course. A computer does a much faster and more accurate job. However, it is worth doing several curves this way for insight into the characteristics of Bezier curves.
An Algebraic Definition Bezier began with the idea that any point p(u) on a curve segment should be given by an equation such as the following:

335 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

p (u ) = pi f i (u )
i =0

(16)

Figure 3: Bezier curves and their control points Equation 16 is a compact way to express the sum of several similar terms because what it says is this: p (u ) = p0 f 0 (u ) + p1 f1 (u ) + ........ + pn f n (u ) (17) Of which Equation 10 and 15 are specific examples, for n=2 and n=3, respectively. The n+1 functions, that is the fi(u) must produce a cure that has certain well-defined characteristics. Here are some of the most important ones: 1. The curve must start on the first control point, p0, and end on the last, pn. Mathematically, we say that the functions must interpolate these two points. 2. The curve must be tangent to the line given by p1 p0 at pn - pn-1 at pn . 3. The functions fi(u) must be symmetric with respect to u and (1-u). This lets us reverse the sequence of control points without changing the shape of the curve. Other characteristics can be found in more advanced works on the subject. Figure

Figure 4: Three different sequences of four control points 336 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

Figure 5: Modifying the shape of a Bezier curve. A family of functions called Bernstein polynomials satisfies these requirements. They are the basis functions of the Bezier curve (Other curves, such as the NURBS curves, use different, but related, basis functions). We rewrite Equation 16 using them so that p (u ) = pi Bi ,n (u )
i =0 n

(18)

Were the basis functions are n i Bi ,n (u ) = (in )u i (1 u )

(19)
(20)

! ( ) = i! (nn 1)!
n i

The symbol ! is the factorial operator. For example, 3! = 3 x 2 x 1, 5! = 5 x 4 x 3 x 2 x 1, so forth. We use the following conventions when evaluating Equation20; If i and u equal zero, then ui = 1 and 0! = 1. We see that for n+1 control points, the basis functions produce an nth-degree polynomial. Expanding Equation 18 for a second degree Bezier curve (When n=2 and there are three control points) produces p (u ) = p0 B0, 2 (u ) + p1 B1, 2 (u ) + p2 B2, 2 (u ) From Equation 20, we find B0, 2 (u ) = (1 u )
B2, 2 (u ) = u 2
2

(21)

(22) (23) (24)

337

B1,2 (u ) = 2u (1 u )

Copyright Virtual University of Pakistan

38-Bezier Curves

VU

These are the basis functions for a second-degree Bezier curve.

Figure 6: Bezier curve basis functions Substituting them into Equation 21 and rearranging terms, we find p (u ) = (1 u ) p0 + 2u (1 u ) p1 + u 2 p 2
2

(25)

This is the same expression we found from the geometric construction, Equation 10. The variable u is now called the parametric variable. Now, lets expand Equation 18 for a cubic Bezier curve, where n=3: p (u ) = p0 B0,3 (u ) + p1 B1,3 (u ) + p2 B2,3 (u ) + p3 B3,3 (u ) And from Equation 20 we find B0,3 (u ) = (1 u )
3

(26)

(27)
2

B1,3 (u ) = 3u (1 u )

(28) (29) (30)

B2,3 (u ) = 3u 2 (1 u )

B3,3 (u ) = u 3

Substituting these into Equation 26 and rearranging terms produces 3 2 p (u ) = (1 u ) p0 + 3u (1 u ) p1 + 3u 2 (1 u ) p 2 + u 3 p3

(31)

Bezier curve equations are well suited for expression in matrix form. We can expand the cubic parametric functions and rewrite Equation 31 as

338 Copyright Virtual University of Pakistan

38-Bezier Curves

VU

(1 3u + 3u 2 u 3 ) ( 3u 6u 2 + 3u 3 ) p (u ) = (3u 2 3u 3 ) u3 or as

p0 p 1 p2 p3

(32)

p (u ) = u 3

u2

1 3 3 3 6 3 u 1 0 3 3 0 0 1

1 p 0 0 p1 0 p 2 0 p 3

(33)

If we let U = u3
P = [ p0

u2
p1

u 1
p2

p3 ]

(34) (35)

and 1 3 3 3 6 3 M = 3 3 0 0 0 1 1 0 0 0

(36)

Then we can write Equation 33 even more compactly as

p (u ) = UMP

(37)

Note that the composition of the matrices U, M, and P varies according to the number of control points (that is, the degree of the Bernstein polynomial basis functions).

339 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

Lecture No.39

Building Polygonal Models of Surfaces

Constructing polygonal approximations to surfaces is an art, and there is no substitute for experience. This section, however, lists a few pointers that might make it a bit easier to get started. Keep polygon orientations consistent. Make sure that when viewed from the outside, all the polygons on the surface are oriented in the same direction (all clockwise or all counterclockwise). Consistent orientation is important for polygon culling and two-sided lighting. Try to get this right the first time, since it's excruciatingly painful to fix the problem later. (If you use glScale*() to reflect geometry around some axis of symmetry, you might change the orientation with glFrontFace() to keep the orientations consistent.) When you subdivide a surface, watch out for any nontriangular polygons. The three vertices of a triangle are guaranteed to lie on a plane; any polygon with four or more vertices might not. Nonplanar polygons can be viewed from some orientation such that the edges cross each other, and OpenGL might not render such polygons correctly. There's always a trade-off between the display speed and the quality of the image. If you subdivide a surface into a small number of polygons, it renders quickly but might have a jagged appearance; if you subdivide it into millions of tiny polygons, it probably looks good but might take a long time to render. Ideally, you can provide a parameter to the subdivision routines that indicates how fine a subdivision you want, and if the object is farther from the eye, you can use a coarser subdivision. Also, when you subdivide, use large polygons where the surface is relatively flat, and small polygons in regions of high curvature. For high-quality images, it's a good idea to subdivide more on the silhouette edges than in the interior. If the surface is to be rotated relative to the eye, this is tougher to do, since the silhouette edges keep moving. Silhouette edges occur where the normal vectors are perpendicular to the vector from the surface to the viewpoint that is, when their vector dot product is zero. Your subdivision algorithm might choose to subdivide more if this dot product is near zero. Try to avoid T-intersections in your models (see Figure 1). As shown, there's no guarantee that the line segments AB and BC lie on exactly the same pixels as the segment AC. Sometimes they do, and sometimes they don't, depending on the transformations and orientation. This can cause cracks to appear intermittently in the surface.

Figure 1 : Modifying an Undesirable T-intersection

If you're constructing a closed surface, make sure to use exactly the same numbers for coordinates at the beginning and end of a closed loop, or you can get gaps and cracks due to numerical round-off. Here's a two-dimensional example of bad code: 340 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

#define PI 3.14159265 #define EDGES 30 /* draw a circle */ glBegin(GL_LINE_STRIP); for (i = 0; i <= EDGES; i++) glVertex2f(cos((2*PI*i)/EDGES), sin((2*PI*i)/EDGES)); glEnd(); The edges meet exactly only if your machine manages to calculate the sine and cosine of 0 and of (2*PI*EDGES/EDGES) and gets exactly the same values. If you trust the floating-point unit on your machine to do this right, the authors have a bridge they'd like to sell you.... To correct the code, make sure that when i == EDGES, you use 0 for the sine and cosine, not 2*PI*EDGES/EDGES. (Or simpler still, use GL_LINE_LOOP instead of GL_LINE_STRIP, and change the loop termination condition to i < EDGES.)

An Example: Building an Icosahedron

To illustrate some of the considerations that arise in approximating a surface, let's look at some example code sequences. This code concerns the vertices of a regular icosahedron (which is a Platonic solid composed of twenty faces that span twelve vertices, each face of which is an equilateral triangle). An icosahedron can be considered a rough approximation for a sphere. Example 1 defines the vertices and triangles making up an icosahedron and then draws the icosahedron.

Example 1 : Drawing an Icosahedron #define X .525731112119133606 #define Z .850650808352039932 static GLfloat vdata[12][3] = { {-X, 0.0, Z}, {X, 0.0, Z}, {-X, 0.0, -Z}, {X, 0.0, -Z}, {0.0, Z, X}, {0.0, Z, -X}, {0.0, -Z, X}, {0.0, -Z, -X}, {Z, X, 0.0}, {-Z, X, 0.0}, {Z, -X, 0.0}, {-Z, -X, 0.0} }; static GLuint tindices[20][3] = { {0,4,1}, {0,9,4}, {9,5,4}, {4,5,8}, {4,8,1}, {8,10,1}, {8,3,10}, {5,3,8}, {5,2,3}, {2,7,3}, {7,10,3}, {7,6,10}, {7,11,6}, {11,0,6}, {0,1,6}, {6,1,10}, {9,0,11}, {9,11,2}, {9,2,5}, {7,2,11} };
int i; glBegin(GL_TRIANGLES); for (i = 0; i < 20; i++) { /* color information here */ glVertex3fv(&vdata[tindices[i][0]][0]); glVertex3fv(&vdata[tindices[i][1]][0]); glVertex3fv(&vdata[tindices[i][2]][0]); } glEnd(); 341 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

The strange numbers X and Z are chosen so that the distance from the origin to any of the vertices of the icosahedron is 1.0. The coordinates of the twelve vertices are given in the array vdata[ ][ ], where the zeroth vertex is {- &Xgr; , 0.0, &Zgr; }, the first is {X, 0.0, Z}, and so on. The array tindices[ ][ ] tells how to link the vertices to make triangles. For example, the first triangle is made from the zeroth, fourth, and first vertex. If you take the vertices for triangles in the order given, all the triangles have the same orientation. The line that mentions color information should be replaced by a command that sets the color of the ith face. If no code appears here, all faces are drawn in the same color, and it'll be impossible to discern the three-dimensional quality of the object. An alternative to explicitly specifying colors is to define surface normals and use lighting, as described in the next subsection.

Note: In all the examples described in this section, unless the surface is to be drawn only once, you should probably save the calculated vertex and normal coordinates so that the calculations don't need to be repeated each time that the surface is drawn. This can be done using your own data structures or by constructing display lists. Calculating Normal Vectors for a Surface If a surface is to be lit, you need to supply the vector normal to the surface. Calculating the normalized cross product of two vectors on that surface provides normal vector. With the flat surfaces of an icosahedron, all three vertices defining a surface have the same normal vector. In this case, the normal needs to be specified only once for each set of three vertices. icosahedron. Example 2 : Generating Normal Vectors for a Surface
GLfloat d1[3], d2[3], norm[3]; for (j = 0; j < 3; j++) { d1[j] = vdata[tindices[i][0]][j] - vdata[tindices[i][1]][j]; d2[j] = vdata[tindices[i][1]][j] - vdata[tindices[i][2]][j]; } normcrossprod(d1, d2, norm); glNormal3fv(norm); The function normcrossprod() produces the normalized cross product of two vectors, as shown in Example 3.

Example 3 : Calculating the Normalized Cross Product of Two Vectors

void normalize(float v[3]) { GLfloat d = sqrt(v[0]*v[0]+v[1]*v[1]+v[2]*v[2]); if (d == 0.0) { error("zero length vector"); return; } v[0] /= d; v[1] /= d; v[2] /= d; } 342 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

void normcrossprod(float v1[3], float v2[3], float out[3]) { GLint i, j; GLfloat length; out[0] = v1[1]*v2[2] - v1[2]*v2[1]; out[1] = v1[2]*v2[0] - v1[0]*v2[2]; out[2] = v1[0]*v2[1] - v1[1]*v2[0]; normalize(out); } If you're using an icosahedron as an approximation for a shaded sphere, you'll want to use normal vectors that are perpendicular to the true surface of the sphere, rather than being perpendicular to the faces. For a sphere, the normal vectors are simple; each points in the same direction as the vector from the origin to the corresponding vertex. Since the icosahedron vertex data is for an icosahedron of radius 1, the normal and vertex data is identical. Here is the code that would draw an icosahedral approximation of a smoothly shaded sphere glBegin(GL_TRIANGLES); for (i = 0; i < 20; i++) { glNormal3fv(&vdata[tindices[i][0]][0]); glVertex3fv(&vdata[tindices[i][0]][0]); glNormal3fv(&vdata[tindices[i][1]][0]); glVertex3fv(&vdata[tindices[i][1]][0]); glNormal3fv(&vdata[tindices[i][2]][0]); glVertex3fv(&vdata[tindices[i][2]][0]); } glEnd();

Improving the Model A twenty-sided approximation to a sphere doesn't look good unless the image of the sphere on the screen is quite small, but there's an easy way to increase the accuracy of the approximation. Imagine the icosahedron inscribed in a sphere, and subdivide the triangles as shown in Figure 2. The newly introduced vertices lie slightly inside the sphere, so push them to the surface by normalizing them (dividing them by a factor to make them have length 1). This subdivision process can be repeated for arbitrary accuracy. The three objects shown in Figure 2 use 20, 80, and 320 approximating triangles, respectively.

Figure 2: Subdividing to Improve a Polygonal Approximation to a Surface

343 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

Example 4 performs a single subdivision, creating an 80-sided spherical approximation.

Example 4 : Single Subdivision

void drawtriangle(float *v1, float *v2, float *v3) { glBegin(GL_TRIANGLES); glNormal3fv(v1); glVertex3fv(v1); glNormal3fv(v2); glVertex3fv(v2); glNormal3fv(v3); glVertex3fv(v3); glEnd(); } void subdivide(float *v1, float *v2, float *v3) { GLfloat v12[3], v23[3], v31[3]; GLint i; for (i = 0; i < 3; i++) { v12[i] = v1[i]+v2[i]; v23[i] = v2[i]+v3[i]; v31[i] = v3[i]+v1[i]; } normalize(v12); normalize(v23); normalize(v31); drawtriangle(v1, v12, v31); drawtriangle(v2, v23, v12); drawtriangle(v3, v31, v23); drawtriangle(v12, v23, v31); } for (i = 0; i < 20; i++) { subdivide(&vdata[tindices[i][0]][0], &vdata[tindices[i][1]][0], &vdata[tindices[i][2]][0]); } Example 5 is a slight modification of Example 4 which recursively subdivides the triangles to the proper depth. If the depth value is 0, no subdivisions are performed, and the triangle is drawn as is. If the depth is 1, a single subdivision is performed, and so on.

Example 5 : Recursive Subdivision

void subdivide(float *v1, float *v2, float *v3, long depth) { GLfloat v12[3], v23[3], v31[3]; GLint i; if (depth == 0) { drawtriangle(v1, v2, v3); return; } 344 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

for (i = 0; i < 3; i++) { v12[i] = v1[i]+v2[i]; v23[i] = v2[i]+v3[i]; v31[i] = v3[i]+v1[i]; } normalize(v12); normalize(v23); normalize(v31); subdivide(v1, v12, v31, depth-1); subdivide(v2, v23, v12, depth-1); subdivide(v3, v31, v23, depth-1); subdivide(v12, v23, v31, depth-1); }

Generalized Subdivision A recursive subdivision technique such as the one described in Example 5 can be used for other types of surfaces. Typically, the recursion ends either if a certain depth is reached or if some condition on the curvature is satisfied (highly curved parts of surfaces look better with more subdivision). To look at a more general solution to the problem of subdivision, consider an arbitrary surface parameterized by two variables u[0] and u[1]. Suppose that two routines are provided:
void surf(GLfloat u[2], GLfloat vertex[3], GLfloat normal[3]); float curv(GLfloat u[2]); If surf() is passed u[], the corresponding three-dimensional vertex and normal vectors (of length 1) are returned. If u[] is passed to curv(), the curvature of the surface at that point is calculated and returned. (See an introductory textbook on differential geometry for more information about measuring surface curvature.) Example 6 shows the recursive routine that subdivides a triangle either until the maximum depth is reached or until the maximum curvature at the three vertices is less than some cutoff.

Example 6 : Generalized Subdivision

void subdivide(float u1[2], float u2[2], float u3[2], float cutoff, long depth) { GLfloat v1[3], v2[3], v3[3], n1[3], n2[3], n3[3]; GLfloat u12[2], u23[2], u32[2]; GLint i; if (depth == maxdepth || (curv(u1) < cutoff && curv(u2) < cutoff && curv(u3) < cutoff)) { surf(u1, v1, n1); surf(u2, v2, n2); surf(u3, v3, n3); glBegin(GL_POLYGON); glNormal3fv(n1); glVertex3fv(v1); glNormal3fv(n2); glVertex3fv(v2); glNormal3fv(n3); glVertex3fv(v3); glEnd(); 345 Copyright Virtual University of Pakistan

39-Building Polygonal Models of Surfaces

VU

return; } for (i = 0; i < 2; i++) { u12[i] = (u1[i] + u2[i])/2.0; u23[i] = (u2[i] + u3[i])/2.0; u31[i] = (u3[i] + u1[i])/2.0; } subdivide(u1, u12, u31, cutoff, depth+1); subdivide(u2, u23, u12, cutoff, depth+1); subdivide(u3, u31, u23, cutoff, depth+1); subdivide(u12, u23, u31, cutoff, depth+1); }

346 Copyright Virtual University of Pakistan

40-Fractals

VU

Lecture No.40

Fractals

Fractal are geometric patterns that is repeated at ever smaller scales to produce irregular shapes and surfaces that can not be represented by classical geometry. Fractals are used in computer modeling of irregular patterns and structure in nature. According to Webster's Dictionary a fractal is defined as being "derived from the Latin word fractus meaning broken, uneven: any of various extremely irregular curves or shapes that repeat themselves at any scale on which they are examined." Mandelbrot, the discoverer of fractals gives two definitions:

"I coined fractal from the Latin adjective fractus. The corresponding Latin verb frangere means 'to break:' to create irregular fragments. It is therefore sensible and how appropriate for our needs! - that, in addition to 'fragmented' (as in fraction or refraction), fractus should also mean 'irregular,' both meanings being preserved in fragment"[3] Every set with a non-integer (Hausdorff-Besicovitch) dimension (D) is a fractal. However not every fractal has an integer D. A fractal is by definition a set for which D strictly exceeds the topological dimension (D^).[3]

Hausdorff-Besicovitch(Fractal Dimension)
To understand the second definition we need to be able to understand the fractal dimension. So first we have to develop an understanding of how to calculate the dimension of an object. Below we have three different objects. 1. As you can see the line is broken into 4 smaller lines. Each of these lines is similar to the original line, but they are all 1/4 the scale. This is the idea of self similarity.

2. The square below is also broken into smaller pieces. Each of which is 1/4th the size of the original. In this case it takes 16 of the smaller pieces to create the original.

3. As with the others the cube is also broken down into smaller cubes of 1/4 the size of the original. It takes 64 of these smaller cubes to create the original cube.

