Colorimetric and Resolution requirements of cameras

Alan Roberts

ADDENDUM 23 : Sony HVR-V1E

Data for this addendum is taken from a short examination of one production model of the Sony V1E HDV
camcorder, physically very similar to (but smaller and lighter than) the HVR Z1 and FX1 HDV camcorders.
It has 3 16:9 1”/4 cmos sensors of 1440x810 pixel dimensions (4.5mm diagonal). It records HDTV using the
HDV algorithm onto standard mini DV tapes (1080i and 1080psf), and SDTV using DVCAM format.
The camera is relatively light (about 1.5kg excluding battery) and has an integral lens and viewfinder, with
side lcd panel, and seems aimed at the middle to high-end consumer/low-end professional market rather than
full broadcast, which would normally demand interchangeable lenses and better control.
It has internal menus for setting the performance, not as complex as in a full broadcast camera, but enough to
control some of the important features, albeit only in “on/off” states. It is not suited to multi-camera
operation. It has analogue-only video outputs (components and SD-composite via a multi-pin connector and
S-video SD) and digits via IEEE1394 Firewire (called i.Link by Sony). This alone puts the camera in the
consumer/semi-pro market, rather than broadcast, which would normally expect either HDSDI or BNC
connectors for analogue.
The same assessment procedure was used as for other HD cameras, partly attempting to get a good “film-
look”, and the settings reflect that. However, because of the lack of internal test signals, and as a result of
some of the initial measurements, that approach was quickly abandoned and efforts were directed at getting a
decent colorimetric performance from the camera. Assuming that a grading operation will be used in post-
production, the settings attempt to give the colourist a reasonable exposure range, but this is inevitably well-
short of what a film stock could be expected to deliver.

1
 Colorimetric and Resolution requirements of cameras
Alan Roberts
Addendum 23: Sony-HVR V1E

Data for this is taken from a short examination of a production model of the Sony HDV camcorder, HVR-
V1E, with three 1”/4 cmos sensors (4.5mm diagonal, each approximately 1440x810). It records in HDV
(1080i/25, 50Hz interlaced) format onto miniDV tapes, and standard definition (576i/25) as DVCAM.
The camera is essentially a pro-sumer model, with some professional features such as having XLR
connectors for audio input. It has an integral lens (Zeiss Sonnar, 3.9~78mm, 20:1, F/1.6 maximum aperture
ramping to F/2.8) and viewfinder, with side lcd panel, and seems aimed at the consumer and low-end
professional market rather than broadcast, which would normally demand interchangeable lenses and higher
resolution sensors, together with greater control through the menus. Minimum exposure is claimed to be 4
lux.
The camera has internal menus for setting the performance and a reasonable selection of external controls.
There are analogue-only video outputs (components and composite plus S-video at SD, all via multi-pin
connectors) and digits via IEEE1394 Firewire (known as “i.Link” by Sony), USB and HDMI.
The normal assessment procedure for cameras could not be used, largely because the V1 does not have a
selectable test signal. Therefore, testing had to be done the hard way, via the lens. Recommended settings
allowing for a “video-look” have been derived; it was not possible to derive decent “film-look” settings due
to shortcomings in the controls of the camera, significant compromises have been made in the camera
design. However, if a “film-look” is really needed, then the camera can be switched into 25P mode
(CAMERA SET menu>Prog.Scan) and this will give correct jerky motion and a little more vertical
resolution than is achieved in interlace, but will not extend the exposure range.
While HDV performance is acceptable in that it is representative of it’s market niche, there are significant
problems with its performance as an SD camera for professional or broadcast purposes due to the presence of
visible aliases that cause a visible “restlessness” in the picture. HD performance with the recommended
settings is probably adequate for consumer use, but better SD performance can be expected byt recording in
HDV and using a professional hardware down-converter or by using the downconversion facilities available
in most good video editing software, or with a genuine SD camera. The reasons for this statement are given
in the measurements section of this document.
The controls for these cameras are not as flexible as for full “broadcast” cameras, so more effort was
expended in measuring performance than in trying to derive a specific “look” for it. Very small lens
apertures (less than F/5.6) soften the picture and produce some visible colour-fringing due to diffraction
effects in the iris, the included neutral density filters are the better alternative to small apertures when
shooting in very bright light.
Many of the menu items have little or no effect on image quality. Those that have significant effect are
highlighted. The full set of menu items is given for completeness. In boxes with a range of numeric settings,
the values indicate the range (often no scales are given) and the factory default setting is underlined. My
recommendations are in the last column, labelled “BBC”, where appropriate. Settings are given only for
normal television use, it was not possible to derive good settings for a film look.
In the tables, items that have an important effect on picture appearance are highlighted with grey
background. Rather than just making assertions about performance, I have included measurement results
that illustrate the reasons for recommending settings. Virtually all picture control is in the Profile menus.
Press the “Menu” button to enter the menus, navigate with the “Menu wheel”, default values are underlined.
Select profiles with the “Profile” button.
This is not intended as a replacement for reading the manual.

