Turčić et al.: Individual stochastic screening...

, acta graphica 22(2011)3-4, 69-78

udc 004.92:519.216:655
original scientific paper
received: 22-08-2011
acta graphica 196 accepted: 22-11-2011

Individual Stochastic Screening

for the Development of Computer Graphics

Authors
Maja Turčić¹*, Vilko Žiljak², Ivana Ž.–Stanimirović²
Polytechnic of Zagreb,
1
University of Zagreb,
2

Zagreb, Croatia Faculty of Graphic Arts,

*E-mail: [email protected] Croatia

Abstract:
With the emergence of new tools and media, art and design have developed into
digital computer-generated works. This article presents a sequence of creating art
graphics because their original authors have not published the procedures. The
goal is to discover the mathematics of an image and the programming libretto with
the purpose of organizing a structural base of computer graphics. We will elaborate
the procedures used to produce graphics known throughout the history of art, but
that are nowadays also found in design and security graphics. The results are close-
ly related graphics obtained by changing parameters that initiate them. The aim is
to control the graphics, i.e. to use controlled stochastic to achieve desired solutions.
Since the artists from the past have never published the procedures of screening
methods, their ideas have remained “only” the works of art. In this article we will
present the development of the algorithm that, more or less successfully, simulates
those screening solutions. It has been proven that mathematically defined graphi-
cal elements serve as screening elements. New technological and mathematical so-
lutions are introduced in the reproduction with individual screening elements to
be used in printing.

Keywords:
Stochastic Models, Computer Art Graphics, Screen Shapes, Security Graphics,
Design and Print

69
 Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

1. Introduction ing of more images gave impetus to new ideas

and media. In mid 1970s, Tomislav Mikulić
Computer technology “sneaked” into hu- could draw only static monochrome lines us-
man lives in 1960s as a luxury intended for ing a graphical terminal, but with a camera and
individuals and corporations and not as a tool a film he developed and created animation for
for general use and expression. Owing to cer- TV shows (Mikulić, 1980). Soon integration of
tain people who had opportunity to experi- commuter graphics into movies, cartoons and
ment, various branches of computer art and commercials took place. Nowadays, owing to
design have been developed. Computer graph- the wide availability of personal computers
ics are responsible for changes in the theory and programmes containing routines enabling
of art, education, entertainment and environ- easier use and expression, animation and 3-D
mental design (Franke, 1985). Computer art graphics are not a problem to create. However,
was a logical step in developing the means of 3-D graphics are also based on the same or very
graphic expression because it offered solutions similar algorithms of two-dimensional graph-
not achievable by means of analogue meth- ics. The 3-D technology has become a stand-
ods. A computer has become a new tool that ard; it has culminated in the film industry and
is constantly changing creative processes and is now available at home as well.
completely renewing the concept of artistic
expression (Nierhoff-Wielk, 2007). Computer In the same way, due to the availability of
graphics tools have created the terms of vec- personal computers and software, the satiation
tor and pixel graphics that present the basis of of ideas has occurred. It is inevitable for those
contemporary printing industry in processing working with programming graphics to use
text and images. the already existing algorithms that have been
proved the most effective and the simplest, for
Computer graphics do not indicate one au- instance circles, sine curves, Bézier curves and
thor or their style, since programming in vari- similar. The difference lies is the unique proce-
ous computer languages can result in the same dure for creating forms.
visual work coming from various authors.
Each work of art or design is an original, i.e. Nowadays, all aspects of graphic industry:
individual, under the assumption that each layouting and editing, the entire DTP proce-
image or design is a reflection of individuality dures, reprophotography, the preparation and
and uniqueness of the author and their creative production of printing plates are based on
thinking process, i.e. that the author’s unique computers, computer software and computer
style is recognisable in each image. tools. Tools currently used in graphic industry
such as Adobe Photoshop, InDesign and Illus-
Although individuals of the art movement trator have developed from the procedures and
Nove tendencije (New Tendencies) fought with experiments of artists/ scientists from 1960s.
numerous critiques of their visual experiments,
their ideas have become the foundations for
generating new and progressive ideas that are
nowadays considered to be common. From the
basic computer graphics whose creation, due
2. Nove Tendencije
to slow computers and inadequate computer (New Tendencies)
languages, interface or printers, used to take
days, a need for animation, i.e. moving image In 1960s, the members of the art move-
arose. The integration of the element of time ment New Tendencies started creating algo-
into graphics programming and the combin- rithmic works of art, thus becoming the pio-