347 Copyright Virtual University of Pakistan

40-Fractals

VU

By looking at this we begin to see a pattern: 4 = 4^1 16 = 4^2 64 = 4^3 This gives us the equation:

N = S^D
Where N is the number of small pieces that go into the larger one, S is the scale to which the smaller pieces compare to the larger one and D is the dimension. We now have the tools to be able to calculate the dimension. Just solve for D in the previous equation. When we do this we find that the Dimension is:

D = log N / log S This dimension is the Hausdorff-Besicovitch dimension. Koch Curve

Euclidean Geometry is the stuff we all learn in school. It is the geometry of lines, planes, circles etc. It's simple and it works, and for a long time, mathematicians thought it was a reasonable representation of nature. However, people soon discovered that they could draw (or at least begin to draw) certain curves and surfaces that could not be described by the classical geometry. How hard can it be to draw a curve? Let us attempt to describe. This is the Koch curve:

Draw a triangle. If we say that each line is of length 1, then the total length of the curve is 3.

348 Copyright Virtual University of Pakistan

40-Fractals

VU

Now take each edge in turn and add another triangle, a third of the size. So now there are 12 edges and 12 points. The length of the curve is now 4. Repeat the process again, and again, forever. Length = 5.3333

Length = 7.1111

349 Copyright Virtual University of Pakistan

40-Fractals

VU

Length = 12.6420 As we continue adding edges, the length of the curve increases. If we add edges forever, then length of the curve reaches infinity, but the whole curve nevertheless covers a finite area. The curve is infinitely detailed. No matter how closely we zoom into the image, it always shows up more detail.

Self Similarity
So what do these mathematical curiosities have to do with the real world? Well, everything as it turns out. Such objects turn up all the time in the natural world. Animals, plants, rocks, crystals and liquids all exhibit fractal properties and self similarity. Lets take a look at a common plant, the fern. The fern is typical of many plants in that it exhibits self similarity. A fern consists of a leaf, which is made up from many similar, but smaller leaves, each of which, in turn, is made from even smaller leaves. The closer we look the more detail we see. The following figure is a standard fern, which we may well find while being dragged on long walks in the country by your parents long before we are able to fully appreciate the beauty of nature. We will see the overall theme of repeating leaves. Each smaller leaf looks similar to the larger leaf.

350 Copyright Virtual University of Pakistan

40-Fractals

VU

Looking a little closer, we can see that those small leaves are made up from even smaller leaves.

Of course, in reality, a fern does have a smallest leaf, though were sure every fern aspires to be like that one. What is interesting it that the program to generate this image is only a few lines long. The same tends to be true for all fractals. A very simple algorithm can explain an infinitely complex object.

Fractal Geometry
Almost all geometric forms used for building man made objects belong to Euclidean geometry, they are comprised of lines, planes, rectangular volumes, arcs, cylinders, spheres, etc. These elements can be classified as belonging to an integer dimension, 1, 2, or 3. This concept of dimension can be described both intuitively and mathematically. Intuitively we say that a line is one dimensional because it only takes 1 number to uniquely define any point on it. That one number could be the distance from the start of the line. This applies equally well to the circumference of a circle, a curve, or the boundary of any object.

A plane is two dimensional since in order to uniquely define any point on its surface we require two numbers. There are many ways to arrange the definition of these two numbers but we normally create an orthogonal coordinate system. Other examples of two dimensional objects are the surface of a sphere or an arbitrary twisted plane.

The volume of some solid object is 3 dimensional on the same basis as above, it takes three numbers to uniquely define any point within the object.

351 Copyright Virtual University of Pakistan

40-Fractals

VU

A more mathematical description of dimension is based on how the "size" of an object behaves as the linear dimension increases. In one dimension consider a line segment. If the linear dimension of the line segment is doubled then obviously the length (characteristic size) of the line has doubled. In two dimensions, ff the linear dimensions of a rectangle for example is doubled then the characteristic size, the area, increases by a factor of 4. In three dimensions if the linear dimension of a box are doubled then the volume increases by a factor of 8. This relationship between dimension D, linear scaling L and the resulting increase in size S can be generalized and written as This is just telling us mathematically what we know from everyday experience. If we scale a two dimensional object for example then the area increases by the square of the scaling. If we scale a three dimensional object the volume increases by the cube of the scale factor. Rearranging the above gives an expression for dimension depending on how the size changes as a function of linear scaling, namely

In the examples above the value of D is an integer, 1, 2, or 3, depending on the dimension of the geometry. This relationship holds for all Euclidean shapes. There are however many shapes which do not conform to the integer based idea of dimension given above in both the intuitive and mathematical descriptions. That is, there are objects which appear to be curves for example but which a point on the curve cannot be uniquely described with just one number. If the earlier scaling formulation for dimension is applied the formula does not yield an integer. There are shapes that lie in a plane but if they are linearly scaled by a factor L, the area does not increase by L squared but by some non integer amount. These geometries are called fractals! One of the simpler fractal shapes is the von Koch snowflake. The method of creating this shape is to repeatedly replace each line segment with the following 4 line segments.

The process starts with a single line segment and continues for ever. The first few iterations of this procedure are shown below.

352 Copyright Virtual University of Pakistan

40-Fractals

VU

This demonstrates how a very simple generation rule for this shape can generate some unusual (fractal) properties. Unlike Euclidean shapes this object has detail at all levels. If one magnifies a Euclidean shape such as the circumference of a circle it becomes a different shape, namely a straight line. If we magnify this fractal more and more detail is uncovered, the detail is self similar or rather it is exactly self similar. Put another way, any magnified portion is identical to any other magnified portion. Note also that the "curve" on the right is not a fractal but only an approximation of one. This is no different from when one draws a circle, it is only an approximation to a perfect circle. At each iteration the length of the curve increases by a factor of 4/3. Thus the limiting curve is of infinite length and indeed the length between any two points of the curve is infinite. This curve manages to compress an infinite length into a finite area of the plane without intersecting itself! Considering the intuitive notion of 1 dimensional shapes, although this object appears to be a curve with one starting point and one end point, it is not possible to uniquely specify any position along the curve with one number as we expect to be able to do with Euclidean curves which are 1 dimensional. Although the method of creating this curve is straightforward, there is no algebraic formula the describes the points on the curve. Some of the major differences between fractal and Euclidean geometry are outlined in the following table.

Firstly the recognition of fractal is very modern, they have only formally been studied in the last 10 years compared to Euclidean geometry which goes back over 2000 years. 353 Copyright Virtual University of Pakistan

40-Fractals

VU

Secondly whereas Euclidean shapes normally have a few characteristic sizes or length scales (e.g. the radius of a circle or the length of of a side of a cube) fractals have so characteristic sizes. Fractal shapes are self similar and independent of size or scaling. Third, Euclidean geometry provides a good description of man made objects whereas fractals are required for a representation of naturally occurring geometries. It is likely that this limitation of our traditional language of shape is responsible for the striking difference between mass produced objects and natural shapes. Finally, Euclidean geometries are defined by algebraic formulae, for example defines a sphere. Fractals are normally the result of a iterative or recursive construction or algorithm.

L-Systems
The following is based on L-Systems as described in "Lecture Notes in Biomathematics" by Przemyslaw Prusinkiewcz and James Hanan. A brief description of an 0L system will be presented here but for a more complete description the user should consult the literature.

Simpleminded example of 0L system

A string of characters (symbols) is rewritten on each iteration according to some replacement rules. Consider an initial string (axiom) F+F+F+F and a rewriting rule F --> F+F-F-FF+F+F-F After one iteration the following string would result F+F-F-FF+F+F-F + F+F-F-FF+F+F-F + F+F-F-FF+F+F-F + F+F-F-FF+F+F-F For the next iteration the same rule is applied but now to the string resulting from the last iteration F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-FF+ FF-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F+ F+ F-FFF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-FF+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ FF+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-FF+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F+ F+ F-FFF+ F+ F-F-F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-FF+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F+ F+ F-F-FF+ F+ F-F-F+ F-F-FF+ F+ F-F Some symbols are now given a graphical meaning, for example, F means move forward drawing a line, + means turn right by some predefined angle (90 degrees in this case), means turn left. Using these symbols the initial string F+F+F+F is just a rectangle ( = 90). The replacement rule F --> F+F-F-FF+F+F-F replaces each forward movement by the following figure

354 Copyright Virtual University of Pakistan

40-Fractals

VU

The first iteration interpreted graphically is

The next iteration interpreted graphically is:

The following characters have a geometric interpretation.

355 Copyright Virtual University of Pakistan

40-Fractals

VU

Recent usage of L-Systems is for the creation of realistic looking objects that occur in nature and in particular the branching structure of plants. One of the important characteristics of L systems is that only a small amount of information is required to represent very complex objects. So while the bushes in figure 9 contain many thousands of lines they can be described in a database by only a few bytes of data, the actual bushes are only "grown" when required for visual presentation. Using suitably designed LSystem algorithms it is possible to design the L-System production rules that will create a particular class of plant. Further examples:

356 Copyright Virtual University of Pakistan

40-Fractals

VU

357 Copyright Virtual University of Pakistan

40-Fractals

VU

358 Copyright Virtual University of Pakistan

40-Fractals

VU

359 Copyright Virtual University of Pakistan

40-Fractals

VU

Featured on the cover of the HPC (High Performance Computing) magazine, 3 August 2001. 360 Copyright Virtual University of Pakistan

40-Fractals

VU

IFS - Iterated Function Systems

Instead of working with lines as in L systems, IFS replaces polygons by other polygons as described by a generator. On every iteration each polygon is replaced by a suitably scaled, rotated, and translated version of the polygons in the generator. Figure 10 shows two such generators made of rectangles and the result after one and six iterations. From this geometric description it is also possible to derive a hop-along description which gives the image that would be created after iterating the geometric model to infinity. The description of this is a set of contractive transformations on a plane of the form

each with an assigned probability. To run the system an initial point is chosen and on each iteration one of the transformation is chosen randomly according to the assigned probabilities, the resulting points (xn, yn) are drawn on the page. As in the case of L systems, if the IFS code for a desired image can be determined (by something called the Collage theorem) then large data compression ratios can be achieved. Instead of storing the geometry of the very complex object just the IFS generator needs to be stored and the image can be generated when required. The fundamental iterative process involves replacing rectangles with a series of rectangles called the generator. The rectangles are replaced by a suitably scaled, translated, and rotated version of the generator. For example consider the generator on the right It consists of three rectangles, each with its own center, dimensions and rotation angle. The initial conditions usually consist of a single square, the first iteration then consists of replacing this square by a suitably positioned, scaled and rotated version of the generator. The next iteration involves replacing each of the rectangles in the current system by suitably positioned, scaled, and rotated versions of the generator resulting in the following

361 Copyright Virtual University of Pakistan

40-Fractals

VU

The next iteration replaces each rectangle above again by the initial generator as shown

and so on, three more iterations later

Hop-along or "The Chaos Game"

A technique exists by which the resulting form after an infinite number of iterations can be derived. This is a function of the form

This gives a series of (x,y) points all which lie on the result of an infinite IFS. Although it still takes an infinite number of terms in this series to form the result the appearance can be readily appreciated after a modest number of terms (10000 say). Note that with both methods it is possible to create the image at any scale. In many but not all cases zoomed in examples will be exhibit self similarity at all scales. Applications generally iinvolve data reduction for model files. If a generator can be found for a complex image then storing the generator and the rules of production results in a great deal of data reduction. For example the weed in the examples above might eventually 362 Copyright Virtual University of Pakistan

40-Fractals

VU

contain over 2000 rectangles but is completely specified by the characteristics of 3 rectangles, only 5 numbers, center (cx,cy), scale (sx,sy), and angle () Note: it is not necessarily trivial to derive a rectangular generator for an arbitrary form, although it is possible to create a polygonal generator for any form. Further examples

363 Copyright Virtual University of Pakistan

40-Fractals

VU

364 Copyright Virtual University of Pakistan

40-Fractals

VU

IFS Fern

Random IFS

Wada basins 365 Copyright Virtual University of Pakistan

41-Viewing

VU

Lecture No.41

Viewing

In this lecture we will practically implement viewing a geometric model in any orientation by transforming it in three-dimensional space and control the location in threedimensional space from which the model is viewed and to Clip undesired portions of the model out of the scene that's to be viewed, manipulate the appropriate matrix stacks that control model transformation for viewing and project the model onto the screen, combine multiple transformations to mimic sophisticated systems in motion, such as a solar system or an articulated robot arm, Reverse or mimic the operations of the geometric processing pipeline l and we will also discuss on how to instruct OpenGL to draw the geometric models we want displayed in our scene. Now we must decide how we want to position the models in the scene, and we must choose a vantage point from which to view the scene. We can use the default positioning and vantage point, but most likely we want to specify them. Look at the image on the cover of this book. The program that produced that image contained a single geometric description of a building block. Each block was carefully positioned in the scene: Some blocks were scattered on the floor, some were stacked on top of each other on the table, and some were assembled to make the globe. Also, a particular viewpoint had to be chosen. Obviously, we wanted to look at the corner of the room containing the globe. But how far away from the scene - and where exactly should the viewer be? We wanted to make sure that the final image of the scene contained a good view out the window, that a portion of the floor was visible, and that all the objects in the scene were not only visible but presented in an interesting arrangement. and how to use OpenGL to accomplish these tasks: how to position and orient models in three-dimensional space and how to establish the location - also in three-dimensional space - of the viewpoint. All of these factors help determine exactly what image appears on the screen. We want to remember that the point of computer graphics is to create a two-dimensional image of three-dimensional objects (it has to be two-dimensional because it's drawn on a flat screen), but we need to think in three-dimensional coordinates while making many of the decisions that determine what gets drawn on the screen. A common mistake people make when creating three-dimensional graphics is to start thinking too soon that the final image appears on a flat, two-dimensional screen. Avoid thinking about which pixels need to be drawn, and instead try to visualize threedimensional space. Create your models in some three-dimensional universe that lies deep inside your computer, and let the computer do its job of calculating which pixels to color. A series of three computer operations convert an object's three-dimensional coordinates to pixel positions on the screen. Transformations, which are represented by matrix multiplication, include modeling, viewing, and projection operations. Such operations include rotation, translation, scaling, reflecting, orthographic projection, and perspective projection. Generally, we use a combination of several transformations to draw a scene. Since the scene is rendered on a rectangular window, objects (or parts of objects) that lie outside the window must be clipped. In three-dimensional Computer graphics, clipping occurs by throwing out objects on one side of a clipping plane. Finally, a correspondence must be established between the transformed coordinates and screen pixels. This is known as a viewport transformation.

366 Copyright Virtual University of Pakistan

41-Viewing

VU

Overview: The Camera Analogy The transformation process to produce the desired scene for viewing is analogous to taking a photograph with a camera. As shown in Figure 1, the steps with a camera (or a computer) might be the following. Set up your tripod and pointing the camera at the scene (viewing transformation). Arrange the scene to be photographed into the desired composition (modeling transformation). Choose a camera lens or adjust the zoom (projection transformation). Determine how large we want the final photograph to be - for example, we might want it enlarged (viewport transformation). After these steps are performed, the picture can be snapped or the scene can be drawn.

367 Copyright Virtual University of Pakistan

41-Viewing

VU

Figure 1: The Camera Analogy Note that these steps correspond to the order in which we specify the desired transformations in our program, not necessarily the order in which the relevant mathematical operations are performed on an object's vertices. The viewing transformations must precede the modeling transformations in our code, but we can specify the projection and viewport transformations at any point before drawing occurs. Figure 2 shows the order in which these operations occur on our computer.

Figure 2: Stages of Vertex Transformation To specify viewing, modeling, and projection transformations, we construct a 4 4 matrix M, which is then multiplied by the coordinates of each vertex v in the scene to accomplish the transformation v'=Mv (Remember that vertices always have four coordinates (x, y, z, w), though in most cases w is 1 and for two-dimensional data z is 0.) Note that viewing and modeling transformations are automatically applied to surface normal vectors, in addition to vertices. (Normal vectors are used only in eye coordinates.) This ensures that the normal vector's relationship to the vertex data is properly preserved. The viewing and modeling transformations we specify are combined to form the modelview matrix, which is applied to the incoming object coordinates to yield eye coordinates. Next, if we've specified additional clipping planes to remove certain objects from the scene or to provide cutaway views of objects, these clipping planes are applied. After that, OpenGL applies the projection matrix to yield clip coordinates. This transformation defines a viewing volume; objects outside this volume are clipped so that they're not drawn in the final scene. After this point, the perspective division is performed by dividing coordinate values by w, to produce normalized device coordinates. Finally, the transformed coordinates are converted to window coordinates by applying the viewport transformation. We can manipulate the dimensions of the viewport to cause the final image to be enlarged, shrunk, or stretched. We might correctly suppose that the x and y coordinates are sufficient to determine which pixels need to be drawn on the screen. However, all the transformations are performed on the z coordinates as well. This way, at the end of this transformation process, the z values correctly reflect the depth of a given 368 Copyright Virtual University of Pakistan

41-Viewing

VU

vertex (measured in distance away from the screen). One use for this depth value is to eliminate unnecessary drawing. For example, suppose two vertices have the same x and y values but different z values. OpenGL can use this information to determine which surfaces are obscured by other surfaces and can then avoid drawing the hidden surfaces. As we've probably guessed by now, we need to know a few things about matrix mathematics to get the most out of this lecture as we have learnt from previous lectures. A Simple Example: Drawing a Cube Example 1 draws a cube that's scaled by a modeling transformation (see Figure 3). The viewing transformation, gluLookAt(), positions and aims the camera towards where the cube is drawn. A projection transformation and a viewport transformation are also specified. The rest of this section walks us through Example 1 and briefly explains the transformation commands it uses. The succeeding sections contain the complete, detailed discussion of all OpenGL's transformation commands.

Figure 3: Transformed Cube Example 1 : Transformed Cube #include <GL/gl.h> #include <GL/glu.h> #include <GL/glut.h> void init(void) { glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_FLAT); }

void display(void) { glClear (GL_COLOR_BUFFER_BIT); glColor3f (1.0, 1.0, 1.0); glLoadIdentity (); /* clear the matrix */ /* viewing transformation */ gluLookAt (0.0, 0.0, 5.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0); glScalef (1.0, 2.0, 1.0); /* modeling transformation */ glutWireCube (1.0); glFlush (); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); 369 Copyright Virtual University of Pakistan

41-Viewing

VU

glLoadIdentity (); glFrustum (-1.0, 1.0, -1.0, 1.0, 1.5, 20.0); glMatrixMode (GL_MODELVIEW); } int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_SINGLE | GLUT_RGB); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutMainLoop(); return 0; } The Viewing Transformation Recall that the viewing transformation is analogous to positioning and aiming a camera. In this code example, before the viewing transformation can be specified, the current matrix is set to the identity matrix with glLoadIdentity(). This step is necessary since most of the transformation commands multiply the current matrix by the specified matrix and then set the result to be the current matrix. If we don't clear the current matrix by loading it with the identity matrix, we continue to combine previous transformation matrices with the new one we supply. In some cases, we do want to perform such combinations, but we also need to clear the matrix sometimes. In Example 1, after the matrix is initialized, the viewing transformation is specified with gluLookAt(). The arguments for this command indicate where the camera (or eye position) is placed, where it is aimed, and which way is up. The arguments used here place the camera at (0, 0, 5), aim the camera lens towards (0, 0, 0), and specify the up-vector as (0, 1, 0). The up-vector defines a unique orientation for the camera. If gluLookAt() was not called, the camera has a default position and orientation. By default, the camera is situated at the origin, points down the negative z-axis, and has an up-vector of (0, 1, 0). So in Example 1, the overall effect is that gluLookAt() moves the camera 5 units along the z-axis. The Modeling Transformation We use the modeling transformation to position and orient the model. For example, we can rotate, translate, or scale the model - or perform some combination of these operations. In Example 1, glScalef() is the modeling transformation that is used. The arguments for this command specify how scaling should occur along the three axes. If all the arguments are 1.0, this command has no effect. In Example 1, the cube is drawn twice as large in the y direction. Thus, if one corner of the cube had originally been at (3.0, 3.0, 3.0), that corner would wind up being drawn at (3.0, 6.0, 3.0). The effect of this modeling transformation is to transform the cube so that it isn't a cube but a rectangular box.