2
1 Switches and Menu settings

SWITCHES, CONNECTORS and BUTTONS

name place feature comment
Right, on tape
Power switch Right thumb control
transport
Right, on tape
Zoom 1 Pressure sensitive Normal zoom control
transport
Zoom 2 Handle
Zoom ring Lens Manual control
Photo/Expand Focus By Zoom 1 Push Expands image in viewfinder
Exposure Lens left Push Auto control
Exposure dial Lens left Wheel Manual control
ND filter Lens left Switch, 1/2/off 2 neutral filters
Auto lock Back Switch Locks exposure
Shutter speed Back Push
Gain Back Push
Focus ring Lens Manual control
Auto focus Lens left Push One-push auto ,focus
Focus Left mid Push/push Auto focus on/off
Expanded focus Left mid Push/push Expands image in viewfinder
White balance Back Push
Record start/stop 1 Power switch Right thumb control
Record start/stop 2 Handle
Assign 1,2,3 Lens top Push 3 assignable buttons
Assign 4,5,6 Under lcd Push 3 more assignable buttons
Zebra Under lcd Switch Zebras on/off
Menu Back Push
Select from list, after enabling by pressing
Menu Back Wheel “Push” button
Sel/Exec Back Push Confirm menu operation
Picture profile Left back Push Select preset “profile” settings
Status check Left back Push Show camera status, sound levels
Tape play controls Under lcd All normal controls
Display batt/info Under lcd Push
Volume/memory Under lcd Push Control of stills replay
LANC By Zoom 1 Remote control
AV connections Back Component, AV, 1394
HDMI out/USB Below lcd
Headphones Left back 3.5mm jack Mutes speaker
Rec ch select Sound pod Channel 1/Channels 1&2
Audio in (XLR x 2) Sound pod
Ch 1 auto/man Sound pod
Ch 1 +48v Sound pod
Ch 2 auto/man Sound pod
Ch 2 +48v Sound pod
Ch 1 audio level Sound pod dial
Ch 2 audio level Sound pod dial
Reset Under lcd Recessed Push Full system reset

3
CAMERA SET menu Basic camera settings
item sub range comments BBC
Exposure 1, Exposure 2,
Exposure/Iris Dial Assign What effect the Exposure dial has
Iris, AE Shift
Dial Sens High, Middle, Low Dial sensitivity
Dial Rotate Normal, Opposite
Smth Slw Rec Rec Time 3,6,12 seconds Smooth Slow Recording, at low resolution
Rec Timing Start, End Record from or to Rec button press
Execute Arms camera to do it, press Menu to cancel
Black lift for deep shadows, not when “Back
Cntrst Enhcr On, Off
Light” is on
Off {v},
Prog. Scan Off, 25P Progressive, scan, film mode1
25P {f}
Steadyshot On, Off Reduces camera shake
Hard, Standard, Soft, Wide Wide Conv lessens the effect, good when using
Type
conv. a WA lens adaptor
Color Bar On, Off
Type Type 1, Type 2, Type 3 Type 1=SMPTE, Type 3 is full height bars Type 1
Tone On, Off Add 1kHz at –18dB
AF Assist On, Off If ON, allows brief manual focussing in Auto
Focus Macro On, Off Allow close focussing
AE Shift -7~0~+7
AE Response Fast, Middle, Slow
AGC Limit Off, 12, 6, 0dB Off=allow 18dB AGC range
AT Iris Lmt F11, F5.6, F4 Stop-down limit2 F5.6
Intelligent, High, Middle,
ATW Sens Set degree of tolerance for auto white balance
Low
Flickr Reduce On, Off Useful under fluorescent lighting
Handle Zoom H 1~6~8 Set zoom speed for Zoom 1 control switch
L 1~3~8
Shot Trans Trans Time 3.5~4~15 sec Automatic shot transition settings
Linear, Soft Stop, Soft
Trans Curve
Trans
Start Timer Off, 5, 10,20 sec Delay to transition start
Rec Link Off, Shot A, Shot B Choose which transition to use
Interval Rec Record bursts of frames, see manual, P73
DV Frame Rec Off, On Similar in SD, see manual P74
Back Light Off, On Raise auto-exposure when scene is back-lit
Spot Light Off, On Reduce auto-exposure when spot lit
Hyper Gain Off, On 36dB gain, not in Spot or Back light
D.Extender Off, On 1.5 times magnification
Fader Off, White, Black Fade to/from white or black