70
Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

neers of such art in Croatia. First they made Many artists have in various ways tried to
their works using analogue methods, but as arrange screening elements and their cover-
technology developed, they themselves started age (Vladimir Bonačić-1969, Frieder Nake-1965,
using machines and programme languages Georg Nees-1965-1968, A. Michael Noll-1965,
to write in. They blended art and science and Manfred Robert Schroder-1968, Leon D. Har-
introduced the concept of research into their mon/Kenneth C. Knowlton-1966, David R. Gar-
works (Rosen, 2011). At that point the viewer rison-1968/69, Vilko Žiljak-1972) (Rosen, 2011).
could not longer tell the difference between an Technologies of phototypesetting from 1980s
artist and a scientist. Those works of art con- had relations in which an entire word could
tain in themselves recognisable routines that be used as a screening microelement (Žiljak,
detach them from personality and individu- 1987). Artists used to solve the problem with
ality. The first exhibition was held in 1961 in instances of individual expression and without
Zagreb showing experimental works of artists/ publishing procedures or algorithms for the
scientists that experiment with the visual and suggested technique. They did not go beyond
participate in the creation of a new dominant the mere presentation of their work of art.
art movement. New tendencies continued into
early 1970s, during which they showed efforts
to systematise and study the possibilities, thus
making their art primarily different from mere
“attempts”. Experiments are defined through
the methodology and planning of conditions
that are then studied. They resulted in repeat-
able works. This hypothesis is opposed to the
traditional concept of an artist as a spontane-
ous genius expressing themselves through art
(Moles, 1968).

They build algorithms through systemic

research and create functionality by defining
boarders (Klutsch, 2005). The concept fell apart
in 1970 and one of the participants, Frieder
Nake, concluded that computer graphics pro-
duction was repeatable and that new, “really Figure 1. Manfred Robert Schroder, Eye II, 1968
good ideas haven’t shown up for a while” (Nake,
1970). On the other hand, the fact is that the
entire graphic industry is based on computer
graphics; it has changed the industry, resulting
in new technologies such as digital printing.

In this article we will present different ways

of possible program development providing
the same visual results although coming from
different authors and programming languages,
if that program logics is studied and under-
stood. We create new algorithms resulting in
similar solutions to those of other authors in
history aimed at determining algorithms and
their upgrades. Figure 2. Eye by M. Turčić, 2011

71
 Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

PImage img;
int charCounter=0;
String s = “ONE PICTURE IS WORTH A THOUSAND WORDS “;
void draw()
{
noLoop();
img.loadPixels();
int x=0;
int y=0;
int rowCounter=0;
while(y<height)
{
while(x<width)
{
char currentChar=s.charAt(charCounter);

color col=img.pixels[y*width+x];
fill(col);
text(currentChar,x,y);
x+=textWidth(currentChar);
charCounter++;
if(charCounter>=s.length()) charCounter=0;
}
charCounter=0;
rowCounter++;
y+=8;
x=-rowCounter*360;
}
}

