What if gamma can't reach the target? - AVS Forum
Display Calibration > What if gamma can't reach the target?
wchris's Avatar wchris 05:12 AM 12-19-2012
Let's assume that gamma target is set to 2.35 for night time viewing but the display can't go beyond 2.2. In that case color patches will be undesaturated. Is it a good idea to increase the "main" saturation knob to make them meet the targets or is it better to leave them in correct places for 2.2 gamma target?

Gregg Loewen's Avatar Gregg Loewen 06:40 AM 12-19-2012
you only consider color "patches" after the gamma and white balance is set.

Yes, you would adjust the "main saturation knob" to make the saturation correct for whatever gamma is actually set at and not the desired target.

fun stuff!!
Doug Blackburn's Avatar Doug Blackburn 01:48 PM 12-19-2012
Gamma is a function of luminance, not saturation. "Low" gamma (numerically) causes colors to have too much luminance at any given grayscale step. In your example, if the target is 2.35 (too dark IMO, it should be 2.25 because the displays used for editing/color grading run at 2.2 or 2.3 gamma) at 50% white, red, green, and blue would be lighter than they should be (assuming 50% white is d65 gray) but color saturation could be perfectly fine if the display is capable of achieving 100% saturation for red, green, and blue.
wchris's Avatar wchris 07:15 PM 12-19-2012
@Doug
That is only true for 100% saturated targets . Everything 'below' changes when gamma does, not only the luminance. That's because the triplet for i.e. 25% saturated patch is not constant. It is based on display gamma.

@Gregg
I don't clearly get your answear, so I made the following animated gif from calman charts. It exhibits how the targets change when different gamma target is set.


So now let's assume that I targeted 2.35 gamma but the display is capable of 2.2 only (according to the gamma chart from calman). The measured patches (25% saturation sweeps) lay within desired areas on chromacity diagram calculated for 2.2 gamma target. But not in areas from 2.35 diagram.

So now should I increase the saturation so that at least the saturation pathes agree with the 2.35 target, or change the target to 2.2 and just calibrate the display (white balance & gamut) to it??
turbe's Avatar turbe 07:20 PM 12-19-2012
why would you target 2.35 when you state your display can't be calibrated to 2.35?
Quote:
gamma target is set to 2.35 for night time viewing but the display can't go beyond 2.2.

So you correct sat for 2.35 but actual gamma is <=2.2?? doesn't make sense to me...

However, after your calibration session(s) you need to view with your reference content as well and make the proper compromises if necessary.. you cannot always rely on the numbers.

You may have to do exactly what gregg posted (and using actual gamma targets).
Doug Blackburn's Avatar Doug Blackburn 08:53 PM 12-20-2012
Quote:
Originally Posted by wchris View Post

@Doug
That is only true for 100% saturated targets . Everything 'below' changes when gamma does, not only the luminance. That's because the triplet for i.e. 25% saturated patch is not constant. It is based on display gamma.

Luminance controls gamma. Saturation does not. You can have 2.25 gamma with 100% saturated targets or 50% saturated, or 25% saturated. The TV may or may not track that correctly. Gamma is difficult to conceive-of in RGB mode, in uvL (hue, saturation, Luminance) space, Gamma is much easier to visualize and conceptualize. Things have to happen in RGB space in ways they do not happen in uvL space.

When you work with gamma on a step by step basis, and you have a graph that shows red, green, and blue bars... you are looking at Red luminance, green luminance, and blue luminance. You want them all equal to achieve any specific grayscale step. When you're at 50% white, you still want red, green, and blue luminance to "match" each other so the step is as close to neutral gray as you can get. This is difficult to conceptualize in RGB space, much easier when you think in terms of uvL space.
Up
Mobile  Desktop