AUDIO SET menu

item sub range comments BBC
DV Au. Mode FS32k, FS48k 32k is 4-channel, always 48k for HDV FS48k
Mic NR Off, On Mic noise reduction
XLR Set Au Man Gain Separate, Linked Links gain controls together, for stereo
Input 1 Level Mic, Line
Input 1 Trim 0, -8, 016dB
Input 1 Wind Off, On Selective wind noise reduction
Input 2 Level Mic, Line
Input 2 Trim 0, -8, 016dB
Input 2 Wind Off, On Selective wind noise reduction
Audio Ch Sel Ch1&2, Ch1, Ch2 Select which recorded track(s) to play
DV Audio Mix Ch1&2, Mix, Ch3&4 In FS32, select which tracks to play

________________________________________________________________________________________________
1
Progressive mode is good, it produces the correct jerky film motion, and increases vertical resolution.
2
Generally, performance of the lens is good, but some colour fringing is visible at F/11 due to diffraction from the edge
of the iris. Setting to F/5.6 is a good compromise, but can cause some exposure problems under high light levels (use
neutrals and/or shutter to control exposure).
4
DISPLAY SET. menu Viewfinders
item sub range comments BBC
Peaking Off, On Edge enhancement, focus aid
Color White, Red, Yellow Colour of sharp edges
Level High, Middle, Low
Histogram Off, On Exposure histogram
Marker Off, On
Center Off, On Cross hairs
Aspect Off, 4:3, 13:9, 14:9, 15:9 14:9 is useful, works only in “All Scan”
Safety Zone Off, 80%, 90%
Guideframe Off, On
Exp. Focus Type Type 1, Type 2 1=enlarge image, 2=enlarge and monochrome
All Scan Mode Off, On Shows whole camera frame, shrunk
Cam Data Dsp Off, On Shows exposure, gain, shutter
Au.Lvl Disp Off, On Audio level meters
Zoom Display Bar, Number Number only shows 0~99, not focal length
Focus Display Meter, Feet Focus distance
Shutter Disp Second, Degree 360°=field or frame duration
LCD Bright Press Sel/Push Exec dial
LCD Color ditto
LCD Bl Level Normal, Bright
VF B.Light Normal, Bright
VF Color On, Off Off=monochrome
VF Powermode Auto, On Auto switches vf off when lcd’s open
Data Code Off, Date, Camera Data Data to show on replay
Letter Size Normal, 2x Big letters for those with rotten eyesight
Remaining Auto, On Auto shows tape left on Play or Power On
Disp Output Lcd panel, V.Out/Panel Control data display

IN/OUT REC menu VTR matters

item sub Range comments BBC
Rec Format HDV1080i, DV DV is SD DVCAM HDV1080i3
Auto senses the tape, otherwise plays only
VCR HDV/DV Auto, HDV 1080i, DV
what’s selected from tape
DV Rec Mode DVCAM, DV SP DVCAM uses 50% more tape/minute
DV Wide Rec On, Off (4:3)
Off, Synchronous, Allows simultaneous recording on a 1394
Ext Rec Ctrl Rec Ctl Mode Relay connected external recorder
Stby
Rec Pause, Stop Control of external recorder, see manual P83
Command
Component 576i, 1080i/576i 576i for ordinary tv sets
i.Link Conv Off, On Auto up/down-conversion on replay
Squeeze, Letter Box,
Down Convert For playing HDV via analogue outputs
Edge Crop