Figure 3. Gotthart Muller 64/6, 1964 Figure 4. Version by M.Turčić, 2011

72
Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

3. Individuality and Development Mass application of computer graphics is

a consequence of the lack of originality, either
Computer graphics bring about the loss of voluntary or unconscious. Tools, nowadays
individuality due to the lack of recognizability used in the applied computer graphics technol-
of an individual author’s style, which is particu- ogy, create units and become routines used to
larly noticeable in its fractal structure. The term create individual solutions for requirements of
‘fractal’ was first introduced by Benoit Man- the advertising industry. If a new graphic solu-
delbrot in 1975 to represent complex structures tion is created, it evokes the need to create new
based on a simple recursion, i.e. repetition. solutions and that way the hypothesis of expo-
Although each graphic is unique, it does not nential development of computer graphics in
express individuality because of the possibility the world is confirmed. The ultimate result of
that the same result is programmed by another these processes is the anonymity of authors be-
author (Čerić, 2008). cause individuals will always have avant-garde
ideas and their realisation will inspire others to
However, computer graphics as we know similar processes (either by stealing or by their
them today and that have been developing extension), eventually becoming subroutines,
since 1960s, prove that an image does not rep- effects and tools in the contemporary graphics
resent an author, because it can be generated software.
and is therefore repeatable. If an author uses
fractals or some known algorithms or recog- The aim is to create one’s own base of routines
nises routines in somebody’s work of art, even that are upgraded to constantly create new tools,
if they add their ideas to the original setting, i.e. packages of routines for graphics which are
they are not an original generator of that art to be individualised. The aftermath of our base
or design. In that way we question the unique- is a detachment from standardised work condi-
ness of graphic design that is derived from rou- tions indicating close relations of science and
tines and subroutines of computer software and art in the methodology leading to individual
reaches the solutions in the same way. Due to digital reproductions.
the possibility of repeating a work of art, meth-
ods to generate computer graphics destroy the
ancient presumption of the recognizability of
an author.
4. Stochastic and anti-stochastic
After forty years of developing methods and
programming languages in computer graphics, Stochastics provides new solutions that are
the methods have become standardised. The normally not possible due to the human Ge-
best example of repetition are fractal graph- stalt need for grouping elements. It generates
ics that have been named (Mandelbrot, Koch, unpredictable shapes, crucial for the vivacity of
Nova, Julia set, Brown’s tree etc.) and in that design, and presents the basis of imitating in-
way standardised and made ready for use by dividuality. Stochastics functions as a simulator
designers and artists. Fractal graphics have a of certain processes and implies the planning
pseudo-stochastic generator making a “sto- and its opposites. Anti-stochastics introduces
chastic image” with set borders within which the ability to control, among other, the return
it widens, develops, i.e. applies anti-stochastic into the original condition, thus creating “a
principles. The problem of Moire was resolved flow” of generated elements functioning within
in the printing industry by introducing fre- them and preventing the work to disperse. An-
quency modular screens. That is a symbiosis ti-stochastics brings order into distortions that
of the coverage of an individual pixel through are created by stochastics, imposing predictable
a random disposition of microelements of the borders within which stochastic distorted signs,
coverage of the same pixel (Žiljak & Pap, 1999). shapes and elements are generated. In the ana-
logue era, works used to inevitability are differ-

73
 Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

ent each time due to human errors, but the digi- sequence, we have built our own generators that
tal technology has introduced sterile uniformity are congruent due to their speed and simplicity,
of products. At the same time, a computer is an as well as high quality results. They are based on
ideal tool for generating stochastic variations, finding the remainder of the product of the two
hence providing freedom and the richness of preceding numbers. Particular attention should
expression, impossible for digital solutions. The be paid to the degeneration, i.e. too frequent
best examples are handwritten fonts that can- repetition of sequences or the transition to one
not look authentic without a stochastic variable. constant. Therefore, the best results are obtained
This was demonstrated by Donald E. Knuth if numbers ‘a’ and ‘m’ are prime numbers (Žiljak
in his Metafont Project that defined letters and & Smiljanić, 1980).
signs by mathematical curves and by introduc-
ing stochastics, all curves were changed leaving ri+1 = (ari + b) mod m (1)
the impression of authenticity (Knuth, 1986).
Congruent sequences always end in a loop;
Stochastics in computer programming is there is always a sequence that is repeated end-
based on random number generators. Some lessly. If b = 0, that sequence is shorter, but it is
programming languages themselves generate still possible to achieve that the sequence is long
parameters of congruence, and in some there enough. The sequence cannot be longer than
is not even an option for the author to define the amount of value m. The best result, i.e. the
seed, i.e. the seed of generator. But in our at- length of sequence, is obtained by using prime
tempt to achieve the utter control of a random numbers (Knuth, 1997).

a)

b)

Figure 5. a) portrait font, the original version and b) a version processed by algorithm

74
Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

Using the available generator within a pro- An individual test of pseudorandom se-
gramming language (random, math. random, quence in the Processing language:
rand...) the same sequence of numbers is always
// congruent method of
obtained because of their pseudo character. In
generating random numbers
the attempt to get diverse results, except for
control, independent generators are needed to float a = 2311;
create the rhythm and add new interventions float m = 914189; //the
into the existing graphics. In computer graph- biggest number
ics, random number generators are used for
float s = 913373; //sjeme
different purposes (for colours, dimensions,
angles…) that do not have the same occurrence int i;
period; and that period can also be random. In float rn;
order to control the parameters of newly intro- float f;
duced segments of programs, we need to ensure
the same sequences at the previous pseudo ran- //for the range
dom sequences. The only way to obtain com- float x1 = -5;
pletely controlled solutions is introducing a few float x2 = 5;
independent pseudo random generators. All
for (i=1; i<50; i++)
our generators have a range from 0 to 1, but each
one has a different initiator. {
s = (a*s) %m;
The numbers in Table 1 leads to the conclu-
//the range from 0 to 1
sion that the length of the sequence obtained
by a congruent method depends on the choice rn= s/m;
of numbers. Better results are achieved using f = rn * (x2-x1)+x1;
prime numbers (especially for ‘a’ and ‘m’); how- println(f);
ever that is not the only factor determining a
generator’s efficiency. }
Table 1: examples of generators and their efficiency in relation to the selected numbers
a S m Sequence length %
7 13 23 10 43
37 19 23 22 95
7 13 24 3 12
2311 41077 96553 32184 33
2311 913373 914189 228546 24
2311 41077 96558 66 0,1
2311 101 37313 9328 25
67 101 193 32 16
401 101 193 192 99
5431 3373 7547 7546 99
5431 3374 7547 5071 67
5431 3373 7542 1254 16
9 25 31 15 48
9 31 73 6 8
9 277 439 73 16
9 53 439 73 16
9 53 331 165 50
227 53 331 330 99
227 56 331 330 99