370 Copyright Virtual University of Pakistan

41-Viewing

VU

Try This Change the gluLookAt() call in Example 1 to the modeling transformation glTranslatef() with parameters (0.0, 0.0, -5.0). The result should look exactly the same as when we used gluLookAt(). Why are the effects of these two commands similar? Note that instead of moving the camera (with a viewing transformation) so that the cube could be viewed, we could have moved the cube away from the camera (with a modeling transformation). This duality in the nature of viewing and modeling transformations is why we need to think about the effect of both types of transformations simultaneously. It doesn't make sense to try to separate the effects, but sometimes it's easier to think about them one way rather than the other. This is also why modeling and viewing transformations are combined into the modelview matrix before the transformations are applied. Also note that the modeling and viewing transformations are included in the display() routine, along with the call that's used to draw the cube, glutWireCube(). This way, display() can be used repeatedly to draw the contents of the window if, for example, the window is moved or uncovered, and we've ensured that each time, the cube is drawn in the desired way, with the appropriate transformations. The potential repeated use of display() underscores the need to load the identity matrix before performing the viewing and modeling transformations, especially when other transformations might be performed between calls to display(). The Projection Transformation Specifying the projection transformation is like choosing a lens for a camera. We can think of this transformation as determining what the field of view or viewing volume is and therefore what objects are inside it and to some extent how they look. This is equivalent to choosing among wide-angle, normal, and telephoto lenses, for example. With a wide-angle lens, we can include a wider scene in the final photograph than with a telephoto lens, but a telephoto lens allows us to photograph objects as though they're closer to us than they actually are. In computer graphics, we don't have to pay $10,000 for a 2000-millimeter telephoto lens; once we've bought our graphics workstation, all we need to do is use a smaller number for our field of view. In addition to the field-of-view considerations, the projection transformation determines how objects are projected onto the screen, as its name suggests. Two basic types of projections are provided for us by OpenGL, along with several corresponding commands for describing the relevant parameters in different ways. One type is the perspective projection, which matches how we see things in daily life. Perspective makes objects that are farther away appear smaller; for example, it makes railroad tracks appear to converge in the distance. If we're trying to make realistic pictures, we'll want to choose perspective projection, which is specified with the glFrustum() command in this code example. The other type of projection is orthographic, which maps objects directly onto the screen without affecting their relative size. Orthographic projection is used in architectural and computer-aided design applications where the final image needs to reflect the measurements of objects rather than how they might look. Architects create perspective drawings to show how particular buildings or interior spaces look when viewed from various vantage points; the need for orthographic projection arises when blueprint plans or elevations are generated, which are used in the construction of buildings. Before glFrustum() can be called to set 371 Copyright Virtual University of Pakistan

41-Viewing

VU

the projection transformation, some preparation needs to happen. As shown in the reshape() routine in Example 1, the command called glMatrixMode() is used first, with the argument GL_PROJECTION. This indicates that the current matrix specifies the projection transformation; the following transformation calls then affect the projection matrix. As we can see, a few lines later glMatrixMode() is called again, this time with GL_MODELVIEW as the argument. This indicates that succeeding transformations now affect the modelview matrix instead of the projection matrix. Note that glLoadIdentity() is used to initialize the current projection matrix so that only the specified projection transformation has an effect. Now glFrustum() can be called, with arguments that define the parameters of the projection transformation. In this example, both the projection transformation and the viewport transformation are contained in the reshape() routine, which is called when the window is first created and whenever the window is moved or reshaped. This makes sense, since both projecting (the width to height aspect ratio of the projection viewing volume) and applying the viewport relate directly to the screen, and specifically to the size or aspect ratio of the window on the screen. Try This Change the glFrustum() call in Example 1 to the more commonly used Utility Library routine gluPerspective() with parameters (60.0, 1.0, 1.5, 20.0). Then experiment with different values, especially for fov(field of view ), near and far plane. Viewing and Modeling Transformations Viewing and modeling transformations are inextricably related in OpenGL and are in fact combined into a single modelview matrix. (See "A Simple Example: Drawing a Cube.") One of the toughest problems newcomers to computer graphics face is understanding the effects of combined three-dimensional transformations. As we've already seen, there are alternative ways to think about transformations - do we want to move the camera in one direction, or move the object in the opposite direction? Each way of thinking about transformations has advantages and disadvantages, but in some cases one way more naturally matches the effect of the intended transformation. If we can find a natural approach for wer particular application, it's easier to visualize the necessary transformations and then write the corresponding code to specify the matrix manipulations. The first part of this section discusses how to think about transformations; later, specific commands are presented. For now, we use only the matrix-manipulation commands we've already seen. Finally, keep in mind that we must call glMatrixMode() with GL_MODELVIEW as its argument prior to performing modeling or viewing transformations. Thinking about Transformations Let's start with a simple case of two transformations: a 45-degree counterclockwise rotation about the origin around the z-axis, and a translation down the x-axis. Suppose that the object we're drawing is small compared to the translation (so that we can see the effect of the translation), and that it's originally located at the origin. If we rotate the object first and then translate it, the rotated object appears on the x-axis. If we translate it down the x-axis first, however, and then rotate about the origin, the object is on the line y=x, as shown in Figure 4. In general, the order of transformations is critical. If we do transformation A and then transformation B, we almost always get something different than if we do them in the opposite order. 372 Copyright Virtual University of Pakistan

41-Viewing

VU

Figure 4 : Rotating First or Translating First

Now let's talk about the order in which we specify a series of transformations. All viewing and modeling transformations are represented as 4 4 matrices. Each successive glMultMatrix*() or transformation command multiplies a new 4 4 matrix M by the current modelview matrix C to yield CM. Finally, vertices v are multiplied by the current modelview matrix. This process means that the last transformation command called in our program is actually the first one applied to the vertices: CMv. Thus, one way of looking at it is to say that we have to specify the matrices in the reverse order. Like many other things, however, once we've gotten used to thinking about this correctly, backward will seem like forward. Consider the following code sequence, which draws a single point using three transformations: glMatrixMode(GL_MODELVIEW); glLoadIdentity(); glMultMatrixf(N); /* apply transformation N */ glMultMatrixf(M); /* apply transformation M */ glMultMatrixf(L); /* apply transformation L */ glBegin(GL_POINTS); glVertex3f(v); /* draw transformed vertex v */ glEnd(); With this code, the modelview matrix successively contains I, N, NM, and finally NML, where I represents the identity matrix. The transformed vertex is NMLv. Thus, the vertex transformation is N(M(Lv)) - that is, v is multiplied first by L, the resulting Lv is multiplied by M, and the resulting MLv is multiplied by N. Notice that the transformations to vertex v effectively occur in the opposite order than they were specified. (Actually, only a single multiplication of a vertex by the modelview matrix occurs; in this example, the N, M, and L matrices are already multiplied into a single matrix before it's applied to v.)

373 Copyright Virtual University of Pakistan

41-Viewing

VU

Grand, Fixed Coordinate System Thus, if we like to think in terms of a grand, fixed coordinate system - in which matrix multiplications affect the position, orientation, and scaling of our model - we have to think of the multiplications as occurring in the opposite order from how they appear in the code. Using the simple example shown on the left side of Figure 4 (a rotation about the origin and a translation along the x-axis), if we want the object to appear on the axis after the operations, the rotation must occur first, followed by the translation. To do this, we'll need to reverse the order of operations, so the code looks something like this (where R is the rotation matrix and T is the translation matrix): glMatrixMode(GL_MODELVIEW); glLoadIdentity(); glMultMatrixf(T); /* translation */ glMultMatrixf(R); /* rotation */ draw_the_object(); Moving a Local Coordinate System Another way to view matrix multiplications is to forget about a grand, fixed coordinate system in which our model is transformed and instead imagine that a local coordinate system is tied to the object were drawing. All operations occur relative to this changing coordinate system. With this approach, the matrix multiplications now appear in the natural order in the code. (Regardless of which analogy we're using, the code is the same, but how we think about it differs.) To see this in the translation-rotation example, begin by visualizing the object with a coordinate system tied to it. The translation operation moves the object and its coordinate system down the x-axis. Then, the rotation occurs about the (now-translated) origin, so the object rotates in place in its position on the axis. This approach is what we should use for applications such as articulated robot arms, where there are joints at the shoulder, elbow, and wrist, and on each of the fingers. To figure out where the tips of the fingers go relative to the body, we'd like to start at the shoulder, go down to the wrist, and so on, applying the appropriate rotations and translations at each joint. Thinking about it in reverse would be far more confusing. This second approach can be problematic, however, in cases where scaling occurs, and especially so when the scaling is non-uniform (scaling different amounts along the different axes). After uniform scaling, translations move a vertex by a multiple of what they did before, since the coordinate system is stretched. Non-uniform scaling mixed with rotations may make the axes of the local coordinate system non-perpendicular. As mentioned earlier, we normally issue viewing transformation commands in our program before any modeling transformations. This way, a vertex in a model is first transformed into the desired orientation and then transformed by the viewing operation. Since the matrix multiplications must be specified in reverse order, the viewing commands need to come first. Note, however, that we don't need to specify either viewing or modeling transformations if we're satisfied with the default conditions. If there's no viewing transformation, the "camera" is left in the default position at the origin, pointed toward the negative z-axis; if there's no modeling transformation, the model isn't moved, and it retains its specified position, orientation, and size. Since the commands for performing modeling transformations can be used to perform viewing transformations, 374 Copyright Virtual University of Pakistan

41-Viewing

VU

modeling transformations are discussed first, even if viewing transformations are actually issued first. This order for discussion also matches the way many programmers think when planning their code: Often, they write all the code necessary to compose the scene, which involves transformations to position and orient objects correctly relative to each other. Next, they decide where they want the viewpoint to be relative to the scene they've composed, and then they write the viewing transformations accordingly. Modeling Transformations The three OpenGL routines for modeling transformations are glTranslate*(), glRotate*(), and glScale*(). As we might suspect, these routines transform an object (or coordinate system, if we're thinking of it that way) by moving, rotating, stretching, shrinking, or reflecting it. All three commands are equivalent to producing an appropriate translation, rotation, or scaling matrix, and then calling glMultMatrix*() with that matrix as the argument. However, these three routines might be faster than using glMultMatrix*(). OpenGL automatically computes the matrices for we. In the command summaries that follow, each matrix multiplication is described in terms of what it does to the vertices of a geometric object using the fixed coordinate system approach, and in terms of what it does to the local coordinate system that's attached to an object. Translate void glTranslate{fd}(TYPEx, TYPE y, TYPEz); Multiplies the current matrix by a matrix that moves (translates) an object by the given x, y, and z values (or moves the local coordinate system by the same amounts). Figure 5 shows the effect of glTranslate*().

Figure 5 : Translating an Object Note that using (0.0, 0.0, 0.0) as the argument for glTranslate*() is the identity operation - that is, it has no effect on an object or its local coordinate system. Rotate void glRotate{fd}(TYPE angle, TYPE x, TYPE y, TYPE z);

375 Copyright Virtual University of Pakistan

41-Viewing

VU

Multiplies the current matrix by a matrix that rotates an object (or the local coordinate system) in a counterclockwise direction about the ray from the origin through the point (x, y, z). The angle parameter specifies the angle of rotation in degrees. The effect of glRotatef(45.0, 0.0, 0.0, 1.0), which is a rotation of 45 degrees about the zaxis, is shown in Figure 6.

Figure 6 : Rotating an Object Note that an object that lies farther from the axis of rotation is more dramatically rotated (has a larger orbit) than an object drawn near the axis. Also, if the angle argument is zero, the glRotate*() command has no effect. Scale void glScale{fd}(TYPEx, TYPE y, TYPEz); Multiplies the current matrix by a matrix that stretches, shrinks, or reflects an object along the axes. Each x, y, and z coordinate of every point in the object is multiplied by the corresponding argument x, y, or z. With the local coordinate system approach, the local coordinate axes are stretched, shrunk, or reflected by the x, y, and z factors, and the associated object is transformed with them. Figure 7 shows the effect of glScalef(2.0, -0.5, 1.0).

376 Copyright Virtual University of Pakistan

41-Viewing

VU

Figure 7 : Scaling and Reflecting an Object glScale*() is the only one of the three modeling transformations that changes the apparent size of an object: Scaling with values greater than 1.0 stretches an object, and using values less than 1.0 shrinks it. Scaling with a -1.0 value reflects an object across an axis. The identity values for scaling are (1.0, 1.0, 1.0). In general, we should limit our use of glScale*() to those cases where it is necessary. Using glScale*() decreases the performance of lighting calculations, because the normal vectors have to be renormalized after transformation. Note: A scale value of zero collapses all object coordinates along that axis to zero. It's usually not a good idea to do this, because such an operation cannot be undone. Mathematically speaking, the matrix cannot be inverted, and inverse matrices are required for certain lighting operations. Sometimes collapsing coordinates does make sense, however; the calculation of shadows on a planar surface is a typical application. In general, if a coordinate system is to be collapsed, the projection matrix should be used rather than the modelview matrix. A Modeling Transformation Code Example Example 2 is a portion of a program that renders a triangle four times, as shown in Figure 8. These are the four transformed triangles. A solid wireframe triangle is drawn with no modeling transformation. The same triangle is drawn again, but with a dashed line stipple and translated (to the left - along the negative x-axis). A triangle is drawn with a long dashed line stipple, with its height (y-axis) halved and its width (x-axis) increased by 50%. A rotated triangle, made of dotted lines, is drawn.

Figure 8 : Modeling Transformation Example Example 2 : Using Modeling Transformations: glLoadIdentity(); glColor3f(1.0, 1.0, 1.0); draw_triangle(); /* solid lines */ glEnable(GL_LINE_STIPPLE); /* dashed lines */ glLineStipple(1, 0xF0F0); glLoadIdentity(); glTranslatef(-20.0, 0.0, 0.0); draw_triangle(); glLineStipple(1, 0xF00F); /*long dashed lines */ 377 Copyright Virtual University of Pakistan

41-Viewing

VU

glLoadIdentity(); glScalef(1.5, 0.5, 1.0); draw_triangle(); glLineStipple(1, 0x8888); /* dotted lines */ glLoadIdentity(); glRotatef (90.0, 0.0, 0.0, 1.0); draw_triangle (); glDisable (GL_LINE_STIPPLE); Note the use of glLoadIdentity() to isolate the effects of modeling transformations; initializing the matrix values prevents successive transformations from having a cumulative effect. Even though using glLoadIdentity() repeatedly has the desired effect, it may be inefficient, because we may have to re specify viewing or modeling transformations. Note: Sometimes, programmers who want a continuously rotating object attempt to achieve this by repeatedly applying a rotation matrix that has small values. The problem with this technique is that because of round-off errors, the product of thousands of tiny rotations gradually drifts away from the value we really want (it might even become something that isn't a rotation). Instead of using this technique, increment the angle and issue a new rotation command with the new angle at each update step. Viewing Transformations A viewing transformation changes the position and orientation of the viewpoint. If we recall the camera analogy, the viewing transformation positions the camera tripod, pointing the camera toward the model. Just as we move the camera to some position and rotate it until it points in the desired direction, viewing transformations are generally composed of translations and rotations. Also remember that to achieve a certain scene composition in the final image or photograph, we can either move the camera or move all the objects in the opposite direction. Thus, a modeling transformation that rotates an object counterclockwise is equivalent to a viewing transformation that rotates the camera clockwise, for example. Finally, keep in mind that the viewing transformation commands must be called before any modeling transformations are performed, so that the modeling transformations take effect on the objects first. We can manufacture a viewing transformation in any of several ways, as described next. we can also choose to use the default location and orientation of the viewpoint, which is at the origin, looking down the negative z-axis. Use one or more modeling transformation commands (that is, glTranslate*() and glRotate*()). We can think of the effect of these transformations as moving the camera position or as moving all the objects in the world, relative to a stationary camera. Use the Utility Library routine gluLookAt() to define a line of sight. This routine encapsulates a series of rotation and translation commands. Create our own utility routine that encapsulates rotations and translations. Some applications might require custom routines that allow we to specify the viewing transformation in a convenient way. For example, we might want to specify the roll, 378 Copyright Virtual University of Pakistan

41-Viewing

VU

pitch, and heading rotation angles of a plane in flight, or we might want to specify a transformation in terms of polar coordinates for a camera that's orbiting around an object. Using glTranslate*() and glRotate*() When we use modeling transformation commands to emulate viewing transformations, we're trying to move the viewpoint in a desired way while keeping the objects in the world stationary. Since the viewpoint is initially located at the origin and since objects are often most easily constructed there as well (see Figure 9), in general we have to perform some transformation so that the objects can be viewed. Note that, as shown in the figure, the camera initially points down the negative z-axis. (we're seeing the back of the camera.)

Figure 9 : Object and Viewpoint at the Origin In the simplest case, we can move the viewpoint backward, away from the objects; this has the same effect as moving the objects forward, or away from the viewpoint. Remember that by default forward is down the negative z-axis; if we rotate the viewpoint, forward has a different meaning. So, to put 5 units of distance between the viewpoint and the objects by moving the viewpoint, as shown in Figure 10, use glTranslatef(0.0, 0.0, 5.0); This routine moves the objects in the scene -5 units along the z axis. This is also equivalent to moving the camera +5 units along the z axis.

