Digital White RGB


Please excuse the newbie question.
I always thought that the "digital definition" of "Digital White" in an RGB system was R = G = B.
In the process of creating ICC profiles for my Nikon D810, I use the ColorChecker chart.

I took a shot in early afternoon direct sunlight, last week (January 18).

When I opened the NEF in RawDigger and looked at the Histogram, I was "shocked" to find out that all histograms were not "equal"?
Here is a link to a screen capture :!AkD78CVR1NBqlsIq1fSwoy2uHWVIyA?e=DciJRN

I supposed the "whitest" white object in the image, the Munsell N9.5 white patch, would show up on the extreme "right side" of the RGBG2 histograms at pretty much the same locations.

But that's not how it is. I previously measured this patch to be CIELab = 94.47 -0.93 2.35, slightly "yellowish" I have an old 20 years target). In fact, I get these RAW values for the white patch :

R: 3516
G: 6777
B: 5073
G2: 7006

I'm surprised these don't correspond to the *Max* of each "channels", as reported on the Histogram as :

R: 4845
G: 10279
B: 7456
G2: 7623

I am curious where these Max values come from.
Here is a screen capture of RawDigger main interface, showing my ColorChecker :!AkD78CVR1NBqlsIrTkWHwckq1730_w?e=zToRk7

As you can see, there does not "seem" to be any area with a "higher reflectance" than the N9.5 patch.

Dear Roger:

Dear Roger:

R=G=B corresponding to neutral is valid only for colour-neutral RGB colour spaces, raw normally is not colour-neutral (that's why white balance exists, )

Raw file may contain some hot pixels, first thing I'd do is selecting ColorChecker and checking maximum on just that selection. You can also switch on filtering of abnormal pixels in RawDigger (Preferences -> Data processing).

Without having the raw file it's hard to say more.

Add new comment