Hi Matt,
I am working with Jeff Saarela in the application of the adv7181D. WE got the color space conversion to work when processing interlaced component RGB video (PAL RGB and RS170 RGB, each with SOG). So our programming of the CSC registers works great. Reds
look red, blues look blue, greens look green
But then when we try processing Graphics RGB (like XGA 1024x768 @ 60Hz progressive, SOG) we can see the video image content (shapes, letters, etc) just fine, but the color is "tinted heavily green" Where we expect black we see green. Where we expect blue we see green. See some red and dont see any blue (even though source video has blue in it). Tried setting AGC_TAR to all possible values (from min to max possible) and it did not have much of an effect. One of the extremes led to giving me a little more red than original settings, but the image overall was still heavily greenish. Tried setting manual gains on each of the channels A, B,C. Increasing gain on channel A led to even more green (and the red went away). I think channel A is representing the luminance Y. Tried increasing channel B gain, and that brought the red back and made the red a 'richer red'. I think channel B is representing the chromanance Cr. Tried increasing the channel C gain, and it had absolutely no effect. I think channel C is the chromanance Cb.
( Fyi: green is on ADC0, red is on ADC1, blue is on ADC2)
Do the above observations perhaps trigger an idea as to what might be wrong and what might fix it?
Thanks
Jeff and Spiros