379 Copyright Virtual University of Pakistan

41-Viewing

VU

Figure 10 : Separating the Viewpoint and the Object Now suppose we want to view the objects from the side. Should we issue a rotate command before or after the translate command? If we're thinking in terms of a grand, fixed coordinate system, first imagine both the object and the camera at the origin. We could rotate the object first and then move it away from the camera so that the desired side is visible. Since we know that with the fixed coordinate system approach, commands have to be issued in the opposite order in which they should take effect, we know that we need to write the translate command first in our code and follow it with the rotate command. Now let's use the local coordinate system approach. In this case, think about moving the object and its local coordinate system away from the origin; then, the rotate command is carried out using the now-translated coordinate system. With this approach, commands are issued in the order in which they're applied, so once again the translate command comes first. Thus, the sequence of transformation commands to produce the desired result is glTranslatef(0.0, 0.0, -5.0); glRotatef(90.0, 0.0, 1.0, 0.0); If we're having trouble keeping track of the effect of successive matrix multiplications, try using both the fixed and local coordinate system approaches and see whether one makes more sense to us. Note that with the fixed coordinate system, rotations always occur about the grand origin, whereas with the local coordinate system, rotations occur about the origin of the local system. We might also try using the gluLookAt() utility routine described in the next section. Using the gluLookAt() Utility Routine Often, programmers construct a scene around the origin or some other convenient location, then they want to look at it from an arbitrary point to get a good view of it. As its name suggests, the gluLookAt() utility routine is designed for just this purpose. It takes three sets of arguments, which specify the location of the viewpoint, define a reference point toward which the camera is aimed, and indicate which direction is up. 380 Copyright Virtual University of Pakistan

41-Viewing

VU

Choose the viewpoint to yield the desired view of the scene. The reference point is typically somewhere in the middle of the scene. (If we've built our scene at the origin, the reference point is probably the origin.) It might be a little trickier to specify the correct up-vector. Again, if we've built some real-world scene at or around the origin and if we've been taking the positive y-axis to point upward, then that's our up-vector for gluLookAt(). However, if we're designing a flight simulator, up is the direction perpendicular to the plane's wings, from the plane toward the sky when the plane is rightside up on the ground. The gluLookAt() routine is particularly useful when we want to pan across a landscape, for instance. With a viewing volume that's symmetric in both x and y, the (eyex, eyey, eyez) point specified is always in the center of the image on the screen, so we can use a series of commands to move this point slightly, thereby panning across the scene. void gluLookAt(GLdouble eyex, GLdouble eyey, GLdouble eyez, GLdouble centerx, GLdouble centery, GLdouble centerz, GLdouble upx, GLdouble upy, GLdouble upz); Defines a viewing matrix and multiplies it to the right of the current matrix. The desired viewpoint is specified by eyex, eyey, and eyez. The centerx, centery, and centerz arguments specify any point along the desired line of sight, but typically they're some point in the center of the scene being looked at. The upx, upy, and upz arguments indicate which direction is up (that is, the direction from the bottom to the top of the viewing volume) In the default position, the camera is at the origin, is looking down the negative z-axis, and has the positive y-axis as straight up. This is the same as calling gluLookat (0.0, 0.0, 0.0, 0.0, 0.0, -100.0, 0.0, 1.0, 0.0); The z value of the reference point is -100.0, but could be any negative z, because the line of sight will remain the same. In this case, we don't actually want to call gluLookAt(), because this is the default and we are already there! (The lines extending from the camera represent the viewing volume, which indicates its field of view.)

Figure 11 : Default Camera Position

381 Copyright Virtual University of Pakistan

41-Viewing

VU

Figure 12 shows the effect of a typical gluLookAt() routine. The camera position (eyex, eyey, eyez) is at (4, 2, 1). In this case, the camera is looking right at the model, so the reference point is at (2, 4, -3). An orientation vector of (2, 2, -1) is chosen to rotate the viewpoint to this 45-degree angle.

Figure 12 : Using gluLookAt() So, to achieve this effect, call gluLookAt(4.0, 2.0, 1.0, 2.0, 4.0, -3.0, 2.0, 2.0, -1.0); Note that gluLookAt() is part of the Utility Library rather than the basic OpenGL library. This isn't because it's not useful, but because it encapsulates several basic OpenGL commands - specifically, glTranslate*() and glRotate*(). To see this, imagine a camera located at an arbitrary viewpoint and oriented according to a line of sight, both as specified with gluLookAt() and a scene located at the origin. To "undo" what gluLookAt() does, we need to transform the camera so that it sits at the origin and points down the negative z-axis, the default position. A simple translate moves the camera to the origin. We can easily imagine a series of rotations about each of the three axes of a fixed coordinate system that would orient the camera so that it pointed toward negative z values. Since OpenGL allows rotation about an arbitrary axis, we can accomplish any desired rotation of the camera with a single glRotate*() command. Note: we can have only one active viewing transformation. we cannot try to combine the effects of two viewing transformations, any more than a camera can have two tripods. If we want to change the position of the camera, make sure we call glLoadIdentity() to wipe away the effects of any current viewing transformation. Advanced To transform any arbitrary vector so that it's coincident with another arbitrary vector (for instance, the negative z-axis), we need to do a little mathematics. The axis about which we want to rotate is given by the cross product of the two normalized vectors. To find the angle of rotation, normalize the initial two vectors. The cosine of the desired angle between the vectors is equal to the dot product of the normalized vectors. The angle of rotation around the axis given by the cross product is always between 0 and 180 degrees. Note that computing the angle between two normalized vectors by taking the inverse 382 Copyright Virtual University of Pakistan

41-Viewing

VU

cosine of their dot product is not very accurate, especially for small angles. But it should work well enough to get us started. Creating a Custom Utility Routine For some specialized applications, we might want to define our own transformation routine. Since this is rarely done and in any case is a fairly advanced topic, it's left mostly as an exercise for the reader. The following exercises suggest two custom viewing transformations that might be useful. Try This Suppose we're writing a flight simulator and we'd like to display the world from the point of view of the pilot of a plane. The world is described in a coordinate system with the origin on the runway and the plane at coordinates (x, y, z). Suppose further that the plane has some roll, pitch, and heading (these are rotation angles of the plane relative to its center of gravity).

Show that the following routine could serve as the viewing transformation: void pilotView{GLdouble planex, GLdouble planey, GLdouble planez, GLdouble roll, GLdouble pitch, GLdouble heading) { glRotated(roll, 0.0, 0.0, 1.0); glRotated(pitch, 0.0, 1.0, 0.0); glRotated(heading, 1.0, 0.0, 0.0); glTranslated(-planex, -planey, -planez); } l Suppose our application involves orbiting the camera around an object that's centered at the origin. In this case, we'd like to specify the viewing transformation by using polar coordinates. Let the distance variable define the radius of the orbit, or how far the camera is from the origin. (Initially, the camera is moved distance units along the positive z-axis.) The azimuth describes the angle of rotation of the camera about the object in the x-y plane, measured from the positive y-axis. Similarly, elevation is the angle of rotation of the camera in the y-z plane, measured from the positive z-axis. Finally, twist represents the rotation of the viewing volume around its line of sight. Show that the following routine could serve as the viewing transformation: void polarView{GLdouble distance, GLdouble twist, GLdouble elevation, GLdouble azimuth) { glTranslated(0.0, 0.0, -distance); glRotated(-twist, 0.0, 0.0, 1.0); glRotated(-elevation, 1.0, 0.0, 0.0); glRotated(azimuth, 0.0, 0.0, 1.0); } 383 Copyright Virtual University of Pakistan

41-Viewing

VU

Manipulating the Matrix Stacks The modelview and projection matrices we've been creating, loading, and multiplying have only been the visible tips of their respective icebergs. Each of these matrices is actually the topmost member of a stack of matrices (see Figure 20).

Figure 20 : Modelview and Projection Matrix Stacks A stack of matrices is useful for constructing hierarchical models, in which complicated objects are constructed from simpler ones. For example, suppose we're drawing an automobile that has four wheels, each of which is attached to the car with five bolts. We have a single routine to draw a wheel and another to draw a bolt, since all the wheels and all the bolts look the same. These routines draw a wheel or a bolt in some convenient position and orientation, say centered at the origin with its axis coincident with the z axis. When we draw the car, including the wheels and bolts, we want to call the wheel-drawing routine four times with different transformations in effect each time to position the wheels correctly. As we draw each wheel, we want to draw the bolts five times, each time translated appropriately relative to the wheel. Suppose for a minute that all we have to do is draw the car body and the wheels. The English description of what we want to do might be something like this:

Draw the car body. Remember where we are, and translate to the right front wheel. Draw the wheel and throw away the last translation so our current position is back at the origin of the car body. Remember where we are, and translate to the left front wheel... Similarly, for each wheel, we want to draw the wheel, remember where we are, and successively translate to each of the positions that bolts are drawn, throwing away the transformations after each bolt is drawn. Since the transformations are stored as matrices, a matrix stack provides an ideal mechanism for doing this sort of successive remembering, translating, and throwing away. All the matrix operations that have been described so far (glLoadMatrix(), glMultMatrix(), glLoadIdentity() and the commands that create specific transformation matrices) deal with the current matrix, or the top matrix on the stack. We can control which matrix is on top with the commands that perform stack operations: glPushMatrix(), which copies the current matrix and adds the copy to the top of the 384 Copyright Virtual University of Pakistan

41-Viewing

VU

stack, and glPopMatrix(), which discards the top matrix on the stack, as shown in Figure 21. In effect, glPushMatrix() means "remember where we are" and glPopMatrix() means "go back to where we were."

Figure 21 : Pushing and Popping the Matrix Stack void glPushMatrix(void); Pushes all matrices in the current stack down one level. The current stack is determined by glMatrixMode(). The topmost matrix is copied, so its contents are duplicated in both the top and second-from-the-top matrix. If too many matrices are pushed, an error is generated. void glPopMatrix(void); Pops the top matrix off the stack, destroying the contents of the popped matrix. What was the second-from-the-top matrix becomes the top matrix. The current stack is determined by glMatrixMode(). If the stack contains a single matrix, calling glPopMatrix() generates an error. Example 4 draws an automobile, assuming the existence of routines that draw the car body, a wheel, and a bolt. Example 4 : Pushing and Popping the Matrix draw_wheel_and_bolts() { long i; draw_wheel(); for(i=0;i<5;i++){ glPushMatrix(); glRotatef(72.0*i,0.0,0.0,1.0); glTranslatef(3.0,0.0,0.0); draw_bolt(); glPopMatrix(); } } draw_body_and_wheel_and_bolts() { draw_car_body(); glPushMatrix(); 385 Copyright Virtual University of Pakistan

41-Viewing

VU

glTranslatef(40,0,30); /*move to first wheel position*/ draw_wheel_and_bolts(); glPopMatrix(); glPushMatrix(); glTranslatef(40,0,-30); /*move to 2nd wheel position*/ draw_wheel_and_bolts(); glPopMatrix(); ... /*draw last two wheels similarly*/ } This code assumes the wheel and bolt axes are coincident with the z-axis, that the bolts are evenly spaced every 72 degrees, 3 units (maybe inches) from the center of the wheel, and that the front wheels are 40 units in front of and 30 units to the right and left of the car's origin. A stack is more efficient than an individual matrix, especially if the stack is implemented in hardware. When we push a matrix, we don't need to copy the current data back to the main process, and the hardware may be able to copy more than one element of the matrix at a time. Sometimes we might want to keep an identity matrix at the bottom of the stack so that we don't need to call glLoadIdentity() repeatedly.

386 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

Lecture No.42

Examples of Composing Several Transformations

This section demonstrates how to combine several transformations to achieve a particular result. The two examples discussed are a solar system, in which objects need to rotate on their axes as well as in orbit around each other, and a robot arm, which has several joints that effectively transform coordinate systems as they move relative to each other. Building a Solar System The program described in this section draws a simple solar system with a planet and a sun, both using the same sphere-drawing routine. To write this program, we need to use glRotate*() for the revolution of the planet around the sun and for the rotation of the planet around its own axis. We also need glTranslate*() to move the planet out to its orbit, away from the origin of the solar system. Remember that we can specify the desired size of the two spheres by supplying the appropriate arguments for the glutWireSphere() routine. To draw the solar system, we first want to set up a projection and a viewing transformation. For this example, gluPerspective() and gluLookAt() are used. Drawing the sun is straightforward, since it should be located at the origin of the grand, fixed coordinate system, which is where the sphere routine places it. Thus, drawing the sun doesn't require translation; we can use glRotate*() to make the sun rotate about an arbitrary axis. To draw a planet rotating around the sun, as shown in Figure 24, requires several modeling transformations. The planet needs to rotate about its own axis once a day. And once a year, the planet completes one revolution around the sun.

Figure 24 : Planet and Sun To determine the order of modeling transformations, visualize what happens to the local coordinate system. An initial glRotate*() rotates the local coordinate system that initially coincides with the grand coordinate system. Next, glTranslate*() moves the local coordinate system to a position on the planet's orbit; the distance moved should equal the radius of the orbit. Thus, the initial glRotate*() actually determines where along the orbit the planet is (or what time of year it is). A second glRotate*() rotates the local coordinate system around the local axes, thus determining the time of day for the planet. Once we've issued all these transformation commands, the planet can be drawn. In summary, these are the OpenGL commands to draw the sun and planet; the full program is shown in Example 6. glPushMatrix(); glutWireSphere(1.0, 20, 16); /* draw sun */ glRotatef ((GLfloat) year, 0.0, 1.0, 0.0); glTranslatef (2.0, 0.0, 0.0); 387 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

glRotatef ((GLfloat) day, 0.0, 1.0, 0.0); glutWireSphere(0.2, 10, 8); /* draw smaller planet */ glPopMatrix(); Example 6 : Planetary System: #include <GL/gl.h> #include <GL/glu.h> #include <GL/glut.h> static int year = 0, day = 0; void init(void) { glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_FLAT); } void display(void) { glClear (GL_COLOR_BUFFER_BIT); glColor3f (1.0, 1.0, 1.0); glPushMatrix(); glutWireSphere(1.0, 20, 16); /* draw sun */ glRotatef ((GLfloat) year, 0.0, 1.0, 0.0); glTranslatef (2.0, 0.0, 0.0); glRotatef ((GLfloat) day, 0.0, 1.0, 0.0); glutWireSphere(0.2, 10, 8); /* draw smaller planet */ glPopMatrix(); glutSwapBuffers(); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity (); gluPerspective(60.0, (GLfloat) w/(GLfloat) h, 1.0, 20.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); gluLookAt (0.0, 0.0, 5.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0); } void keyboard (unsigned char key, int x, int y) { switch (key) { case `d': day = (day + 10) % 360; glutPostRedisplay(); break; case `D': day = (day - 10) % 360; glutPostRedisplay(); break; case `y': year = (year + 5) % 360; 388 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

glutPostRedisplay(); break; case `Y': year = (year - 5) % 360; glutPostRedisplay(); break; default: break; } } int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_DOUBLE | GLUT_RGB); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutKeyboardFunc(keyboard); glutMainLoop(); return 0; } Try This Try adding a moon to the planet. Or try several moons and additional planets. Hint: Use glPushMatrix() and glPopMatrix() to save and restore the position and orientation of the coordinate system at appropriate moments. If we're going to draw several moons around a planet, we need to save the coordinate system prior to positioning each moon and restore the coordinate system after each moon is drawn. Try tilting the planet's axis. Building an Articulated Robot Arm This section discusses a program that creates an articulated robot arm with two or more segments. The arm should be connected with pivot points at the shoulder, elbow, or other joints. Figure 25 shows a single joint of such an arm.

Figure 25 : Robot Arm We can use a scaled cube as a segment of the robot arm, but first we must call the appropriate modeling transformations to orient each segment. Since the origin of the local coordinate system is initially at the center of the cube, we need to move the local 389 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

coordinate system to one edge of the cube. Otherwise, the cube rotates about its center rather than the pivot point. After we call glTranslate*() to establish the pivot point and glRotate*() to pivot the cube, translate back to the center of the cube. Then the cube is scaled (flattened and widened) before it is drawn. The glPushMatrix() and glPopMatrix() restrict the effect of glScale*(). Here's what our code might look like for this first segment of the arm (the entire program is shown in Example 7): glTranslatef (-1.0, 0.0, 0.0); glRotatef ((GLfloat) shoulder, 0.0, 0.0, 1.0); glTranslatef (1.0, 0.0, 0.0); glPushMatrix(); glScalef (2.0, 0.4, 1.0); glutWireCube (1.0); glPopMatrix(); To build a second segment, we need to move the local coordinate system to the next pivot point. Since the coordinate system has previously been rotated, the x-axis is already oriented along the length of the rotated arm. Therefore, translating along the x-axis moves the local coordinate system to the next pivot point. Once it's at that pivot point, we can use the same code to draw the second segment as we used for the first one. This can be continued for an indefinite number of segments (shoulder, elbow, wrist, fingers). glTranslatef (1.0, 0.0, 0.0); glRotatef ((GLfloat) elbow, 0.0, 0.0, 1.0); glTranslatef (1.0, 0.0, 0.0); glPushMatrix(); glScalef (2.0, 0.4, 1.0); glutWireCube (1.0); glPopMatrix(); Example 7 : Robot Arm: #include <GL/gl.h> #include <GL/glu.h> #include <GL/glut.h> static int shoulder = 0, elbow = 0; void init(void) { glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_FLAT); } void display(void) { glClear (GL_COLOR_BUFFER_BIT); glPushMatrix(); glTranslatef (-1.0, 0.0, 0.0); glRotatef ((GLfloat) shoulder, 0.0, 0.0, 1.0); glTranslatef (1.0, 0.0, 0.0); glPushMatrix(); 390 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

glScalef (2.0, 0.4, 1.0); glutWireCube (1.0); glPopMatrix(); glTranslatef (1.0, 0.0, 0.0); glRotatef ((GLfloat) elbow, 0.0, 0.0, 1.0); glTranslatef (1.0, 0.0, 0.0); glPushMatrix(); glScalef (2.0, 0.4, 1.0); glutWireCube (1.0); glPopMatrix(); glPopMatrix(); glutSwapBuffers(); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity (); gluPerspective(65.0, (GLfloat) w/(GLfloat) h, 1.0, 20.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); glTranslatef (0.0, 0.0, -5.0); } void keyboard (unsigned char key, int x, int y) { switch (key) { case `s': /* s key rotates at shoulder */ shoulder = (shoulder + 5) % 360; glutPostRedisplay(); break; case `S': shoulder = (shoulder - 5) % 360; glutPostRedisplay(); break; case `e': /* e key rotates at elbow */ elbow = (elbow + 5) % 360; glutPostRedisplay(); break; case `E': elbow = (elbow - 5) % 360; glutPostRedisplay(); break; default: break; } } int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_DOUBLE | GLUT_RGB); 391 Copyright Virtual University of Pakistan

42-Examples of Composing Several Transformations

VU

glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutKeyboardFunc(keyboard); glutMainLoop(); return 0; } Try This Modify Example 7 to add additional segments onto the robot arm. l Modify Example 7 to add additional segments at the same position. For example, give the robot arm several "fingers" at the wrist, as shown in Figure 26. Hint: Use glPushMatrix() and glPopMatrix() to save and restore the position and orientation of the coordinate system at the wrist. If we're going to draw fingers at the wrist, we need to save the current matrix prior to positioning each finger and restore the current matrix after each finger is drawn.