75
 Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

The backbone of this animation are letters ments in the shape of lines. New screening ele-
from the Streetvertizing font (http://www.clean- ments can already result from the graphic itself
vertising.com/) and Processing library Geomera- and do not have to be created from screening
tive (http://www.generative-gestaltung.de), cells. The example showing solutions to the
used to transfer font characters into a certain glyph already proves that there is a new field of
number of curves, i.e. segments. Afterwards, the screening technology. Dissipation and multipli-
end points of these curves are used as starting cation of these screening elements resolves the
points of new graphic elements, in this instance problem of Moire in its own way, if the same let-
lines. The target (finishing point) of these lines ter were interpreted in multicoloured print. In
is determined using an own random number order to achieve, for instance, a brown hue of
generator and by each pass through the loop, the graphic 6th letter, a 6/6 would preset a yellow
the letter starts to be drawn again, but the num- channel, letter a 6/4 a magenta channel, letter
ber range that the generator uses is in each pass a 6/3 a cyan channel. Channel K would not be
bigger, resulting thus in the greater length of needed. The same principle is used in the spot
lines and bigger angles compared to the start- multicolour system pairing each colour with
ing point. Two independent generators are used, its own solution. The length of pseudo random
one for the x and the other for the y axis. The generator determining the coverage, and the
animation shows alterations of one parameter in seed is different for various channels in order to
this graphic, and that is the range of numbers of prevent the repetition of the same graphic posi-
the random number generator. tions in various channels. Similarly to the way in
which these examples generate a screening ele-
These algorithms for generating graphic ele- ment as a part of a straight line, this discussion
ments are targeted at the new area of defining can be expanded to any other graphic element,
screening technology. The example of a lower such as a square, an ellipsis, an arc, a circle or a
case letter ‘a’ shows coverage with microele- Bézier curve.

// Animation of letters thorough segmentation and own random

number generator in Processing
void draw() {
noFill();
pushMatrix();
strokeWeight(0.1);
stroke(75,0,130);
beginShape(LINES);
for (int i = 1; i < pnts.length; i++ ) {
seed = (mul * seed ) % modul;
ran = (seed /div)*(minus)-(minus/2);
seed2 = (mul * seed2 ) % modul;
ran2 = (seed2 /div)*(minus)-(minus/2);
vertex(pnts[i-1].x, pnts[i-1].y);
vertex(pnts[i].x + ran, pnts[i].y + ran2);
}
endShape(CLOSE);
popMatrix();
div-=1000;
minus+=1;
}

76
Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

Figure 6. random animation of letter a

Graphic technology is based on solving cov- ing and human behaviour has also changed.
erage through various mathematical screening Programming enables the obtaining of a series
methods. The artists from the age of the earliest of images by changing parameters that can lead
computer graphics were the first to come across to animations. New methods of expression such
the problem. The creation of new screening el- as music or words can be generated by comput-
ements for solving coverage is based on com- ers and in that way they expand the forms of ex-
puter graphics developed in 1960s. After the use pression. The acceptance of new methods does
of cameras in reprographics (and net screen- not threaten the “classical” art techniques, but it
ing) was abandoned, all technology of solving develops new functions of stimulating creative
coverage and screening problems was derived capacities and sensitizing viewpoints. In com-
from computer graphics. The focus was placed parison with traditional techniques, computer
on finding new solutions in design and solving offers a different approach and generates results
problems of graphic reproduction by means of emphasizing cognitive processes, creativity and
computer methods. imagination rather than physical skills and tal-
ent. Printing technology has introduced math-
ematical definitions of screening elements. Au-
thors of those programs have suggested many
procedures and forms based on PostScript de-
5. Conclusion scription of screening cells. We have presented
a proposition of screening using simple math-
in its beginnings, computer graphics were ematical forms. Screening solutions of early
based on geometrical and mathematical forms computer graphics were compared which no
and were therefore losing support of critics and analogue mathematical interpretation was pos-
audiences. Those beginnings provoked the de- sible for. Preconditions for the systematisation
velopment of unknown spheres of graphically of screening elements as a basis for new forms
attractive forms, i.e. new aesthetic area that does have been created.
not even have to be named art. The way of think-

