Xasan

Well done!

Same as I. Borg found it https://www.dpreview.com/forums/post/65475442 That would result in the same relatively large dE. A typo or I'm missing something here?

One would think that's what designers are doing, but they are not. Limits are stricter than this, due to shot noise and various artifacts. But let's forget it, and pretend that we can seriously ...

I wonder if they respect the limits to highlights recommended by camera designers and how do they account for shot noise in highlights when calculating saturation point. Sensor begins to saturate ...

How do they define saturation?

I don't see why a practitioner, unless he for some reason is interested in DR, would use SNR 0 as a threshold.

SNR is a rather direct measurement, and part of it is shot noise, right?

Calculating M0 as pinv(raw' * raw) * raw' * xyzRef; 7.3014e-01 2.8669e-01 7.4507e-02 2.2482e-01 9.5975e-01 -3.1573e-01 7.3872e-03 -2.3704e-01 1.1420e+00 avg. dE00 = 16.873 Calculating M0 as xyzRef' ...

Good light source overall, but not very suitable for calculating a generic color transform  by solving  ref = M * dev for M, right?

Right! Thank you, I missed that.

Wasn't my intention to be critical at all. Some more thoughts if I may (for you, or for somebody who wants to use your code): 1. maybe to exclude cyan patch from fitting. it doesn't fit sRGB 2. ...

X-Rite 'ColorChecker Camera Calibration Software'. Professional. Funny.

Hi, Jack, A few questions and comments on your code to estimate an FM matrix, https://www.strollswithmydog.com/wordpress/wp-content/uploads/estimateMatrix_.zip 1. With the normalization you have ra ...

Of course not. You need it too. While you are calling a simple line on C gibberish, how is your Japanese, perfect? End of communication.

What explains things is the whole image processing pipeline, exposure, and the kind of scene. Why not to take all of it into account is beyond me. Maybe it's some kind of a trend, maybe it's just ...

LOL Should you knew, your wish would be sincere.

Division by 8 is normalization. #include <stdio.h> int main() { printf("%d %d\n", (217>>1) + (217>>1), (217+217)>>1); return 0; } 216 217 You don't know what it is :-P

They are very relevant as they are a part of in-camera image processing pipeline. Do the same stacking not in camera but manually, using different pipelines and you will see. Now if you'll excuse ...

Given the dynamic scene combined with multi-exposure and in-camera clipping, normalization, and stacking using limited fixed point range for computations and output - the resulting artifacts are ...

@Jon Stern:
Any papers on RGBW CFAs with noise and color error analysis that you can recommend? Thank you.