Getting cameras to match can be a never-ending task. Especially if the cameras aren't identical to start with. However, what if you discover that none of the cameras you're trying to align seem to respond properly to the same color chart? That question was recently submitted by a Broadcast Engineering reader:
Please settle a long-running debate regarding the Sony Color Pattern Chart used with PTB500. Red is 104 degrees. What is the correct phase angle of the color following red? If it is magenta, it doesn't seem to register with any three-CCD camera I have used. All three-CCD (sub-broadcast) cameras I have tested on this chart have given a different phase on this color - the best being 85 degrees, and the worst 110 degrees. I have bought two Sony color charts with same result. What is this color? What chart do you suggest I use with Sony PTB 500 for testing three-CCD color cameras?
I turned this one over to Sony's camera expert, Larry Thorpe, vice president of acquisition systems. Here is what Larry had to say:
"My understanding is that the chart is made by a Japanese third party, the same one that makes the other PTB charts. It is obviously made of strips of filter film, similar to Kodak Wratten filters. We do not know the details of spectral response, and therefore cannot begin to specify various camera responses. Further, we have no information on camera colorimetry among cameras in the business and industrial categories. These specifications are simply not published by any manufacturer. Such information is notoriously unavailable from those design groups.
In my opinion, the best use of that chart is strictly for side-by-side comparisons among the same model. Various models will have different responses due to IR filters, lenses, CCD chip generation and a variety of other details.
I recommend use of the MacBeth ColorChecker chart and lights for colorimetry tests due to the built-in grayscale, many more and more useful colors, less possibility of systematic errors in the PTB (filters, lamp, mirror and diffuser condition as well as adjustment and reference calibration instrument). Again, comparison testing is useful, but absolute colorimetry is a difficult and elusive goal beyond any real need in checking inexpensive cameras."
Dr. Digital responds: Matching the grayscale response of older tube cameras, even of the same make and model, was often a challenge. We used to avoid putting more than one color monitor in a production room because getting two color monitors to match perfectly was nearly impossible. Getting a perfect color match between different makes and models of today's digital cameras may be asking a bit too much. Jerry Cohen (then at JVC) once detailed the various compromises made when designing three-CCD color cameras (BE July 1993). Send me an email with your fax number and I will get a copy to you.
Getting lost in the numbers With all of today's digital equipment, it is sometimes easy to get lost in the numbers - audio sampling rates to 96kHz, hard drives capable of handling more than 100GB, HD data rates at 1.5Gb/s. It is easy to test equipment to death while failing to notice the simple things like the camera focus is wrong, or the mic is picking up fan noise from a nearby computer. Old cars can easily do the speed limit on today's highways, and old analog equipment can make high-quality pictures in today's digital world. In the end, it is the quality of sound and picture that make the difference, especially when combined with compelling content. How many times have we watched the footage of the Tacoma Narrows bridge, Zapruder's 8mm film of the Kennedy assassination or the archive footage of the final moments of the Hindenburg? Two of the three were black and white. The third is so grainy it is difficult to make out. But those images are forever etched in our minds because of the content, not the color or resolution.
As usual, your comments and suggestions are welcome. Drop me a note at drdigital@compuserve.com.