Figure 26: Robot Arm with Fingers

392 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Lecture No.43

Real-World and OpenGL Lighting

When we look at a physical surface, our eye's perception of the color depends on the distribution of photon energies that arrive and trigger our cone cells. Those photons come from a light source or combination of sources, some of which are absorbed and some are reflected by the surface. In addition, different surfaces may have very different properties - some are shiny and preferentially reflect light in certain directions, while others scatter incoming light equally in all directions. Most surfaces are somewhere in between. OpenGL approximates light and lighting as if light can be broken into red, green, and blue components. Thus, the color of light sources is characterized by the amount of red, green, and blue light they emit, and the material of surfaces is characterized by the percentage of the incoming red, green, and blue components that is reflected in various directions. The OpenGL lighting equations are just an approximation but one that works fairly well and can be computed relatively quickly. If we desire a more accurate (or just different) lighting model, we have to do our own calculations in software. Such software can be enormously complex, as a few hours of reading any optics textbook should convince us. In the OpenGL lighting model, the light in a scene comes from several light sources that can be individually turned on and off. Some light comes from a particular direction or position, and some light is generally scattered about the scene. For example, when we turn on a light bulb in a room, most of the light comes from the bulb, but some light comes after bouncing off one, two, three, or more walls. This bounced light (called ambient) is assumed to be so scattered that there is no way to tell its original direction, but it disappears if a particular light source is turned off. Finally, there might be a general ambient light in the scene that comes from no particular source, as if it had been scattered so many times that its original source is impossible to determine. In the OpenGL model, the light sources have an effect only when there are surfaces that absorb and reflect light. Each surface is assumed to be composed of a material with various properties. A material might emit its own light (like headlights on an automobile), it might scatter some incoming light in all directions, and it might reflect some portion of the incoming light in a preferential direction like a mirror or other shiny surface. The OpenGL lighting model considers the lighting to be divided into four independent components: emissive, ambient, diffuse and specular. All four components are computed independently and then added together. A Simple Example: Rendering a Lit Sphere These are the steps required to add lighting to our scene. Define NORMAL vectors for each vertex of all the objects. These NORMALS determine the orientation of the object relative to the light sources. Create, select, and position one or more light sources. Create and select a lighting model, which defines the level of global ambient light and the effective location of the viewpoint (for the purposes of lighting calculations) Define material properties for the objects in the scene. Example 1 accomplishes these tasks. It displays a sphere illuminated by a single light source, as shown earlier in Figure 1. 393 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Example 1 : Drawing a Lit Sphere: #include <GL/gl.h> #include <GL/glu.h> #include <GL/glut.h> void init(void) { GLfloat mat_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat mat_shininess[] = { 50.0 }; GLfloat light_position[] = { 1.0, 1.0, 1.0, 0.0 }; glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_SMOOTH); glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialfv(GL_FRONT, GL_SHININESS, mat_shininess); glLightfv(GL_LIGHT0, GL_POSITION, light_position); glEnable(GL_LIGHTING); glEnable(GL_LIGHT0); glEnable(GL_DEPTH_TEST); } void display(void) { glClear (GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glutSolidSphere (1.0, 20, 16); glFlush (); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity(); if (w <= h) glOrtho (-1.5, 1.5, -1.5*(GLfloat)h/(GLfloat)w, 1.5*(GLfloat)h/(GLfloat)w, -10.0, 10.0); else glOrtho (-1.5*(GLfloat)w/(GLfloat)h, 1.5*(GLfloat)w/(GLfloat)h, -1.5, 1.5, -10.0, 10.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); } int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_SINGLE | GLUT_RGB | GLUT_DEPTH); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutMainLoop(); return 0; } 394 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

The lighting-related calls are in the init() command; they're discussed briefly in the following paragraphs and in more detail later in the chapter. One thing to note about Example 1 is that it uses RGBA color mode, not color-index mode. The OpenGL lighting calculation is different for the two modes, and in fact the lighting capabilities are more limited in color-index mode. Thus, RGBA is the preferred mode when doing lighting. Define Normal Vectors for Each Vertex of Every Object An object's NORMALS determine its orientation relative to the light sources. For each vertex, OpenGL uses the assigned normal to determine how much light that particular vertex receives from each light source. In this example, the NORMALS for the sphere are defined as part of the glutSolidSphere() routine. (recall "Normal Vectors") Create, Position, and Enable One or More Light Sources Example 1 uses only one, white light source; its location is specified by the glLightfv() call. This example uses the default color for light zero (GL_LIGHT0), which is white; if we want a differently colored light, use glLight*() to indicate this. We can include at least eight different light sources in our scene of various colors; the default color of these other lights is black. (The particular implementation of OpenGL we're using might allow more than eight.) we can also locate the lights wherever we desire - we can position them near the scene, as a desk lamp would be, or an infinite distance away, like the sun. In addition, we can control whether a light produces a narrow, focused beam or a wider beam. Remember that each light source adds significantly to the calculations needed to render the scene, so performance is affected by the number of lights in the scene. After we've defined the characteristics of the lights we want, we have to turn them on with the glEnable() command. We also need to call glEnable() with GL_LIGHTING as a parameter to prepare OpenGL to perform lighting calculations. Select a Lighting Model As we might expect, the glLightModel*() command describes the parameters of a lighting model. In Example 1, the only element of the lighting model that's defined explicitly is the global ambient light. The lighting model also defines whether the viewer of the scene should be considered to be an infinite distance away or local to the scene, and whether lighting calculations should be performed differently for the front and back surfaces of objects in the scene. Example 1 uses the default settings for these two aspects of the model - an infinite viewer and one-sided lighting. Using a local viewer adds significantly to the complexity of the calculations that must be performed, because OpenGL must calculate the angle between the viewpoint and each object. With an infinite viewer, however, the angle is ignored, and the results are slightly less realistic. Further, since in this example, the back surface of the sphere is never seen (it's the inside of the sphere), one-sided lighting is sufficient. Define Material Properties for the Objects in the Scene An object's material properties determine how it reflects light and therefore what material it seems to be made of. Because the interaction between an object's material surface and incident light is complex, specifying material properties so that an object has a certain desired appearance is an art. We can specify a material's ambient, diffuse, and specular colors and how shiny it is. In this example, only these last two material properties - the specular material color and shininess - are explicitly specified (with the glMaterialfv() calls). 395 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Some Important Notes As we write our own lighting program, remember that we can use the default values for some lighting parameters; others need to be changed. Also, don't forget to enable whatever lights we define and to enable lighting calculations. Finally, remember that we might be able to use display lists to maximize efficiency as we change lighting conditions. Creating Light Sources Light sources have a number of properties, such as color, position, and direction. The following sections explain how to control these properties and what the resulting light looks like. The command used to specify all properties of lights is glLight*(); it takes three arguments: to identify the light whose property is being specified, the property, and the desired value for that property. void glLight{if}(GLenum light, GLenum pname, TYPEparam); void glLight{if}v(GLenum light, GLenum pname, TYPE *param); Creates the light specified by light, which can be GL_LIGHT0, GL_LIGHT1, ... , or GL_LIGHT7. The characteristic of the light being set is defined by pname, which specifies a named parameter (see Table 1). param indicates the values to which the pname characteristic is set; it's a pointer to a group of values if the vector version is used, or the value itself if the nonvector version is used. The nonvector version can be used to set only single-valued light characteristics. Table 1 : Default Values for pname Parameter of glLight*()

396 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Note: The default values listed for GL_DIFFUSE and GL_SPECULAR in Table 1 apply only to GL_LIGHT(). For other lights, the default value is (0.0, 0.0, 0.0, 1.0) for both GL_DIFFUSE and GL_SPECULAR. Example 2 shows how to use glLight*(): Example 2 : Defining Colors and Position for a Light Source GLfloat light_ambient[] = { 0.0, 0.0, 0.0, 1.0 }; GLfloat light_diffuse[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat light_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat light_position[] = { 1.0, 1.0, 1.0, 0.0 }; glLightfv(GL_LIGHT0, GL_AMBIENT, light_ambient); glLightfv(GL_LIGHT0, GL_DIFFUSE, light_diffuse); glLightfv(GL_LIGHT0, GL_SPECULAR, light_specular); glLightfv(GL_LIGHT0, GL_POSITION, light_position); As we can see, arrays are defined for the parameter values, and glLightfv() is called repeatedly to set the various parameters. In this example, the first three calls to glLightfv() are superfluous, since they're being used to specify the default values for the GL_AMBIENT, GL_DIFFUSE, and GL_SPECULAR parameters. Note: Remember to turn on each light with glEnable(). All the parameters for glLight*() and their possible values are explained in the following sections. These parameters interact with those that define the overall lighting model for a particular scene and an object's material properties. Color OpenGL allows we to associate three different color-related parameters - GL_AMBIENT, GL_DIFFUSE, and GL_SPECULAR - with any particular light. The GL_AMBIENT parameter refers to the RGBA intensity of the ambient light that a particular light source adds to the scene. As we can see in Table 1, by default there is no ambient light since GL_AMBIENT is (0.0, 0.0, 0.0, 1.0). This value was used in Example 1. If this program had specified blue ambient light as GLfloat light_ambient[] = { 0.0, 0.0, 1.0, 1.0}; glLightfv(GL_LIGHT0, GL_AMBIENT, light_ambient); The GL_DIFFUSE parameter probably most closely correlates with what we naturally think of as "the color of a light." It defines the RGBA color of the diffuse light that a particular light source adds to a scene. By default, GL_DIFFUSE is (1.0, 1.0, 1.0, 1.0) for GL_LIGHT0, which produces a bright. The default value for any other light (GL_LIGHT1, ... , GL_LIGHT7) is (0.0, 0.0, 0.0, 0.0). The GL_SPECULAR parameter affects the color of the specular highlight on an object. Typically, a real-world object such as a glass bottle has a specular highlight that's the color of the light shining on it (which is often white). Therefore, if we want to create a realistic effect, set the GL_SPECULAR parameter to the same value as the GL_DIFFUSE parameter. By default, GL_SPECULAR is (1.0, 1.0, 1.0, 1.0) for GL_LIGHT0 and (0.0, 0.0, 0.0, 0.0) for any other light. 397 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Note: The alpha component of these colors is not used until blending is enabled. Position and Attenuation As previously mentioned, we can choose whether to have a light source that's treated as though it's located infinitely far away from the scene or one that's nearer to the scene. The first type is referred to as a directional light source; the effect of an infinite location is that the rays of light can be considered parallel by the time they reach an object. An example of a real-world directional light source is the sun. The second type is called a positional light source, since its exact position within the scene determines the effect it has on a scene and, specifically, the direction from which the light rays come. A desk lamp is an example of a positional light source. The light used in Example 1 is a directional one: GLfloat light_position[] = { 1.0, 1.0, 1.0, 0.0 }; glLightfv(GL_LIGHT0, GL_POSITION, light_position); As shown, we supply a vector of four values (x, y, z, w) for the GL_POSITION parameter. If the last value, w, is zero, the corresponding light source is a directional one, and the (x, y, z) values describe its direction. This direction is transformed by the modelview matrix. By default, GL_POSITION is (0, 0, 1, 0), which defines a directional light that points along the negative z-axis. (Note that nothing prevents we from creating a directional light with the direction of (0, 0, 0), but such a light won't help we much.) If the w value is nonzero, the light is positional, and the (x, y, z) values specify the location of the light in homogeneous object coordinates. This location is transformed by the modelview matrix and stored in eye coordinates. Also, by default, a positional light radiates in all directions, but we can restrict it to producing a cone of illumination by defining the light as a spotlight. Note: Remember that the colors across the face of a smooth-shaded polygon are determined by the colors calculated for the vertices. Because of this, we probably want to avoid using large polygons with local lights. If we locate the light near the middle of the polygon, the vertices might be too far away to receive much light, and the whole polygon will look darker than we intended. To avoid this problem, break up the large polygon into smaller ones. For real-world lights, the intensity of light decreases as distance from the light increases. Since a directional light is infinitely far away, it doesn't make sense to attenuate its intensity over distance, so attenuation is disabled for a directional light. However, we might want to attenuate the light from a positional light. OpenGL attenuates a light source by multiplying the contribution of that source by an attenuation factor:

where d = distance between the light's position and the vertex kc = GL_CONSTANT_ATTENUATION 398 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

kl = GL_LINEAR_ATTENUATION kq = GL_QUADRATIC_ATTENUATION By default, kc is 1.0 and both kl and kq are zero, but we can give these parameters different values: glLightf(GL_LIGHT0, GL_CONSTANT_ATTENUATION, 2.0); glLightf(GL_LIGHT0, GL_LINEAR_ATTENUATION, 1.0); glLightf(GL_LIGHT0, GL_QUADRATIC_ATTENUATION, 0.5); Note that the ambient, diffuse, and specular contributions are all attenuated. Only the emission and global ambient values aren't attenuated. Also note that since attenuation requires an additional division (and possibly more math) for each calculated color, using attenuated lights may slow down application performance. Spotlights As previously mentioned, we can have a positional light source act as a spotlight - that is, by restricting the shape of the light it emits to a cone. To create a spotlight, we need to determine the spread of the cone of light we desire. (Remember that since spotlights are positional lights, we also have to locate them where we want them. Again, note that nothing prevents us from creating a directional spotlight, but it won't give us the result we want.) To specify the angle between the axis of the cone and a ray along the edge of the cone, use the GL_SPOT_CUTOFF parameter. The angle of the cone at the apex is then twice this value, as shown in Figure 2.

Figure 2 : GL_SPOT_CUTOFF Parameter Note that no light is emitted beyond the edges of the cone. By default, the spotlight feature is disabled because the GL_SPOT_CUTOFF parameter is 180.0. This value means that light is emitted in all directions (the angle at the cone's apex is 360 degrees, so it isn't a cone at all). The value for GL_SPOT_CUTOFF is restricted to being within the range [0.0,90.0] (unless it has the special value 180.0). The following line sets the cutoff parameter to 45 degrees: glLightf(GL_LIGHT0, GL_SPOT_CUTOFF, 45.0); We also need to specify a spotlight's direction, which determines the axis of the cone of light:

399 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

GLfloat spot_direction[] = { -1.0, -1.0, 0.0 }; glLightfv(GL_LIGHT0, GL_SPOT_DIRECTION, spot_direction); The direction is specified in object coordinates. By default, the direction is (0.0, 0.0, -1.0), so if we don't explicitly set the value of GL_SPOT_DIRECTION, the light points down the negative z-axis. Also, keep in mind that a spotlight's direction is transformed by the modelview matrix just as though it were a normal vector, and the result is stored in eye coordinates. In addition to the spotlight's cutoff angle and direction, there are two ways we can control the intensity distribution of the light within the cone. First, we can set the attenuation factor described earlier, which is multiplied by the light's intensity. We can also set the GL_SPOT_EXPONENT parameter, which by default is zero, to control how concentrated the light is. The light's intensity is highest in the center of the cone. It's attenuated toward the edges of the cone by the cosine of the angle between the direction of the light and the direction from the light to the vertex being lit, raised to the power of the spot exponent. Thus, higher spot exponents result in a more focused light source. Multiple Lights As mentioned, we can have at least eight lights in our scene (possibly more, depending on our OpenGL implementation). Since OpenGL needs to perform calculations to determine how much light each vertex receives from each light source, increasing the number of lights adversely affects performance. The constants used to refer to the eight lights are GL_LIGHT0, GL_LIGHT1, GL_LIGHT2, GL_LIGHT3, and so on. In the preceding discussions, parameters related to GL_LIGHT0 were set. If we want an additional light, we need to specify its parameters; also, remember that the default values are different for these other lights than they are for GL_LIGHT0, as explained in Table 1. Example 3 defines a white attenuated spotlight. Example 3 : Second Light Source GLfloat light1_ambient[] = { 0.2, 0.2, 0.2, 1.0 }; GLfloat light1_diffuse[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat light1_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat light1_position[] = { -2.0, 2.0, 1.0, 1.0 }; GLfloat spot_direction[] = { -1.0, -1.0, 0.0 }; glLightfv(GL_LIGHT1, GL_AMBIENT, light1_ambient); glLightfv(GL_LIGHT1, GL_DIFFUSE, light1_diffuse); glLightfv(GL_LIGHT1, GL_SPECULAR, light1_specular); glLightfv(GL_LIGHT1, GL_POSITION, light1_position); glLightf(GL_LIGHT1, GL_CONSTANT_ATTENUATION, 1.5); glLightf(GL_LIGHT1, GL_LINEAR_ATTENUATION, 0.5); glLightf(GL_LIGHT1, GL_QUADRATIC_ATTENUATION, 0.2); glLightf(GL_LIGHT1, GL_SPOT_CUTOFF, 45.0); glLightfv(GL_LIGHT1, GL_SPOT_DIRECTION, spot_direction); glLightf(GL_LIGHT1, GL_SPOT_EXPONENT, 2.0); glEnable(GL_LIGHT1); If these lines were added to Example 1, the sphere would be lit with two lights, one directional and one spotlight. 400 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Try This Modify Example 1 in the following manner: Change the first light to be a positional colored light rather than a directional white one. Add an additional colored spotlight. Hint: Use some of the code shown in the preceding section. Measure how these two changes affect performance. Controlling a Light's Position and Direction OpenGL treats the position and direction of a light source just as it treats the position of a geometric primitive. In other words, a light source is subject to the same matrix transformations as a primitive. More specifically, when glLight*() is called to specify the position or the direction of a light source, the position or direction is transformed by the current modelview matrix and stored in eye coordinates. This means we can manipulate a light source's position or direction by changing the contents of the modelview matrix. (The projection matrix has no effect on a light's position or direction.) This section explains how to achieve the following three different effects by changing the point in the program at which the light position is set, relative to modeling or viewing transformations: A light position that remains fixed A light that moves around a stationary object A light that moves along with the viewpoint Keeping the Light Stationary In the simplest example, as in Example 1, the light position remains fixed. To achieve this effect, we need to set the light position after whatever viewing and/or modeling transformation we use. In Example 4, the relevant code from the init() and reshape() routines might look like this. Example 4 : Stationary Light Source glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity(); if (w <= h) glOrtho (-1.5, 1.5, -1.5*h/w, 1.5*h/w, -10.0, 10.0); else glOrtho (-1.5*w/h, 1.5*w/h, -1.5, 1.5, -10.0, 10.0); glMatrixMode (GL_MODELVIEW); glLoadIdentity(); /* later in init() */ GLfloat light_position[] = { 1.0, 1.0, 1.0, 1.0 }; glLightfv(GL_LIGHT0, GL_POSITION, position); As we can see, the viewport and projection matrices are established first. Then, the identity matrix is loaded as the modelview matrix, after which the light position is set. Since the identity matrix is used, the originally specified light position (1.0, 1.0, 1.0) isn't changed by being multiplied by the modelview matrix. Then, since neither the light position nor the modelview matrix is modified after this point, the direction of the light remains (1.0, 1.0, 1.0). 401 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Independently Moving the Light Now suppose we want to rotate or translate the light position so that the light moves relative to a stationary object. One way to do this is to set the light position after the modeling transformation, which is itself changed specifically to modify the light position. We can begin with the same series of calls in init() early in the program. Then we need to perform the desired modeling transformation (on the modelview stack) and reset the light position, probably in display(). Example 5 shows what display() might be. Example 5 : Independently Moving Light Source static GLdouble spin; void display(void) { GLfloat light_position[] = { 0.0, 0.0, 1.5, 1.0 }; glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glPushMatrix(); gluLookAt (0.0, 0.0, 5.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0); glPushMatrix(); glRotated(spin, 1.0, 0.0, 0.0); glLightfv(GL_LIGHT0, GL_POSITION, light_position); glPopMatrix(); glutSolidTorus (0.275, 0.85, 8, 15); glPopMatrix(); glFlush(); } spin is a global variable and is probably controlled by an input device. display() causes the scene to be redrawn with the light rotated spin degrees around a stationary torus. Note the two pairs of glPushMatrix() and glPopMatrix() calls, which are used to isolate the viewing and modeling transformations, all of which occur on the modelview stack. Since in Example 5 the viewpoint remains constant, the current matrix is pushed down the stack and then the desired viewing transformation is loaded with gluLookAt(). The matrix stack is pushed again before the modeling transformation glRotated() is specified. Then the light position is set in the new, rotated coordinate system so that the light itself appears to be rotated from its previous position. (Remember that the light position is stored in eye coordinates, which are obtained after transformation by the modelview matrix.) After the rotated matrix is popped off the stack, the torus is drawn. Example 6 is a program that rotates a light source around an object. When the left mouse button is pressed, the light position rotates an additional 30 degrees. A small, unlit, wireframe cube is drawn to represent the position of the light in the scene. Example 6 : Moving a Light with Modeling Transformations: #include <GL/gl.h> #include <GL/glu.h> #include "glut.h" static int spin = 0; void init(void) { glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_SMOOTH); 402 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