77
 Turčić et al.: Individual stochastic screening..., acta graphica 22(2011)3-4, 69-78

References

Nierhoff-Wielk, B., 2007. Ex Machina - Žiljak V. Smiljanić G., 1980, Modeliranje

The Encounter of Computer and Art. A i simuliranje sa računalom, Elektrotehnički
Look Back. In: W. Herzogenrath and B. Ni- fakultet u Zagrebu, Liber
erhoff-Wielk, eds. 2007. Ex Machina - Early
Computer Graphics up to 1979. The Franke Žiljak, V.,1987. Računarska Tipografija, Za-
collections and other foundations at the greb: Školska Knjiga
Kunsthalle Bremen. For the 80th birthday of
Herber W. Franke 1979, Bremen: Deutscher Žiljak V., Pap K. 1999. Mathematical model
Kunstverlag, pp. 22-23, ISBN 978-3-422- of stochastic curves in digital printing, In:
06689-2 26th International Research Conference of
IARIGAI, Munchen, Germany september
Klutsch,C., 2005. The Summer 1968 in Lon- 5-8 1999.
don and Zagreb: Starting or End Point for
Computer Art. In: C&C ‘05 Proceedings of Knuth, D.E. 1986. The Metafont book, Boston,
the 5th conference on Creativity & cogni- MA, USA: Addison-Wesley Longman Pub-
tion, London, United Kingdom 12-15 April lishing Co., Inc., ISBN 0-201-13444-6
2005, NewYork: ACM, ISBN 1-59593-025-6,
Available at: <http://www.computerkunst. Knuth, D.E., 1997. The Art of Computer Pro-
org/Kluetsch_London_Zagreb.pdf> [Ac- gramming 3rd ed Vol 2, pp 10-25, Reading,
cessed 3 August 2011] Massachusetts: Addison-Wesley, ISBN
0-201-89683-2
Franke, W.F., 1985. The New Visual Age: The
Influence of Computer Graphics on Art and Rosen, M., 2011. The Art of Programming.
Society, Leonardo, 18(2), pp 105-107, Cam- In: M. Rosen ed. 2011. A little known story
bridge: MIT Press, Available at: < http:// about a movement, a magazine, and the
dam.org/essays/franke-the-new-visual- computer’s arrival in art: New Tendencies
age-1985> [Accessed 3 August 2011] and Bit International, 1961-1973, Germany:
ZKM Publications, Center for art and me-
Čerić, V., 2008, Algorithmic Art: Technology, dia Karlsruhe, pp. 27-40, ISBN 978-0-262-
Mathematics and Art. In: invited lecture at 51581-8
the Information Technology Interface ITI
2008 30th International Conference, Cavtat, Mikulić, T., 1980, Computer animation, exi-
Dubrovnik 23-26 June 2008, pp 75-81, ISBN: bition at: Galerija Nova, centar za kulturnu
978-953-7138-12-7, Available at: <http:// djelatnost sso, zagreb, mihanovićeva 28/I,
www.vceric.net/texts/mathmod09_algo- 16. june 1980. Available at: <http://tomislav-
rithms_mathematics_and_art.pdf> [Ac- mikulic.com/1980-animacija-100.pdf> [Ac-
cessed 3 August 2011] cessed 3 August 2011]

Moles, A., 1968, Introduction, Kelemen, B. Nake, F., 1970, “Statement for PAGE” , PAGE
Putar, R. eds. Bit Interantional 3, Interna- 8 bulletin of the computer arts society, may
tional Colloquium on Computers and Vis- 1970, Gustav Metzger ed., London Available
ual Research, Zagreb August 3-4 1968, pp at: <http://www.bbk.ac.uk/hosted/cache/
3-10, Zagreb: Galerije grada Zagreba archive/PAGE/PAGE8%20May%201970.
pdf> [Accessed 3 August 2011]

78