TC/UB SET menu Timecode etc

item sub range comments BBC
TC Preset Preset See manual P85
Reset Sets TC back to zero
UB Preset Preset See manual P85
Reset Sets User Bits back to zero
TC Run Rec Run, Free Run Rec run times only when tape’s rolling
TC Make Regenerate, Preset Makes Rec Run when over-recording
TC Link Sync TC on cameras via 1394, see manual P 85
TC/UB Disp TC, U-Bit Which to show on-screen
UB Time Rec Off, On On=record real; time in User Bits

________________________________________________________________________________________________
3
Recording at DV is not very good, the camera generates significant visible spatial aliasing. Better to record in HDV
and down-convert in your favourite software or hardware converter.
5
MEMORY SET menu Memory stick matters
item sub range comments BBC
Record image sizes are1440x810, 1080x810,
Quality Fine, Standard
640x480,640x360
All Erase All Files, Currnt Folder Deletes files
Format Yes Select to format a card
File No. Series, Reset Reset restarts at 0001 when a new card’s used
New Folder Make a new folder in the 102-999 series
Rec Folder Sel/Push Exec dial, select which to record into
PB Folder Same for playback

OTHERS menu Basic stuff

item sub range comments BBC
Save/Load/Rename/Delete camera profile, see
Camera Prof.
manual P89
Last Scn Review,
Marker, Hyper Gain,
D.Extender, Allscan
Mode, Focus Infinity,
Rec Review, End
Assign any of these to each of the 6 assignable
Assign Button Search, Index Mark,
buttons. See manual P49~53 for details
Peaking, Steadyshot,
Color Bar, Focus
Macro, Spotlight,
Backlight, Fader,
Profile, Shot Trans
Photo/Exp. Focus Photo/Exp. Focus Assign the Photo/Exp.Focus button
Clock Set Set time/date, see manual P28
World Time Set offset to local time
Language Set on-screen language
Memory Stick, Pict
USB Select Print direct to an enabled printer via USB
Bridge
PB Zoom Allows zooming of play-back stills
Shorten rec start time, breaks MPEG2 GoP
Quick Rec Off, On Off
structure, may not be ingestable in some NLEs
Date Rec Off, On Superimpose date on stills
Beep On, Off Warbles when recording starts Off
Rec Lamp Off, On Cue lamp
Remote Ctrl Off, On Allows remote control
Shows values for: Operation, Drum Run, Tape
Hours Meter
Run, Threading

6
PICTURE PROFILES menus, default settings Camera control, default settings
item sub range comments BBC
PP1 Portrait These are the default settings, not individually
PP2 Cinema investigated. Each can be separately edited, and
PP3 Sunset saved to Memory stick (up to 20 can be stored
PP4 Monotone on one stick).
PP5
PP6
PP7
PP8
PP9

PICTURE PROFILES menus, manual settings Camera control

item sub range comments BBC
Color Level -7~0~+7 Saturation, -8=monochrome -3
Color Phase -7~0~+7 Greenish to reddish 0
Sharpness 0~7~15 Detail enhancement 3
Off, Type1, Type2, Anti-wrinkle cream: 1=narrow, 2=wider, 3-very
Skintone Dtl Off
Type3 wide
Skintone Lvl 1~3~6 Boost/restrain skintone detail
WB Shift -7~0~+7 - is darker, + lighter 0
Auto, High, Middle, Maximum point is fixed, adjusts intercept with
Knee Point Middle
Low main curve
Blk Compnstn Off, Stretch, Compress Slope near black, need stretch for good colours Stretch
Cinematone 4
Off, Type1, Type2 Film-type gamma curves Off
Gamma
Cinematone Color On, Off Film-type gamma curve Off
Profile Name Set the name, see manual P44
Copy Copy these settings to another profile
Reset Resets profile to default settings

________________________________________________________________________________________________
4
The film gamma curves reduce the camera photographic speed by lowering the slope of the curve in mid-tones. This
raises saturation a great deal and there was no sufficient control to get colours looking reasonable. Pictures look most
like film with the normal gamma, middle knee, and black stretch on. Even so, colouring is a little strange and not
particularly nice, it is a great shame that there is no colour matrix to play with, because considerable improvements are
to be had.
7
2 Measurements
Gamma curves and exposure range were not explored in great detail, since there seemed to be little control
available over either, this is a “take it or leave it” camera, with little real control over the image quality.

