Homebrew standard video test charts?

I'm a bit out of it these days, but never saw such a chart used on studio cameras. Only 'chart' they used colour wise was a painted head for colour matching reasons. Surrounded by a grey card and grey scale. I'd say it extremely difficult to print anything accurate enough.

--
*Sherlock Holmes never said "Elementary, my dear Watson" *

    Dave Plowman        dave@davenoise.co.uk           London SW
                  To e-mail, change noise into sound.

I might first suggest looking at the test patterns for digital camers off the camera test sites. I have used those befors for different things.

For example. I look at my cameras image here. Printing is necessary.

Might also try sci.engr.television.broadcast

Well I could have selected a better camera.........

I've often wondered how one can absolutely determine if a video camera is properly color calibrated.

You don't explain what you want to acomplish.. certifying that a camera is operating within the manufacturer's original specifications, or finding the best balance of colors that a used camera is capable of.

I've seen the technical literature procedures and color/light boxes, but any piece of test equipment or color analyzer will have some variations. Even the charts from different sources will have inconsistent variations. If the charts are colored plastic sheets, exposing them to fluorescent lighting might fade them a bit. The output of the lamps in the light boxes will most likely vary by brand, or even differences in voltage.

As far as attempting to fabricate a light box for colors, I would assume that a well-stocked art supply company would have plastic sheets in a wide range of colors, but other than an art supply, the only other source I can think of might be a specialized stage lighting supplies business. Either of these might have products that have some standardized specifications associated with their products.. #112 red related to a specific color, as a guess (kinda similar to the way camera filter numbers are standardized).

A piece of equipment that's nearly always present with commercial video production is the waveform monitor, and often paired with a vector monitor.

Aside from the source colors being nearly perfect, monitoring displays will have differences, certainly between CRT and LCD or other display technology.

So then an analyzer is needed to determine if the monitor is delivering an accurate representation of the source signals.

The source product my be nearly perfect compared to some EIA or other standard, but then in commercial TV broadcasting, the video progrmming has to pass thru other levels of signal processing and transmission equipment, it's all quite extrordinary.

Finally, the consumer adjusts their monitor levels to their own preferences (or they don't have any) and the video may look like a 4-year old colored the displayed image with only 4 or 5 crayons.

The studio and field use ENG broadcast cameras are originally set up by the manufacturer, and may go thru numerous alignments in the useful service life, which I suspect are sometimes TLAR that looks about right, or close enuff.

Between the cameras and the transmitters there are likely to be a considerable number of pieces of equipment dedicated as auto-level compensators for various signal parameters. At one time, the controls for these adjustments were probably the responsibility of a broadcast station engineer, for the engineer to monitor and adjust as required.

It's very interesting to play with a color corrector to adjust the levels from different sources to get them to look nearly identical. For a small-scale production setup, I would think that a color corrector would be an absolute necessity.

In the movie industry, many varied shots are made with vastly different equipment, sometimes by choice and other times by necessity. I watched a director explain that anything from 8mm film, various types of video, and high-end wide film are sometimes combined to create a movie. In another example Spielberg(?) was explaining that the color levels were intentionally reduced in Saving Private Ryan, and when the movie made it's way to cable, the program providers were swamped with phone call complaints that the colors were too weak, so the providers adjusted the levels. OK, TLAR.

I suspect that much of the final product is corrected by specialized computers in most movie and TV productions. Viewers might expect a different level of color rendering for a product such as a nature film where everything should look natural.

-- Cheers, WB .............

Maybe the C1 chart of Coloraid could be the thing for you?

It's for photography, but I guess you could create a colourprofile for your videocamera with it as well. I got a chart of them to calibrate my scanner once in a while, after reading positive comments about the accuracy of the charts.

cheers

-martin-

I found the gray scale reflectance chart at our public access tv operation and checked it out for a few weeks to permit attempts to duplicate it. Indeed it uses 'chips' (swatches) of continuous tone pigment pasted onto a uniform background. The center square is still a bit of a mystery; on this chart, it is a piece of black velvet fabric. For the archives I will quote the instructions as pasted on the rear of the chart (typed with 80 char max line length):

--------------------------------------------------------------------- INSTRUCTIONS FOR GRAY SCALE REFLECTANCE CHART LOGARITHMIC REFLECTANCE CHART

General Description. The chart consists of a pair of gray scales running in opposite directions on a uniform 12% reflectance gray background. The scales begin with Black (3% reflectance), increasing logarithmically to White (60% reflectance). Table 1 shows the Nominal (ideal) Reflectance of the Gray Scales on the accompanying chart.

