[Chuck Anstey]

There is something I am not understanding about RGB balance to set the grayscale and how it is related to gamut errors.


I have a pretty flat RGB gray scale from 20% to 100% on my G70 CRT projector using a combination of projector Bias, Gain and a Lumagen HDP 11 pt gamma curve. However when I measure the primaries in the gamut, it says the blue primary is way too bright, with hue and color having small errors. I don't understand how this is possible and still have a balanced D65 white point. When I change the Gain control, I am changing only the intensity (lightness) of the color at the upper end, not the actual color as that is fixed by the phosphor and lens, right? If I lower the gain control so the blue primary's dL is low, the RGB graph says I have very little blue. So if the white point is balanced in the correct proportion of R, G, and B, which should only be intensity, how is any primary too bright or too dim?


What am I not understanding?

 

[sotti]

RGB balance is different than gamut luminance.


That's why CMS systems have seperate luminance controls on top of the RGB LUT.


All of this is derived from the fact that displays are never perfectly additive.

 

[samhain1]

I have the opposite problem, near perfect d65 and blue luminance is way below the zero mark.


Any corrections via CMS reduces overall brightness by too much

 

[Chuck Anstey]



Quote:
Originally Posted by sotti /forum/post/20891989


RGB balance is different than gamut luminance.

But in what way? If I have my 75% white perfectly balanced between R, G, and B, I am assuming that means I have the correct luminance balance between them as that is all I change to bring them into balance. So at 75% white I turn off the R and G gun to display 75% blue, which measured equivalent to putting up a 75% blue window (i.e. only the blue gun is outputting any light and at 75%), how can the luminance be anything but correct for that primary? Too much blue luminance implies too high a color temperature.

 

[sotti]



Quote:
Originally Posted by Chuck Anstey /forum/post/20893837


But in what way? If I have my 75% white perfectly balanced between R, G, and B, I am assuming that means I have the correct luminance balance between them as that is all I change to bring them into balance. So at 75% white I turn off the R and G gun to display 75% blue, which measured equivalent to putting up a 75% blue window (i.e. only the blue gun is outputting any light and at 75%), how can the luminance be anything but correct for that primary? Too much blue luminance implies too high a color temperature.

There are many different issues in the signal processing and the way a pixel emits light when a single subpixel or all three subpixel are stimulated.


Your assumption is correct for a purely additive mixing model. The issue is that displays rarely function as purely additive mixing devices.