glEnable(GL_LIGHTING); glEnable(GL_LIGHT0); glEnable(GL_DEPTH_TEST); } /* Here is where the light position is reset after the modeling * transformation (glRotated) is called. This places the * light at a new position in world coordinates. The cube * represents the position of the light. */ void display(void) { GLfloat position[] = { 0.0, 0.0, 1.5, 1.0 }; glClear (GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glPushMatrix (); glTranslatef (0.0, 0.0, -5.0); glPushMatrix (); glRotated ((GLdouble) spin, 1.0, 0.0, 0.0); glLightfv (GL_LIGHT0, GL_POSITION, position); glTranslated (0.0, 0.0, 1.5); glDisable (GL_LIGHTING); glColor3f (0.0, 1.0, 1.0); glutWireCube (0.1); glEnable (GL_LIGHTING); glPopMatrix (); glutSolidTorus (0.275, 0.85, 8, 15); glPopMatrix (); glFlush (); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity(); gluPerspective(40.0, (GLfloat) w/(GLfloat) h, 1.0, 20.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); } void mouse(int button, int state, int x, int y) { switch (button) { case GLUT_LEFT_BUTTON: if (state == GLUT_DOWN) { spin = (spin + 30) % 360; glutPostRedisplay(); } break; default: break; } } 403 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_SINGLE | GLUT_RGB | GLUT_DEPTH); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutMouseFunc(mouse); glutMainLoop(); return 0; } Selecting a Lighting Model The OpenGL notion of a lighting model has three components: The global ambient light intensity Whether the viewpoint position is local to the scene or whether it should be considered to be an infinite distance away Whether lighting calculations should be performed differently for both the front and back faces of objects This section explains how to specify a lighting model. It also discusses how to enable lighting - that is, how to tell OpenGL that we want lighting calculations performed. The command used to specify all properties of the lighting model is glLightModel*(). glLightModel*() has two arguments: the lighting model property and the desired value for that property. void glLightModel{if}(GLenum pname, TYPEparam); void glLightModel{if}v(GLenum pname, TYPE *param); Sets properties of the lighting model. The characteristic of the lighting model being set is defined by pname, which specifies a named parameter (see Table 2). param indicates the values to which the pname characteristic is set; it's a pointer to a group of values if the vector version is used, or the value itself if the nonvector version is used. The nonvector version can be used to set only single-valued lighting model characteristics, not for GL_LIGHT_MODEL_AMBIENT. Table 2 : Default Values for pname Parameter of glLightModel*()

404 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Global Ambient Light As discussed earlier, each light source can contribute ambient light to a scene. In addition, there can be other ambient light that's not from any particular source. To specify the RGBA intensity of such global ambient light, use the GL_LIGHT_MODEL_AMBIENT parameter as follows: GLfloat lmodel_ambient[] = { 0.2, 0.2, 0.2, 1.0 }; glLightModelfv(GL_LIGHT_MODEL_AMBIENT, lmodel_ambient); In this example, the values used for lmodel_ambient are the default values for GL_LIGHT_MODEL_AMBIENT. Since these numbers yield a small amount of white ambient light, even if we don't add a specific light source to our scene, we can still see the objects in the scene. Enabling Lighting With OpenGL, we need to explicitly enable (or disable) lighting. If lighting isn't enabled, the current color is simply mapped onto the current vertex, and no calculations concerning normals, light sources, the lighting model, and material properties are performed. Here's how to enable lighting: glEnable(GL_LIGHTING); To disable lighting, call glDisable() with GL_LIGHTING as the argument. We also need to explicitly enable each light source that we define, after we've specified the parameters for that source. Example 1 uses only one light, GL_LIGHT0: glEnable(GL_LIGHT0);

Defining Material Properties We've seen how to create light sources with certain characteristics and how to define the desired lighting model. This section describes how to define the material properties of the objects in the scene: the ambient, diffuse, and specular colors, the shininess, and the color of any emitted light. Most of the material properties are conceptually similar to ones we've already used to create light sources. The mechanism for setting them is similar, except that the command used is called glMaterial*(). void glMaterial{if}(GLenum face, GLenum pname, TYPEparam); void glMaterial{if}v(GLenum face, GLenum pname, TYPE *param); Specifies a current material property for use in lighting calculations. face can be GL_FRONT, GL_BACK, or GL_FRONT_AND_BACK to indicate which face of the object the material should be applied to. The particular material property being set is identified by pname and the desired values for that property are given by param, which is either a pointer to a group of values (if the vector version is used) or the actual value (if the nonvector version is used). The nonvector version works only for setting 405 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

GL_SHININESS. The possible values for pname are shown in Table 3. Note that GL_AMBIENT_AND_DIFFUSE allows we to set both the ambient and diffuse material colors simultaneously to the same RGBA value. Table 3 : Default Values for pname Parameter of glMaterial*()

As discussed in "Selecting a Lighting Model," we can choose to have lighting calculations performed differently for the front- and back-facing polygons of objects. If the back faces might indeed be seen, we can supply different material properties for the front and the back surfaces by using the face parameter of glMaterial*(). Note that most of the material properties set with glMaterial*() are (R, G, B, A) colors. Regardless of what alpha values are supplied for other parameters, the alpha value at any particular vertex is the diffuse-material alpha value (that is, the alpha value given to GL_DIFFUSE with the glMaterial*() command, as described in the next section). Also, none of the RGBA material properties apply in color-index mode. Diffuse and Ambient Reflection The GL_DIFFUSE and GL_AMBIENT parameters set with glMaterial*() affect the color of the diffuse and ambient light reflected by an object. Diffuse reflectance plays the most important role in determining what we perceive the color of an object to be. It's affected by the color of the incident diffuse light and the angle of the incident light relative to the normal direction. (It's most intense where the incident light falls perpendicular to the surface.) The position of the viewpoint doesn't affect diffuse reflectance at all. Ambient reflectance affects the overall color of the object. Because diffuse reflectance is brightest where an object is directly illuminated, ambient reflectance is most noticeable where an object receives no direct illumination. An object's total ambient reflectance is affected by the global ambient light and ambient light from individual light sources. Like 406 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

diffuse reflectance, ambient reflectance isn't affected by the position of the viewpoint. For real-world objects, diffuse and ambient reflectance are normally the same color. For this reason, OpenGL provides we with a convenient way of assigning the same value to both simultaneously with glMaterial*(): GLfloat mat_amb_diff[] = { 0.1, 0.5, 0.8, 1.0 }; glMaterialfv(GL_FRONT_AND_BACK, GL_AMBIENT_AND_DIFFUSE, mat_amb_diff); In this example, the RGBA color (0.1, 0.5, 0.8, 1.0) - a deep blue color - represents the current ambient and diffuse reflectance for both the front- and back-facing polygons. Specular Reflection Specular reflection from an object produces highlights. Unlike ambient and diffuse reflection, the amount of specular reflection seen by a viewer does depend on the location of the viewpoint - it's brightest along the direct angle of reflection. To see this, imagine looking at a metallic ball outdoors in the sunlight. As we move our head, the highlight created by the sunlight moves with us to some extent. However, if we move our head too much, we lose the highlight entirely. OpenGL allows us to set the effect that the material has on reflected light (with GL_SPECULAR) and control the size and brightness of the highlight (with GL_SHININESS). We can assign a number in the range of [0.0, 128.0] to GL_SHININESS - the higher the value, the smaller and brighter (more focused) the highlight.

Twelve spheres, each with different material parameters. The row properties are as labeled above. The first column uses a blue diffuse material color with no specular properties. The second column adds white specular reflection with a low shininess exponent. The third column uses a high shininess exponent and thus has a more concentrated highlight. The fourth column uses the blue diffuse color and, instead of specular reflection, adds an emissive component. In above figure, the spheres in the first column have no specular reflection. In the second column, GL_SPECULAR and GL_SHININESS are assigned values as follows: 407 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

GLfloat mat_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat low_shininess[] = { 5.0 }; glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialfv(GL_FRONT, GL_SHININESS, low_shininess); In the third column, the GL_SHININESS parameter is increased to 100.0. Emission By specifying an RGBA color for GL_EMISSION, we can make an object appear to be giving off light of that color. Since most real-world objects (except lights) don't emit light, we'll probably use this feature mostly to simulate lamps and other light sources in a scene. GLfloat mat_emission[] = {0.3, 0.2, 0.2, 0.0}; glMaterialfv(GL_FRONT, GL_EMISSION, mat_emission); Notice that the spheres appear to be slightly glowing; however, they're not actually acting as light sources. We would need to create a light source and position it at the same location as the sphere to create that effect. Changing Material Properties Example 1 uses the same material properties for all vertices of the only object in the scene (the sphere). In other situations, we might want to assign different material properties for different vertices on the same object. More likely, we have more than one object in the scene, and each object has different material properties. For example, the code that produced "above figure has to draw twelve different objects (all spheres), each with different material properties. Example 8 shows a portion of the code in display(). Example 8 : Different Material Properties: GLfloat no_mat[] = { 0.0, 0.0, 0.0, 1.0 }; GLfloat mat_ambient[] = { 0.7, 0.7, 0.7, 1.0 }; GLfloat mat_ambient_color[] = { 0.8, 0.8, 0.2, 1.0 }; GLfloat mat_diffuse[] = { 0.1, 0.5, 0.8, 1.0 }; GLfloat mat_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat no_shininess[] = { 0.0 }; GLfloat low_shininess[] = { 5.0 }; GLfloat high_shininess[] = { 100.0 }; GLfloat mat_emission[] = {0.3, 0.2, 0.2, 0.0}; glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); /* draw sphere in first row, first column * diffuse reflection only; no ambient or specular */ glPushMatrix(); glTranslatef (-3.75, 3.0, 0.0); glMaterialfv(GL_FRONT, GL_AMBIENT, no_mat); glMaterialfv(GL_FRONT, GL_DIFFUSE, mat_diffuse); glMaterialfv(GL_FRONT, GL_SPECULAR, no_mat); glMaterialfv(GL_FRONT, GL_SHININESS, no_shininess); glMaterialfv(GL_FRONT, GL_EMISSION, no_mat); 408 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

glutSolidSphere(1.0, 16, 16); glPopMatrix(); /* draw sphere in first row, second column * diffuse and specular reflection; low shininess; no ambient */ glPushMatrix(); glTranslatef (-1.25, 3.0, 0.0); glMaterialfv(GL_FRONT, GL_AMBIENT, no_mat); glMaterialfv(GL_FRONT, GL_DIFFUSE, mat_diffuse); glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialfv(GL_FRONT, GL_SHININESS, low_shininess); glMaterialfv(GL_FRONT, GL_EMISSION, no_mat); glutSolidSphere(1.0, 16, 16); glPopMatrix(); /* draw sphere in first row, third column * diffuse and specular reflection; high shininess; no ambient */ glPushMatrix(); glTranslatef (1.25, 3.0, 0.0); glMaterialfv(GL_FRONT, GL_AMBIENT, no_mat); glMaterialfv(GL_FRONT, GL_DIFFUSE, mat_diffuse); glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialfv(GL_FRONT, GL_SHININESS, high_shininess); glMaterialfv(GL_FRONT, GL_EMISSION, no_mat); glutSolidSphere(1.0, 16, 16); glPopMatrix(); /* draw sphere in first row, fourth column * diffuse reflection; emission; no ambient or specular refl. */ glPushMatrix(); glTranslatef (3.75, 3.0, 0.0); glMaterialfv(GL_FRONT, GL_AMBIENT, no_mat); glMaterialfv(GL_FRONT, GL_DIFFUSE, mat_diffuse); glMaterialfv(GL_FRONT, GL_SPECULAR, no_mat); glMaterialfv(GL_FRONT, GL_SHININESS, no_shininess); glMaterialfv(GL_FRONT, GL_EMISSION, mat_emission); glutSolidSphere(1.0, 16, 16); glPopMatrix(); As we can see, glMaterialfv() is called repeatedly to set the desired material property for each sphere. Note that it only needs to be called to change a property that needs to be respecified. The second, third, and fourth spheres use the same ambient and diffuse properties as the first sphere, so these properties do not need to be respecified. Since glMaterial*() has a performance cost associated with its use, Example 8 could be rewritten to minimize material-property changes. Another technique for minimizing performance costs associated with changing material properties is to use glColorMaterial(). void glColorMaterial(GLenum face, GLenum mode); 409 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

Causes the material property (or properties) specified by mode of the specified material face (or faces) specified by face to track the value of the current color at all times. A change to the current color (using glColor*()) immediately updates the specified material properties. The face parameter can beGL_FRONT, GL_BACK, or GL_FRONT_AND_BACK (the default). The mode parameter can be GL_AMBIENT, GL_DIFFUSE, GL_AMBIENT_AND_DIFFUSE (the default), GL_SPECULAR, or GL_EMISSION. At any given time, only one mode is active. glColorMaterial() has no effect on color-index lighting. Note that glColorMaterial() specifies two independent values: the first specifies which face or faces are updated, and the second specifies which material property or properties of those faces are updated. OpenGL does not maintain separate mode variables for each face. After calling glColorMaterial(), we need to call glEnable() with GL_COLOR_MATERIAL as the parameter. Then, we can change the current color using glColor*() (or other material properties, using glMaterial*()) as needed as we draw: glEnable(GL_COLOR_MATERIAL); glColorMaterial(GL_FRONT, GL_DIFFUSE); /* now glColor* changes diffuse reflection */ glColor3f(0.2, 0.5, 0.8); /* draw some objects here */ glColorMaterial(GL_FRONT, GL_SPECULAR); /* glColor* no longer changes diffuse reflection */ /* now glColor* changes specular reflection */ glColor3f(0.9, 0.0, 0.2); /* draw other objects here */ glDisable(GL_COLOR_MATERIAL); We should use glColorMaterial() whenever we need to change a single material parameter for most vertices in our scene. If we need to change more than one material parameter, as was the case for "in above figure, use glMaterial*(). When we don't need the capabilities of glColorMaterial() anymore, be sure to disable it so that we don't get undesired material properties and don't incur the performance cost associated with it. The performance value in using glColorMaterial() varies, depending on our OpenGL implementation. Some implementations may be able to optimize the vertex routines so that they can quickly update material properties based on the current color. Example 9 shows an interactive program that uses glColorMaterial() to change material parameters. Pressing each of the three mouse buttons changes the color of the diffuse reflection. Example 9 : Using glColorMaterial(): #include <GL/gl.h> #include <GL/glu.h> #include "glut.h" GLfloat diffuseMaterial[4] = { 0.5, 0.5, 0.5, 1.0 }; void init(void) { GLfloat mat_specular[] = { 1.0, 1.0, 1.0, 1.0 }; GLfloat light_position[] = { 1.0, 1.0, 1.0, 0.0 }; 410 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

glClearColor (0.0, 0.0, 0.0, 0.0); glShadeModel (GL_SMOOTH); glEnable(GL_DEPTH_TEST); glMaterialfv(GL_FRONT, GL_DIFFUSE, diffuseMaterial); glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialf(GL_FRONT, GL_SHININESS, 25.0); glLightfv(GL_LIGHT0, GL_POSITION, light_position); glEnable(GL_LIGHTING); glEnable(GL_LIGHT0); glColorMaterial(GL_FRONT, GL_DIFFUSE); glEnable(GL_COLOR_MATERIAL); } void display(void) { glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glutSolidSphere(1.0, 20, 16); glFlush (); } void reshape (int w, int h) { glViewport (0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode (GL_PROJECTION); glLoadIdentity(); if (w <= h) glOrtho (-1.5, 1.5, -1.5*(GLfloat)h/(GLfloat)w, 1.5*(GLfloat)h/(GLfloat)w, -10.0, 10.0); else glOrtho (-1.5*(GLfloat)w/(GLfloat)h, 1.5*(GLfloat)w/(GLfloat)h, -1.5, 1.5, -10.0, 10.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); } void mouse(int button, int state, int x, int y) { switch (button) { case GLUT_LEFT_BUTTON: if (state == GLUT_DOWN) { /* change red */ diffuseMaterial[0] += 0.1; if (diffuseMaterial[0] > 1.0) diffuseMaterial[0] = 0.0; glColor4fv(diffuseMaterial); glutPostRedisplay(); } break; case GLUT_MIDDLE_BUTTON: if (state == GLUT_DOWN) { /* change green */ diffuseMaterial[1] += 0.1; if (diffuseMaterial[1] > 1.0) diffuseMaterial[1] = 0.0; glColor4fv(diffuseMaterial); glutPostRedisplay(); } 411 Copyright Virtual University of Pakistan

43-Real-World and OpenGL Lighting

VU

break; case GLUT_RIGHT_BUTTON: if (state == GLUT_DOWN) { /* change blue */ diffuseMaterial[2] += 0.1; if (diffuseMaterial[2] > 1.0) diffuseMaterial[2] = 0.0; glColor4fv(diffuseMaterial); glutPostRedisplay(); } break; default: break; } } int main(int argc, char** argv) { glutInit(&argc, argv); glutInitDisplayMode (GLUT_SINGLE | GLUT_RGB | GLUT_DEPTH); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutMouseFunc(mouse); glutMainLoop(); return 0; } Try This Modify Example 8 in the following manner: Change the global ambient light in the scene. Hint: Alter the value of the GL_LIGHT_MODEL_AMBIENT parameter. Change the diffuse, ambient, and specular reflection parameters, the shininess exponent, and the emission color. Hint: Use the glMaterial*() command, but avoid making excessive calls. Use two-sided materials and add a user-defined clipping plane so that we can see the inside and outside of a row or column of spheres. if we need to recall user-defined clipping planes.) Hint: Turn on two-sided lighting with GL_LIGHT_MODEL_TWO_SIDE, set the desired material properties, and add a clipping plane. Remove all the glMaterialfv() calls, and use the more efficient glColorMaterial() calls to achieve the same lighting.

