under the Caveats Topic of the Color Calibration chapter, I followed the procedure to avoid invalid CCTs for certain camera brands.
As described, I created the preset with the picked coefficients. My problem now is, that if I set the CAT adaption method in the color calibration module to anything else but bypass to adjust color temperature, than the WB and color calibration module rightfully complain, that white balance is applied twice.
Is that a case of misconception, live with it as everything behaves as intended, a bug or a suggestion to have an option in white balance to let color calibration know that this is constellation is intended to subdue the warning? In case of the later two I would forward that to github as well.
Cautiously I think you can ignore the warning…Likely hard coded for someone not using camera reference or as shot. I guess I should read your topic. I thought it had to be camera reference or for cameras with unexpected D65 values then it is set to as shot. I think anything else will likely trigger the error…
You have followed the special case process to approximate a D65 WB which is what is needed so I think you are fine. I think you just have to ignore the warning as it would be flagged by any non expected WB setting… I believe this would be the case
Indeed - in most cases you need WB set to camera reference for color calibration to work so this warning is shown if WB is not set to this. If you know that camera reference is wrong for your camera you’ll have to set WB coefficients differently and you can ignore the warning.
Well, you have applied the white balance twice, so the warning is accurate. But it’s only a warning, meant to say that you are not in the usual use case and maybe it’s wrong. Except for your case, it’s right and you know why. So, simply disregard it.
Apart from the warning, do you get better result with the custom WB procedure?
I was thinking to do the same as I have a sony, which is said to have “strange” WB coefficients.
In my hack raw processor rawproc, you can have thirteen white balance tools in the toochain, and the only complaint you might get is the segfault for the thirteenth tool…
Actually, I’ll do two whitebalances sometimes. If I think WB needs additional work over the original as-shot, I’ll just add a second tool someplace after demosaic and tweak it there, as the original tool is usually part of a group and I’m too lazy to split it out…
Technically, the result should be the same. Only the CCT in K has a connection with real-life black body that it might not have had before. But it’s really just an interface.
In my unscientific dabble I can confirm:
same picture results in
D65 Camera Reference Preset WB → CAT CCT Daylight 4599 with an illuminant temp of 4597
Preset (5594k/0,962) → CCT Black Body 5016 with an illuminant temp of 5001
No visible change in the picture, only slightest movement in the histogram
I might need to add, that I did not encounter a problem picture so far. Studying the documentation I stumbled over that topic and simply tried it out.
So for day by day use, as long as I am not hitting a problem, I will stick with the general D65 setting