2.1 Colour performance

Assessments were done visually, since there was not sufficient time
available with the camera to do a full set of analytical tests. The camera
was pointed at a Macbeth chart, and the picture viewed on a large Sony
crt HDTV monitor with another Macbeth chart close to it. Both charts
were evenly illuminated with studio colour temperature for the camera,
D65 for the monitor reference chart.
With the camera set to factory defaults
(Fig. 1), the colours were rather Figure 1 default settings
muddy, over-contrasty, oversaturated,
and tinged with red near black level. With no colour matrix and no
black level controls, there was little that could be done about this.
Setting Color Level (saturation) to –3, and Blk Compnstn to Stretch (Fig.
2), colouring became a little better, but even so some of the colours
Figure 2 saturation=-3, black
stayed distinctly of wrong hue and
stretch=on saturation (e.g. the Yellow patch)
although skin tones were generally
good.
The gamma curve does not appear to be “powerful” enough, resulting in
the visible muddy colouring. Black Stretch helped a little, but both the
Cinegamma curves made things worse. The oddly named Cinematone
Color also made things worse, by increasing both saturation and contrast Figure 3 sat=-3, bs=0,
(Fig. 3). cinegamma=2, cinecolor=on

2.2 Sharpness and resolution

The camera has 3 1”/4 cmos sensors of 1440 by 810 pixels, presumably with green offset from red and blue
to enhance horizontal resolution beyond the limit of the sensors, this is normal practice in 3-sensor cameras.
The horizontal count in this camera is just adequate for HDV (1440 limit), so it should be adequate for
resolution, but the 810-line count is rather low for any attempt at a decent “film look”, where resolution up to
1080 is to be expected. It also works in two modes (HD video, SD video) and the requirements for each are
quite different.

The test card was a BBC Zone Plate, designed for

1080-line television. This reproduction of it (Fig.4)
shows the layout, but also shows considerable aliasing
caused by the scaled reproduction here. Each circular
zone is a phase-space of spatial frequencies, with zero
(dc) in the middle, extending to 1080 lines/picture
height (l/ph) vertically, and 1920 lines/picture width
(l/pw) horizontally. The scales are linear, so it is
relatively easy to make reasonable measurements. In
the camera, the image is recorded as 1440 pixels by
1080 lines, i.e. a pixel-based aspect ratio of 4:3, but this
illustration is shown with the correct image aspect ratio. Figure 4 zone plate

2.2.1 Resolution in HDV

8
Fig.5 shows the chart captured with Sharpness set at zero (i.e. this is
the native performance of the camera) and the camera set to normal,
interlaced, HDV. This time, I have not altered the scale of the HDV
file, but have included it here with square pixels rather than the 4:3
pixels that are assumed for HDV.

Along the vertical axis, resolution is visible up to about 75% of the

scale (i.e. 1080*0.75=810, which is the resolution limit of the
camera’s 810-line sensors). The resolution near this limit is easier to
see when the camera Prog.Scan mode is set to 25P (progressive)
since it causes interlace twitter when set to interlace. Thus the
camera appears to be sharper vertically in progressive than in
interlace. However, close examination of screen grabs shows no
difference in the frame content between interlace and progressive.
This is a reasonable compromise in a small camera where the cost of
proper filtering would be prohibitive.

Horizontally, there is a null zone (where the image is plane grey, no Figure 5 HDV, sharpness=0
detail) at 1440. This the horizontal resolution limit of the sensors,
however, the green sensor is presumably half-pixel offset from the
red and green to raise the effective resolution above 1440, which could explain the lower contrast resolution
above 1440. In practice, this results in aliasing in the image (where unwanted detail is “reflected” into the
wanted detail), causing problems when the picture moves, so it is rather dangerous to rely on this process for
producing high frequency detail.