Table 1: Nominal Data for Logarithmic Reflectance Chart Gray Scale Chip No. 1 2 3 4 5 6 7 8 9 Nominal Reflectance 60.0 41.3 28.4 19.5 13.4 9.2 6.3 4.4 3.0 of Chip (% relative to MgO)

Use. Electrical measurement and adjustment of the camera chain equipment should be made first. Reflection measurements and adjustments should be made only after proper scanning, electrical and optical focus, coarse shading, and registration have been accomplished. The chart should be televised by the camera and reproduced on suitable picture and waveform monitors. It should be uniformly lighted in accordance with usual studio and mobile practices.

Checking the Transfer Characteristics. For a monochrome camera the camera output without gamma correction, when viewed on a waveform monitor at line rate, should appear as crossed stairsteps. This presentation is a logarithmic measure of the transfer characteristic of the camera and associated amplifiers. The crossed stairsteps should have the same total height and no step should exhibit appreciable tilt. Any abnormal tilt or difference in total height indicates incorrect shading adjustment. If gamma correction is used to give an overall system gamma of about unity, the steps will appear almost linear on the waveform monitor. With gamma correction, the output, when viewed on a picture monitor should appear as a faithful reproduction of the brightness steps of the chart on a flat field. Any severe departure from faithful reproduction of the original gray scale indicates that the system transfer characteristic is non-linear. Any abnormal departure from a gray field indicates incorrect shading.

For a Color Camera. A monochrome reflectance chart of the correct color temperature should appear on the color picture monitor. The picture should appear as a faithful reproduction of the brightness steps of the chart on a flat gray field. Any variation in chromaticity along the gray scale indicates that the transfer characteristic of the three color channels do not match. Any variation in color of the uniform gray background of the chart indicates that the shading characteristics of the three channels are not matched.

When viewed on a waveform monitor at line rate, the composite color video signal should appear as crossed stairsteps. The crossed stairsteps should have the same total height, and no steps should exhibit appreciable tilt. No trace of the subcarrier should appear on any of the steps or on the gray- surround line.

Rating and Matching Pick-Up Tubes. The techniques of determining the relative sensitivities, the matching of multi-tube groups, and the general adjustment of cameras depends upon the type of pick-up tube. The user should refer to the manufacturer's manual for these details.

Manufactured and Distributed by Hale Color Consultants, Inc. 1505 Phoenix Road Phoenix, MD 21130 (301) 472-4850

Television Test Charts should be handled carefully and protected from undue exposure to strong light, which can fade the colors and reduce their usefulness. Charts have an average wear-out life of two years, after first use. This chart should be replaced not later than --Feb 1988--

----------------------------------------------------------------------------

Now, it's off to the paint store...

Michael

Thanks for both of those; I have been collecting test pattern files for awhile and appreciate your link too.

Michael

Thanks for your suggestion; I do have an IT 8.7 (Agfa) target and will take that with me to the paint store to use for comparisons. Perhaps scanning it on a flatbed scanner at the store, noting the color data as the scanner saw it, and then scanning paint store 'chips' and noting their values would make for a quick and dirty color comparitor.

Michael

Actually it is for setup and calibration of the homebrew studio camera I described in a previous thread...

As I have learned by the caveats printed on the back of the reflectance chart which I borrowed today.

I will need to find a certain blue filter for one set of adjustments and other than a camera shop, the art supply store sounds promising.

Would be nice to find one of those suction-cup photometers someday for calibrating monitors.

Oddly enough, the engineer at our public access tv operation said essentially the same thing regarding their new cameras; they don't do any setup or adjustments other than the TLAR of flesh tones on talent in the studio. They also route all of the new cameras through frame syncs and proc amps so that is where all of the adjustments are done (is that where it is all headed - no more camera setup?)

See comment above regarding the no-setup approach.

Michael

Yes. The black velvet fabric provides a completely NON-reflective spot so that you can use a waveform monitor to see where the black level is coming out of your camera. VERY handy.

You have a chart that is 21 years out of date. If you match paint colors to it, you will have a copy of a 21-year expired chip chart. Do they not have a current chart?

It is not clear to me how aiming a camera at a graphic image of color bars can be used for calibration of any kind? How were you going to use this?

Colorbars are typically generated by a binary counter with the least significant bit ("1") driving the blue channel, the next bit ("2") driving the red channel, and the most significant bit ("4") driving the green channel. Known amplitudes (75%,

100%, etc.) are fed into the color encoding circuit to allow the encoder to be calibrated.

The imaging devices, lens, and lighting have no role in this part of a camera calibration. Trying to calibrate the encoder with color bars coming through the lens and onto the imaging devices is like trying to solve an equation with two (or more) unknowns.

The grayscale ("chip") chart is used for fine adjustment of the analog parts of the imaging devices before the signal goes into the color encoder. The color encoder must have already been calibrated (with the electronically generated color bars) or else the next step is futile.