412 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

Lecture No.44
Evaluators

Evaluators, curves and Surfaces

A Bzier curve is a vector-valued function of one variable C(u) = [X(u) Y(u) Z(u)] where u varies in some domain (say [0,1]). A Bzier surface patch is a vector-valued function of two variables S(u,v) = [X(u,v) Y(u,v) Z(u,v)] Where u and v can both vary in some domain. The range isn't necessarily threedimensional as shown here. You might want two-dimensional output for curves on a plane or texture coordinates, or you might want four-dimensional output to specify RGBA information. Even one-dimensional output may make sense for gray levels. For each u (or u and v, in the case of a surface), the formula for C() (or S()) calculates a point on the curve (or surface). To use an evaluator, first define the function C() or S(), enable it, and then use the glEvalCoord1() or glEvalCoord2() command instead of glVertex*(). This way, the curve or surface vertices can be used like any other vertices - to form points or lines, for example. In addition, other commands automatically generate series of vertices that produce a regular mesh uniformly spaced in u (or in u and v). One- and twodimensional evaluators are similar, but the description is somewhat simpler in one dimension, so that case is discussed first. One-Dimensional Evaluators This section presents an example of using one-dimensional evaluators to draw a curve. It then describes the commands and equations that control evaluators. One-Dimensional Example: A Simple Bzier Curve The program shown in Example 1 draws a cubic Bzier curve using four control points, as shown in Figure 1.

Figure 1 : Bzier Curve

413 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

Example 1 : Bzier Curve with Four Control Points: #include <GL/gl.h> #include <GL/glu.h> #include <stdlib.h> #include <GL/glut.h> GLfloat ctrlpoints[4][3] = { { -4.0, -4.0, 0.0}, { -2.0, 4.0, 0.0}, {2.0, -4.0, 0.0}, {4.0, 4.0, 0.0}}; void init(void) { glClearColor(0.0, 0.0, 0.0, 0.0); glShadeModel(GL_FLAT); glMap1f(GL_MAP1_VERTEX_3, 0.0, 1.0, 3, 4, &ctrlpoints[0][0]); glEnable(GL_MAP1_VERTEX_3); } void display(void) { int i; glClear(GL_COLOR_BUFFER_BIT); glColor3f(1.0, 1.0, 1.0); glBegin(GL_LINE_STRIP); for (i = 0; i <= 30; i++) glEvalCoord1f((GLfloat) i/30.0); glEnd(); /* The following code displays the control points as dots. */ glPointSize(5.0); glColor3f(1.0, 1.0, 0.0); glBegin(GL_POINTS); for (i = 0; i < 4; i++) glVertex3fv(&ctrlpoints[i][0]); glEnd(); glFlush(); } void reshape(int w, int h) { glViewport(0, 0, (GLsizei) w, (GLsizei) h); glMatrixMode(GL_PROJECTION); glLoadIdentity(); if (w <= h) glOrtho(-5.0, 5.0, -5.0*(GLfloat)h/(GLfloat)w, 5.0*(GLfloat)h/(GLfloat)w, -5.0, 5.0); else glOrtho(-5.0*(GLfloat)w/(GLfloat)h, 5.0*(GLfloat)w/(GLfloat)h, -5.0, 5.0, -5.0, 5.0); glMatrixMode(GL_MODELVIEW); glLoadIdentity(); } int main(int argc, char** argv) { 414 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

glutInit(&argc, argv); glutInitDisplayMode (GLUT_SINGLE | GLUT_RGB); glutInitWindowSize (500, 500); glutInitWindowPosition (100, 100); glutCreateWindow (argv[0]); init (); glutDisplayFunc(display); glutReshapeFunc(reshape); glutMainLoop(); return 0; } A cubic Bzier curve is described by four control points, which appear in this example in the ctrlpoints[][] array. This array is one of the arguments to glMap1f(). All the arguments for this command are as follows: GL_MAP1_VERTEX_3 Three-dimensional control points are provided and three-dimensional vertices are produced 0.0 : Low value of parameter u 1.0 : High value of parameter u 3 : The number of floating-point values to advance in the data between one control point and the next 4 : The order of the spline, which is the degree+1: in this case, the degree is 3 (since this is a cubic curve) &ctrlpoints[0][0] : Pointer to the first control point's data Note that the second and third arguments control the parameterization of the curve - as the variable u ranges from 0.0 to 1.0, the curve goes from one end to the other. The call to glEnable() enables the one-dimensional evaluator for three-dimensional vertices. The curve is drawn in the routine display() between the glBegin() and glEnd() calls. Since the evaluator is enabled, the command glEvalCoord1f() is just like issuing a glVertex() command with the coordinates of a vertex on the curve corresponding to the input parameter u. 44.1 Defining and Evaluating a One-Dimensional Evaluator

The Bernstein polynomial of degree n (or order n+1) is given by

If Pi represents a set of control points (one-, two-, three-, or even four- dimensional), then the equation

415 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

represents a Bzier curve as u varies from 0.0 to 1.0. To represent the same curve but allowing u to vary between u1 and u2 instead of 0.0 and 1.0, evaluate

The command glMap1() defines a one-dimensional evaluator that uses these equations. void glMap1{fd}(GLenum target, TYPEu1, TYPEu2, GLint stride, GLint order, const TYPE*points); Defines a one-dimensional evaluator. The target parameter specifies what the control points represent, as shown in Table 1, and therefore how many values need to be supplied in points. The points can represent vertices, RGBA color data, normal vectors, or texture coordinates. For example, with GL_MAP1_COLOR_4, the evaluator generates color data along a curve in four-dimensional (RGBA) color space. You also use the parameter values listed in Table 1 to enable each defined evaluator before you invoke it. Pass the appropriate value to glEnable() or glDisable() to enable or disable the evaluator. The second two parameters for glMap1*(), u1 and u2, indicate the range for the variable u. The variable stride is the number of single- or double-precision values (as appropriate) in each block of storage. Thus, it's an offset value between the beginning of one control point and the beginning of the next. The order is the degree plus one, and it should agree with the number of control points. The points parameter points to the first coordinate of the first control point. Using the example data structure for glMap1*(), use the following for points: (GLfloat *)(&ctlpoints[0].x)

416 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

Table 1 : Types of Control Points for glMap1*()

More than one evaluator can be evaluated at a time. If you have both a GL_MAP1_VERTEX_3 and a GL_MAP1_COLOR_4 evaluator defined and enabled, for example, then calls to glEvalCoord1() generate both a position and a color. Only one of the vertex evaluators can be enabled at a time, although you might have defined both of them. Similarly, only one of the texture evaluators can be active. Other than that, however, evaluators can be used to generate any combination of vertex, normal, color, and texture-coordinate data. If more than one evaluator of the same type is defined and enabled, the one of highest dimension is used. Use glEvalCoord1*() to evaluate a defined and enabled one-dimensional map. void glEvalCoord1{fd}(TYPE u); void glEvalCoord1{fd}v(TYPE *u); Causes evaluation of the enabled one-dimensional maps. The argument u is the value (or a pointer to the value, in the vector version of the command) of the domain coordinate. For evaluated vertices, values for color, color index, normal vectors, and texture coordinates are generated by evaluation. Calls to glEvalCoord*() do not use the current values for color, color index, normal vectors, and texture coordinates. glEvalCoord*() also leaves those values unchanged. 44.2 Defining Evenly Spaced Coordinate Values in One Dimension

You can use glEvalCoord1() with any values for u, but by far the most common use is with evenly spaced values, as shown previously in Example 1. To obtain evenly spaced 417 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

values, define a one-dimensional grid using glMapGrid1*() and then apply it using glEvalMesh1(). void glMapGrid1{fd}(GLint n, TYPEu1, TYPEu2); Defines a grid that goes from u1 to u2 in n steps, which are evenly spaced. void glEvalMesh1(GLenum mode, GLint p1, GLint p2); Applies the currently defined map grid to all enabled evaluators. The mode can be either GL_POINT or GL_LINE, depending on whether you want to draw points or a connected line along the curve. The call has exactly the same effect as issuing a glEvalCoord1() for each of the steps between and including p1 and p2, where 0 <= p1, p2 <= n. Programmatically, it's equivalent to the following: glBegin(GL_POINTS); /* OR glBegin(GL_LINE_STRIP); */ for (i = p1; i <= p2; i++) glEvalCoord1(u1 + i*(u2-u1)/n); glEnd(); except that if i = 0 or i = n, then glEvalCoord1() is called with exactly u1 or u2 as its parameter. Two-Dimensional Evaluators In two dimensions, everything is similar to the one-dimensional case, except that all the commands must take two parameters, u and v, into account. Points, colors, normals, or texture coordinates must be supplied over a surface instead of a curve. Mathematically, the definition of a Bzier surface patch is given by

S (u, v) = Bin B m j (v ) P ij
i =0 j =0
where Pij are a set of m*n control points, and the Bi are the same Bernstein polynomials for one dimension. As before, the Pij can represent vertices, normals, colors, or texture coordinates. The procedure to use two-dimensional evaluators is similar to the procedure for one dimension. 1. Define the evaluator(s) with glMap2*(). 2. Enable them by passing the appropriate value to glEnable(). 3. Invoke them either by calling glEvalCoord2() between a glBegin() and glEnd() pair or by specifying and then applying a mesh with glMapGrid2() and glEvalMesh2(). 44.3 Defining and Evaluating a Two-Dimensional Evaluator

Use glMap2*() and glEvalCoord2*() to define and then invoke a two-dimensional evaluator. void glMap2{fd}(GLenum target, TYPEu1, TYPEu2, GLint ustride, GLint uorder, TYPEv1, TYPEv2, GLint vstride, GLint vorder, TYPE points); 418 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

The target parameter can have any of the values in Table 1, except that the string MAP1 is replaced with MAP2. As before, these values are also used with glEnable() to enable the corresponding evaluator. Minimum and maximum values for both u and v are provided as u1, u2, v1, and v2. The parameters ustride and vstride indicate the number of single- or double-precision values (as appropriate) between independent settings for these values, allowing users to select a subrectangle of control points out of a much larger array. For example, if the data appears in the form GLfloat ctlpoints[100][100][3]; and you want to use the 4x4 subset beginning at ctlpoints[20][30], choose ustride to be 100*3 and vstride to be 3. The starting point, points, should be set to &ctlpoints[20][30][0]. Finally, the order parameters, uorder and vorder, can be different, allowing patches that are cubic in one direction andquadratic in the other, for example. void glEvalCoord2{fd}(TYPE u, TYPE v); void glEvalCoord2{fd}v(TYPE *values); Causes evaluation of the enabled two-dimensional maps. The arguments u and v are the values (or a pointer to the values u and v, in the vector version of the command) for the domain coordinates. If either of the vertex evaluators is enabled (GL_MAP2_VERTEX_3 or GL_MAP2_VERTEX_4), then the normal to the surface is computed analytically. This normal is associated with the generated vertex if automatic normal generation has been enabled by passing GL_AUTO_NORMAL to glEnable(). If it's disabled, the corresponding enabled normal map is used to produce a normal. If no such map exists, the current normal is used. 44.4 Two-Dimensional Example: A Bzier Surface

Example 2 draws a wire frame Bzier surface using evaluators, as shown in Figure 2. In this example, the surface is drawn with nine curved lines in each direction. Each curve is drawn as 30 segments. To get the whole program, add the reshape() and main() routines from Example 1.

Figure 2 : Bzier Surface Example 2 : Bzier Surface: #include <GL/gl.h> #include <GL/glu.h> #include <stdlib.h> 419 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

#include <GL/glut.h> GLfloat ctrlpoints[4][4][3] = { {{-1.5, -1.5, 4.0}, {-0.5, -1.5, 2.0}, {0.5, -1.5, -1.0}, {1.5, -1.5, 2.0}}, {{-1.5, -0.5, 1.0}, {-0.5, -0.5, 3.0}, {0.5, -0.5, 0.0}, {1.5, -0.5, -1.0}}, {{-1.5, 0.5, 4.0}, {-0.5, 0.5, 0.0}, {0.5, 0.5, 3.0}, {1.5, 0.5, 4.0}}, {{-1.5, 1.5, -2.0}, {-0.5, 1.5, -2.0}, {0.5, 1.5, 0.0}, {1.5, 1.5, -1.0}} }; void display(void) { int i, j; glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glColor3f(1.0, 1.0, 1.0); glPushMatrix (); glRotatef(85.0, 1.0, 1.0, 1.0); for (j = 0; j <= 8; j++) { glBegin(GL_LINE_STRIP); for (i = 0; i <= 30; i++) glEvalCoord2f((GLfloat)i/30.0, (GLfloat)j/8.0); glEnd(); glBegin(GL_LINE_STRIP); for (i = 0; i <= 30; i++) glEvalCoord2f((GLfloat)j/8.0, (GLfloat)i/30.0); glEnd(); } glPopMatrix (); glFlush(); } void init(void) { glClearColor (0.0, 0.0, 0.0, 0.0); glMap2f(GL_MAP2_VERTEX_3, 0, 1, 3, 4,0, 1, 12, 4, &ctrlpoints[0][0][0]); glEnable(GL_MAP2_VERTEX_3); glMapGrid2f(20, 0.0, 1.0, 20, 0.0, 1.0); glEnable(GL_DEPTH_TEST); glShadeModel(GL_FLAT); } Defining Evenly Spaced Coordinate Values in Two Dimensions In two dimensions, the glMapGrid2*() and glEvalMesh2() commands are similar to the one-dimensional versions, except that both u and v information must be included. void glMapGrid2{fd}(GLint nu, TYPEu1, TYPEu2,GLint nv, TYPEv1, TYPEv2); void glEvalMesh2(GLenum mode, GLint i1, GLint i2, GLint j1, GLint j2);

420 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

Defines a two-dimensional map grid that goes from u1 to u2 in nu evenly spaced steps, from v1 to v2 in nv steps (glMapGrid2*()), and then applies this grid to all enabled evaluators (glEvalMesh2()). The only significant difference from the one-dimensional versions of these two commands is that in glEvalMesh2() the mode parameter can be GL_FILL as well as GL_POINT or GL_LINE. GL_FILL generates filled polygons using the quad-mesh primitive. Stated precisely, glEvalMesh2() is nearly equivalent to one of the following three code fragments. (It's nearly equivalent because when i is equal to nu or j to nv, the parameter is exactly equal to u2 or v2, not to u1+nu*(u2-u1)/nu, which might be slightly different due to round-off error.) glBegin(GL_POINTS); /* mode == GL_POINT */ for (i = nu1; i <= nu2; i++) for (j = nv1; j <= nv2; j++) glEvalCoord2(u1 + i*(u2-u1)/nu, v1+j*(v2-v1)/nv); glEnd(); or for (i = nu1; i <= nu2; i++) { /* mode == GL_LINE */ glBegin(GL_LINES); for (j = nv1; j <= nv2; j++) glEvalCoord2(u1 + i*(u2-u1)/nu, v1+j*(v2-v1)/nv); glEnd(); } for (j = nv1; j <= nv2; j++) { glBegin(GL_LINES); for (i = nu1; i <= nu2; i++) glEvalCoord2(u1 + i*(u2-u1)/nu, v1+j*(v2-v1)/nv); glEnd(); } or for (i = nu1; i < nu2; i++) { /* mode == GL_FILL */ glBegin(GL_QUAD_STRIP); for (j = nv1; j <= nv2; j++) { glEvalCoord2(u1 + i*(u2-u1)/nu, v1+j*(v2-v1)/nv); glEvalCoord2(u1 + (i+1)*(u2-u1)/nu, v1+j*(v2-v1)/nv); glEnd(); } Example 3 shows the differences necessary to draw the same Bzier surface as Example 2, but using glMapGrid2() and glEvalMesh2() to subdivide the square domain into a uniform 8x8 grid. This program also adds lighting and shading, as shown in Figure 3.