The camera has only one control for detail (Sharpness), thus there is
little that can be done to optimise the performance, and it has a large
effect with coarse steps. Setting to maximum produce remarkably
high aliasing and overshooting on edges, a very unpleasant picture.
Fig.6 shows the performance when set to 3, which is about the
maximum value that can be used without causing unduly disturbing
effects.

The effect on vertical resolution is small, this is inevitable since the

camera cannot produce resolution above 810, and perhaps the
control has no effect on vertical detail anyway. But horizontally it
has increased the impression of sharpness by accentuating the
aliasing due to detail between 1440 and 1920. This is a mixed
blessing, the picture appears to be sharper, but much of this extra
“detail” derives from the aliased higher frequencies and not from the
wanted detail. When the picture stands still, this is not an issue, but
the aliased frequencies move counter to camera motion, so are
accentuated when the camera moves. This is why it is not advisable
to use higher values of Sharpness. Figure 6 HDV sharpness=3

There was no evidence of loss of sharpness when zoomed in tightly,

and iris diffraction did not appear to be a problem at normal exposure levels. However, it was not possible to
test at substantial over-exposure levels with the lens stopped down; it is advisable not to use smaller iris sizes
than about F/5.6 under high-contrast conditions.

2.2.2 Resolution in SD

Recordings were made of the same zone plate chart, with the camera set to SD (625/50, “PAL”). For SD, it
was framed such that exactly half the width and height were recorded, thus the circular zones excited
frequencies up to 960 pixels/picture width and 540 lines/picture height, adequate for these tests.

9
Fig.7 shows the result for the same zone quadrant, with the camera
set to record interlaced images.

Vertically, there is detail up to 360 line/picture height. A broadcast

SD camera can be expected to perform somewhat better, perhaps up
to 430. There is no evidence of vertical aliasing due to the presence
of the higher frequencies in the zone plate chart. Subjectively, there
is some interlace twitter, a restlessness in the image.

Horizontally, there is a null zone at 720 where wanted frequencies

and unwanted alias frequencies beat with each other, and there is
evidence of detail between 720 and 960. Since the DV format cannot
support frequencies above 720 (exactly half the sampling frequency),
this can only be aliased frequencies. This is evidence that the down-
converter does not use
sufficiently good filters,
which would normally reject
these high frequencies in the
camera’s HD image. The
visible result is a distinct Figure 7 zone plate, 50i SD
“business” in the signal at
high frequencies, a
disturbing restlessness (similar to the vertical interlace twitter) that
cannot be eliminated with camera settings (other than a diffusing
filter on the lens).

Fig.8 shows the result for shooting in 25P. Again, as for HDV
recording, the vertical resolution is the same as for interlace, but
there is a little more aliasing visible. This indicates that the down-
converter takes no account of the camera setting, and deals with
each HDTV field separately to make one SDTV field. Viewing this
on the same HDTV monitor (which automatically switched to
SDTV) there was still significant interlace twitter, because crt-based
television monitors rarely show truly progressive images except
when supplied with 720p signals.
Figure 8 zone plate, 25p SD
Horizontally, the result is exactly the same as for interlace.

Overall, down-converted pictures do not look particularly good, neither when the camera is set to HDV and
playback is SD, nor when the camera is set to SD. The cause is the presence of high frequencies in the HD
image, brought about by the HD lens and sensors. If the camera is to be used for SD work, then the only
solution is to use an optical diffusing filter on the lens, possibly a ¼ Black Promist or equivalent, to lower the
high frequency content before it reaches the sensors, where the aliases are generated.

2.3 Noise
Noise was measured by exposing the camera to a uniform grey scene, at 4 exposure levels from 15% to 95%
video level, with +12dB gain. Allowing for this gain value, the Peak Signal to Noise Ratio (PSNR) at 0dB
gain should vary between 54 and 56dB depending on signal level, which is very good for a camera in this
price and performance range. Note that the 8-bit DV and HDV recording compressor cannot accurately
represent noise at higher than 56dB PSNR anyway, so the camera is a good match to its recording formats.

10