After the encoder is calibrated, one adjusts gain, offset, gamma, knee, etc, etc. (whatever is available in the camera electronics) so that red, green, and blue all track perfectly in order to make a perfectly colorless (zero chroma) stairstep on a waveform monitor.

This was the Google's first hit searching for: TV chip chart

This guy has both grayscale ("chip") and colorbar images for what its worth.

If you are adjusting the colorimetry of the R, G, and B channels, then the exact value of the gray chips doesn't really matter. In fact, using a continuous ramp would serve the same purpose (i.e. getting the RGB channels to track perfectly.) The exact value (reflectance %) of the gray chips DOES matter when making absolute calibration for gamma. It also assumes that you have a waveform monitor to view the video signal with.

What are you calibrating? Do you have a calibration procedure which calls for a graphic image of color bars? Seems rather unconventional (not to mention difficult) to me.

A whole lot of really good stuff which I don't need to repeat. In my former life, I was a studio tech and spent (way too) many hours doing alignment on 3-tube cameras. All I ever used was a properly lit 10-step black & white chip chart in combination with waveform & vector monitors and an oscilloscope.

I sure hope that the OP has no intentions of trying to properly match the luminance values of a chart that's more than 20 years old as it ain't gonna happen.

msg wrote: > Oddly enough, the engineer at our public access tv operation said > essentially the same thing regarding their new cameras; they don't do > any setup or adjustments other than the TLAR of flesh tones on talent > in the studio. They also route all of the new cameras through frame syncs > and proc amps so that is where all of the adjustments are done (is that > where it is all headed - no more camera setup?)

Sad but true. The days of studio techs (especially in low budget places) actually giving a damn about technical quality is rapidly going the way of the dodo bird. My experiences lately have been that most of them don't know what to do with the bars & tone I put on the head end of a tape :-(

Mike

now there i was thinking bars and tone was slang for a pub crawl and the associated noise, and at the head was where it ended up ;-)

Watch it there 'kid' or the price for that Windows 7 package of yours will mysteriously double :-)

Mike

Calibration procedure calls for the following in this order:

Adjustment Chart

1. AGC Grayscale
2. Pedestal Level Grayscale
3. Output Level Grayscale
4. Carrier Balance & R-Y, B-Y Pedestal Grayscale
5. White Balance Grayscale
6. Auto White Balance Grayscale
7. Auto White Balance 5500 deg. Grayscale+filter
8. Color phase + R-Y, B-Y gain Color Bars
9. Hi Inten Surpress/Backlight Comp Grayscale Plus a bunch of adjustments of digital effects section using fabricated Red/Blue, cross, box and other charts.

The electronics doesn't generate color bars and the phase and gain adjustments are done by shooting the chart.

Michael

Whilst this is true, there simply isn't the need to line up modern cameras in the same way as of yore. They do it themselves in software. There is, however, still the same need to match them by eye using a decent monitor. Which can be a problem since CRT monitors are near obsolete, and LCD are near useless for this.

Again these days the pictures and sound are simply 'digitized' straight into the editor. But that bars and tone can be useful if there prove to be subsequent problems - if only to cover your arse. I'm using Panasonic P2 these days which records to memory cards - and due to the cost of these get re-used the next day after digitizing the contents - so there is no way to refer back to the original line up, unlike tape where they will usually be kept at least until the edit is finished.

--
*Bigamy is having one wife too many - monogamy is the same

    Dave Plowman        dave@davenoise.co.uk           London SW
                  To e-mail, change noise into sound.

The only home-brew test charts I know to work, are focus charts and black level chart (a small square or rectangular hole cut into a large sheet of flat white card-stock, backed by a box lined with dull black felt or velvet, so any light entering the hole is unlikely to escape).

The best advice I can offer you, if you have more than one camera, is to choose the one that appears to produce the truest colors, and adjust the other(s) to match.

It won't be right, but it won't jar the viewer's sensibilities quite so much when you switch between cameras. In a pinch, it'll get you through a remote-van live-shot.

Sounds like a low-end one-chip consumer camera. In which case it likely isn't terribly critical.

(extra stuff snipped)

That might be true at your local public access station, but public access is not exactly noted for its technical quality, and I can assure you that in in the commercial broadcast field, cameras are still set up rigorously, and anyone that would rout everything through frame syncs as a matter of standard practice would need to find a field other than system design to work in. Cameras are referenced and timed as always.

The gray scale chart is used because it allows the VO to match all the color channels in a number of parameters. Scopes (waveform and vector) are used to take the color variations of monitors out of the equation in the initial setup. High-precision video monitors (ideally CRT but, sadly, LCDs are working their way into this, and so far they don't compare) are then used for final painting.

In many ways, nothing has changed in video control since color cameras became the norm.

mg