421 Copyright Virtual University of Pakistan

44-Evaluators, curves and Surfaces

VU

Figure 3 : Lit, Shaded Bzier Surface Drawn with a Mesh Example 3 : Lit, Shaded Bzier Surface Using a Mesh: void initlights(void) { GLfloat ambient[] = {0.2, 0.2, 0.2, 1.0}; GLfloat position[] = {0.0, 0.0, 2.0, 1.0}; GLfloat mat_diffuse[] = {0.6, 0.6, 0.6, 1.0}; GLfloat mat_specular[] = {1.0, 1.0, 1.0, 1.0}; GLfloat mat_shininess[] = {50.0}; glEnable(GL_LIGHTING); glEnable(GL_LIGHT0); glLightfv(GL_LIGHT0, GL_AMBIENT, ambient); glLightfv(GL_LIGHT0, GL_POSITION, position); glMaterialfv(GL_FRONT, GL_DIFFUSE, mat_diffuse); glMaterialfv(GL_FRONT, GL_SPECULAR, mat_specular); glMaterialfv(GL_FRONT, GL_SHININESS, mat_shininess); } void display(void) { glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT); glPushMatrix(); glRotatef(85.0, 1.0, 1.0, 1.0); glEvalMesh2(GL_FILL, 0, 20, 0, 20); glPopMatrix(); glFlush(); } void init(void) { glClearColor(0.0, 0.0, 0.0, 0.0); glEnable(GL_DEPTH_TEST); glMap2f(GL_MAP2_VERTEX_3, 0, 1, 3, 4, 0, 1, 12, 4, &ctrlpoints[0][0][0]); glEnable(GL_MAP2_VERTEX_3); glEnable(GL_AUTO_NORMAL); glMapGrid2f(20, 0.0, 1.0, 20, 0.0, 1.0); initlights(); } 422 Copyright Virtual University of Pakistan

45-Animations

VU

Lecture No.45

Animations

The passage of time has fascinated artists, scientists and theologians for thousands of years. Naturally they have attributed to it different interpretations, different implications and different conclusions. Nevertheless there seems to be general agreement on one aspect of time; that we are all conditioned by it and that, whether we like it or not, there is a time space into which we inevitably have to fit. Einstein, among other well known names in the world of science, made a special study of time in relation to his research in physics. His theory of relativity maintains that space and time are merely different aspects of the same thing. Since then other physicists have pointed out that objects can be moved backward and forward in space, but nothing can be moved back in time. Another method of describing the concept of time is through the three arrows of time. The first arrow is thermodynamic and can be seen operating when sugar dissolves in hot water. Second, is the historical arrow, whereby a single-celled organism evolves to produce more complex and varied species? The third is the cosmological arrow which is the theory that the universe is expanding from a big bang in the past. This cosmic expansion cannot be reversed in time. While the principle of relationships between the time arrows is still to be worked out on a scientific level, the actual application of it is constantly related to all work which utilizes it, such as music and the performing arts. In the latter it is one of the most important raw materials. In terms of animation, the idea of film time is one of the most vital concepts to understand and to use. It is an essential raw material which can be compressed or expanded and used for effects and moods in a highly creative way. It is, therefore, essential to learn and to understand how time can be applied to animation. The great advantage of animation is that the animator can creatively manipulate time since an action must be timed prior to carrying out the actual physical work on a film. It is also essential to understand how the audience will react to the manipulation of time from their point of view. Time sense or a sense of timing, therefore, is just as important as color sense and skill of drawing or craftsmanship in film animation. It has to be realized that while a performance on stage and on the screen requires a basic understanding of how timing works, this lecture is primarily confined to hand drawn animation which up to this point in film history still comprises 90% of all output in the animation medium. Timing for TV series For economic reasons, TV series are made as simply as possible from the animation point of view. This approach is generally known as limited animation. Animation is expensive, non-animation is cheaper. So to keep the films lively the plots are usually carried along by means of dialogue. It is often necessary to work with prerecorded blocks of dialogue which must remain intact. If this dialogue is well recorded for maximum dramatic effect, lengths of pauses between phrases 423 Copyright Virtual University of Pakistan

45-Animations

VU

cannot be changed (except within very narrow limits) without destroying that effect. In this case, the overall timing of long sections of the film is governed entirely by the dialogue. (There could be, however, considerable flexibility for more detailed timing within this fixed overall length.) The director has room to maneuver sections. So, if the total timing for all the recorded dialogue is subtracted from the required length for the whole film, this gives the amount of time that is available without dialogue. This can then be split up in the normal way and distributed throughout the film to give the best effect. Limited animation With limited animation as many repeats as possible are used within the 24 frames per second. A hold is also lengthened to reduce the number of drawings. As a rule not more than 6 drawings are produced for one second of animation. Limited animation requires almost as much skill on the part of the animator as full animation, since he must create an illusion of action with the greatest sense of economy. Full animation Full animation implies a large number of drawings per second of action. Some action may require that every single frame of the 24 frames within the second is animated in order to achieve an illusion of fluidity on the screen. Neither time nor money is spared on animation. As a rule only, TV commercials and feature-length animated films can afford this luxury. Animation is expensive and time-consuming. It is not economically possible to animate more than is needed and edit the scenes later, as it is in live-action films. In cartoons the director carefully pre-times every action so that the animator works within exact limits and does no more drawings than necessary. Ideally, the director should be able to view line test loops of the film as it progresses and so have a chance to make adjustments. But often there is no time to make corrections in limited animation and the aim is to make the animation work the first time. Timing for Animation in general Timing in animation is an elusive subject. It only exists whilst the film is being projected, in the same way that a melody only exists when it is being played. A melody is more easily appreciated by listening to it than by trying to explain it in words. So with cartoon timing, it is difficult to avoid using a lot of words to explain what may seem fairly simple when seen on the screen. Timing is also a dangerous factor to try to formulatesomething which works in one situation or in one mood may not work at all in another situation or mood. The only real criterion for timing is: if it works effectively on the screen it is good, if it doesn't, it isn't. So if having looked through the following pages you can see a better way to achieve an effect, then go ahead and do it! In this book we attempt to look at the laws of movement in nature. What do movements mean? What do they express? How can these movements be 424 Copyright Virtual University of Pakistan

45-Animations

VU

simplified and exaggerated to be made animatable and to express ideas, feelings and dramatic effects? The timing mainly described is that which is used in socalled classical or full animation. To cover all possible kinds of timing in all possible kinds of animation would be quite impossible. Nevertheless we hope to provide a basic understanding of how timing in animation is ultimately based on timing in nature and how, from this starting point, it is possible to apply such a difficult and invisible concept to the maximum advantage in film animation. Animation Principles What is good timing? Timing is the part of animation which gives meaning to movement. Movement can easily be achieved by drawing the same thing in two different positions and inserting a number of other drawings between the two. The result on the screen will be movement, but it will not be animation. In nature, things do not just move. Newton's first law of motion stated that things do not move unless a force acts upon them. So in animation the movement itself is of secondary importance; the vital factor is how the action expresses the underlying causes of the movement. With inanimate objects these causes may be natural forces, mainly gravity. With living characters the same external forces can cause movement, plus the contractions of muscles but, more importantly, there are the underlying will, mood, instincts and so on of the character who is moving. In order to animate a character from A to B, the forces which are operating to produce the movement must be considered. Firstly, gravity tends to pull the character down towards the ground. Secondly, his body is built and jointed in a certain way and is acted on by a certain arrangement of muscles which tend to work against gravity. Thirdly, there is the psychological reason or motivation for his actionwhether he is dodging a blow, welcoming a guest or threatening someone with a revolver. A live actor faced with these problems moves his muscles and limbs and deals with gravity automatically from habit, and so can concentrate on acting. An animator has to worry about making his flat, weightless drawings move like solid, heavy objects, as well as making them act in a convincing way. In both these aspects of animation, timing is of primary importance. Part of the working storyboard of The Story of the Bible by Halas and Batchelor. At this stage the director works out the smooth visual flow of the film, the editing, 425 Copyright Virtual University of Pakistan

45-Animations

VU

camera movements and so on. All these elements combine to tell the story in an interesting way. The storyboard A smooth visual flow is the major objective in any film, especially if it is an animated one. Good continuity depends on coordinating the action of the character, choreography, scene changes and camera movement. All these different aspects cannot be considered in isolation. They must work together to put across a story point. Furthermore the right emphasis on such planning, including the behaviour of the character, must also be realised. The storyboard should serve as a blueprint for any film project and as the first visual impression of the film. It is at this stage that the major decisions are taken as far as the film's content is concerned. It is generally accepted that no production should proceed until a satisfactory storyboard is achieved and most of the creative and technical problems which may arise during the film's production have been considered. There is no strict rule as to how many sketches are required for a film. It depends on the type, character and content of the project. A rough guideline is approximately 100 storyboard sketches for each minute of film. If, however, a film is technically complex, the number of sketches could double. For a TV commercial, more sketches are produced as a rule because there are usually more scene changes and more action than in longer films. The basic unit of time in animation The basis of timing in animation is the fixed projection speed of 24 frames per second (fps) for film and video. While other projection speeds have been used in the past the standard projection rate for film of all formats16mm, 35mm and 70mm remains 24 fps. On television and video this becomes 25 frames per second (PAL) or 30 fps (NTSC), but the difference is usually imperceptible. The thing to remember is that if an action on the screen takes one second it covers 24 frames of film, and if it takes half a second it covers 12 frames and so on. 24 frames of film go through the projector every second (25 on television). This fixed number of frames provides the basis on which all actions are planned and timed by the director. For single frame animation, where one drawing is done for each frame, a second of action needs 24 drawings. If the same action is animated on double frames, where each drawing is photographed twice in succession, 12 drawings are necessary out the number of frames and hence the speed of the action would be the same in both cases. 426 Copyright Virtual University of Pakistan

45-Animations

VU

Whatever the mood or pace of the action that appears on the screen, whether it be a frantic chase or a romantic love scene, all timing calculations must be based on the fact that the projector continues to hammer away at its constant projection rate. That is24 fps for film and either 25 fps or 30 fps for television and video depending on format. The unit of time within which an animator works is, therefore, 1/24 sec, 1/25 sec or 1/30 sec and an important part of the skill, which the animator has to learn is what this specific timing feels like on screen. With practice the animator also learns what multiples of this unit look like3 frames, 8 frames, 12 frames and so on. Animation and properties of matter The basic question which an animator is continually asking himself is: What will happen to this object when a force acts upon it? And the success of his animation largely depends on how well he answers this question. All objects in nature have their own weight, construction and degree of flexibility, and therefore each behaves in its own individual way when a force acts upon it. This behavior, a combination of position and timing, is the basis of animation. Animation consists of drawings, which have neither weight nor do they have any forces acting on them. In certain types of limited or abstract animation, the drawings can be treated as moving patterns. However, in order to give meaning to movement, the animator must consider Newton's laws of motion which contain all the information necessary to move characters and objects around. There are many aspects of his theories which are important in this book. However, it is not necessary to know the laws of motion in their verbal form, but in the way which is familiar to everyone, that is by watching things move. For instance, everyone knows that things do not start moving suddenly from resteven a cannonball has to accelerate to its maximum speed when fired. Nor do things suddenly stop deada car hitting a wall of concrete carries on moving after the first impact, during which time it folds itself rapidly up into a wreck. It is not the exaggeration of the weight of the object which is at the centre of animation, but the exaggeration of the tendency of the weightany weightto move in a certain way. The timing of a scene for animation has two aspects: 1. 2. The timing of inanimate objects The timing of the movement of a living character

With inanimate objects the problems are straightforward dynamics. How long does a door take to slam?, How quickly does a cloud drift across the sky?, How long does it take a steamroller, running out of control downhill, to go through a brick wall?. With living characters the same kind of problems occur because a character is a piece of flesh which has to be moved around by the action of forces on it. In addition, however, time must be allowed for the mental operation of the character, if he is to come alive on the screen. He must appear to be thinking his way through his actions, making decisions and finally moving his body around under the influence of his own will power and muscle. 427 Copyright Virtual University of Pakistan

45-Animations

VU

Animation consists of sequences of weightless drawings. The success of animation on the screen depends largely on how well these drawings give the impression of reacting in an exaggerated way when weight and forces are made to act upon them. Movement and caricature The movement of most everyday objects around us is caused by the effect of forces acting upon matter. The movement of objects becomes so familiar to us that, subconsciously, these movements give us a great deal of information about the objects themselves and about the forces acting on them. This is true not only of inanimate objects, but also of living thingsespecially people. The animator's job is to synthesize movements and to apply just the right amount of creative exaggeration to make the movement look natural within the cartoon medium. Cartoon film is a medium of caricature. The character of each subject and the movement it expresses are exaggerated. The subjects can be considered as caricatured matter acted upon by caricatured forces. Cartoon film can also be a dramatic medium. This particular quality can be achieved, among other means, by speeded up action and highly exaggerated timing. The difference between an action containing caricature, or humour or drama, may be very subtle. Eventually, with enough experience in animation timing, it becomes possible to emphasise the difference. Caricatured matter has the same properties as natural matter, only more so. To understand how cartoon matter behaves it is necessary to look more closely into the way matter behaves naturally. Cartoon is a medium of caricaturenaturalistic action looks weak in animation. Look at what actually happens, simplify down to the essentials of the movement and exaggerate these to the extreme. 428 Copyright Virtual University of Pakistan

45-Animations

VU

Cause and effect There is a train of cause and effect which runs through an object when it is acted upon by a force. This is the result of the transmission of the force through a more or less flexible medium (ie caricatured matter). This is one aspect of good movement in animation. An animator must understand the mechanics of the natural movement of an object and then keep this knowledge in the back of his mind whilst he concentrates on the real business of animation. This is the creation of mood and conveying the right feeling by the way an action is done. Examples of cause and effect: Figs A and B A rope wrapped around anything and pulled tight has the tendency shown. How far the reaction goes depends on: i. ii. the strength of the forces pulling the rope. the flexibility or rigidity of the object being squeezed. Exaggerate the tendency.

429 Copyright Virtual University of Pakistan

45-Animations

VU

A A rope something.

wrapped

round

B The ends of the rope are pulled tight. Fig. C On the seesaw, the end of the plank with the smaller stone tends to stay where it is to begin with because of its inertia, so bending plank D. A moment later it starts to accelerate and the plank springs back with the opposite curvature causing the stone to whizz out of the screen, E.

C A stone on one end of a seesaw. D and E A bigger stone drops onto the other end. Newton's laws of motion Every object or character has weight and moves only when a force is applied to it. This is Newton's first law of motion. An object at rest tends to remain at rest until a force moves it and once it is moving it tends to keep moving in a straight line until another force stops it. The heavier an object is, or strictly speaking, the greater its mass, the more force is required to change its motion. A heavy body has more inertia and more momentum than a light one. A heavy object at rest, such as a cannonball, needs a lot of force to move it (following Fig. A). When fired from a cannon, the force of the charge acts on the cannonball only whilst it is in the gun barrel. Since the force of 430 Copyright Virtual University of Pakistan

45-Animations

VU

the explosive charge is very large indeed, this is sufficient to accelerate the cannonball to a considerable speed. A smaller force acting for a short time, say a strong kick, may have no effect on the cannonball at all. In fact it is more likely to damage the kicker's toe. However, persistent force, even if not very strong, would gradually start the cannonball rolling and it would eventually be travelling fairly quickly. A A cannon ball needs a lot of force to start it moving. Once moving, it takes a lot of stopping. Once the cannonball is moving, it tends to keep moving at the same speed and requires some force to stop it. If it meets an obstacle it may, depending on its speed, crash straight through it. If it is rolling on a rough surface it comes to rest fairly soon, but if rolling on a smooth flat surface, friction takes quite a long time to bring it to rest. When dealing with very heavy objects, therefore, the director must allow plenty of time to start, stop or change their movements, in order to make their weight look convincing. The animator, for his part, must see that plenty of force is applied to the cannonball to make it start, stop or change direction. Light objects have much less resistance to change of movement and so behave very differently when forces act on them. A toy balloon (Fig. B) needs much less time to start it moving. The flick of a finger is enough to make it accelerate quickly away. When moving, it has little momentum and the friction of the air quickly slows it up, so it does not travel very far. B A balloon needs only a small force to move it, but air resistance quickly brings it to rest. In both these examples a circle is being animated. The timing of its movements can make it look heavy or light on the screen. The way an object behaves on the screen, and the effect of weight that it gives, depends entirely on the spacing of the animation drawings and not on the drawing itself. It does not matter how beautifully drawn the cannonball is in the static sense, it does not look like a cannonball if it does not behave like one and the same applies to the balloon, and indeed to any other object or character. 431 Copyright Virtual University of Pakistan

45-Animations

VU

Fast run cycles An eight frame run cyclethat is four frames to each stepgives a fast and vigorous dash. At this speed the successive leg positions are quite widely separated and may need drybrush or speed lines to make the movement flow. Drawing 5 shows the same position as drawing 1 but with opposite arms and feet. Similarly drawings 6 and 2, 7 and 3, and 8 and 4 show the same positions. These alternate positions should be varied slightly in each case, to avoid the rather mechanical effect of the same positions occurring every four frames. A twelve frame cycle gives a less frantic run, but if the cycle is more than sixteen frames the movement tends to lose its dash and appear too leisurely.

The body normally leans forward in the direction of movement, although for comic effect a backward lean can sometimes work. If a faster run than an eight frame repeat is needed, then perhaps several foot positions can be given on each drawing, to fill up the gaps in the movement, or possibly the legs can become a complete blur treated entirely in dry-brush. In the first example, drawing 4 is equivalent to the step position in a walk, with the maximum forward and backward leg and arm movement. In a run it is also the point at which the centre of gravity of the body is farthest from the ground, that is, 432 Copyright Virtual University of Pakistan

45-Animations

VU

in mid-stride. In drawing 1 the weight is returning to its lowest point, which is in drawing 2. In drawing 3 the body starts to rise again as the thrust of the back foot gives the forward impetus for the next stride. These are both examples of eight frame run cycles. This means four drawings to each step. Drawings 1 and 5 show the same leg and arm positions but with opposite feet, and so do 2 and 6, 3 and 7, 4 and 8. In such a short cycle these positions should be varied slightly to avoid a mechanical effect. Timing and music Ever since the very first animated productions, Disney's Steamboat Mickey and Fischinger's abstract film Brahm's Hungarian Dances, it was clear that there is a strong relationship between animation and music. This relationship can be explained on two accounts. First, both elements have a basic mathematical foundation and move forward at a determined speed. Second, since animation is created manually frame by frame, it can be fitted to music in a very exact manner. It is further able to capture its rhythm, its mood and hit the beat right to the frame. Most animation makes good use of this advantage. In general principle it is more difficult to follow the rhythm of a musical composition with its mood than its beat. The latter aspect of the music is easily measured, since beats are fitted into bar units of defined time length and are interpreted in time units. Bars can contain various numbers of beats and these must be measured to the film frame. Having done this, it is comparatively easy to fit the animation to the speed of the beat and find the right type of movement to follow the music, whether it is a slow waltz of 36 frames, or 4 frames for rock music. A beat can be emphasised by synchronisation of the feet but it works better if the whole body is used. In quick beats of 3,4 or 6 frames it is possible to follow every second beat without losing the rhythm. It is always better to work to specially prepared music if this can be afforded.

A This eight frame cycle, animated on double frames, of a whirling Spanish dancer is fitted closely to strong flamenco music. The figure fully conveys the character of the music with all its functional simplicity.

B Single and double frame animation are alternated to fit the beats of the sound of a Spanish guitar. It is essential for the movement to follow the musical lead of a specially 433 Copyright Virtual University of Pakistan

45-Animations

VU

prerecorded music track, for accurate synchronisation.

Camera movements Tracks are used to move into a closer field or pull back to a more distant one. They are done by moving the camera frame by frame, up or down its vertical pillar, above the animation drawings. Usually the field centre also moves during a track and as the camera travels on a fixed axis this movement in a north south, east west direction is done by moving the table. Tracks and table moves are worked out in terms of general timingtrack lengths and the action which the various fields must includeby the director on bar sheets before production begins. When the animator finalises the action of the scene in detail, he converts the director's timing into specific instructions to the cameraman. He writes down the field sizes and marks the frames where camera movements start and stop in the camera instructions column on the exposure chart (Fig. A). He also provides a drawn field key with field centres marked (Fig. B). It then becomes the cameraman's responsibility to achieve the required effect smoothly and accurately on the screen. Briefly, the procedure is as follows: in Fig. B the track is made from field X to field Y so the screen centre moves towards the south-east. This means that under the camera, the table must move north-west. Fig. C is an enlargement of this table move, showing how the cameraman divides the line to achieve a smooth movement from X to Y. At the same time he measures the distance the camera travels on its column during the track, and divides this 434 Copyright Virtual University of Pakistan

45-Animations

VU

in exactly the same way as Fig. C, so that camera and table top move smoothly together. Fig. D is a similar track and table move which includes a tilt. This would also be done as a table move. Tracks and table moves are usually animated on single frames. A The director's timing of tracks is finalised by the animator in the camera instructions column on the exposure sheet. B The accompanying field key. C Enlargement of field centres from Fig. B, inverted for use as table move. D Another example of a track including a table tilt.

435 Copyright Virtual University of Pakistan