Marti Maria

Thanks @kloczek for sending me the report. Most of those warnings are in the testbed code and in the jpegicc sample, so not part of the library. Anyway I will...

> I noticed that your code does not seem to do any sort of black point compensation for relative colorimetric intent. Lcms black point compensation works fine, but It is...

Those profiles are wrong. I have not analyzed the profiles in depth but at first glance profile iz.icc does not roundtrip on relative colorimetric on low values. Also profile id.icc...

I suspect the issue comes from id.icc, which returns a non-zero black point but below perceptual black. V4 ICC profiles are tricky, when using perceptual PCS they should conform to...

It happens that the profile has only one intent and this intent is perceptual no matter what header says. Try this when creating the id.icc profile and before saving it...

Michael, Thanks for reporting. I have uploaded a fixed version to the GIT server, hope that will address the problem. If I understood correctly, the compiler does not align strings...

Thanks for the fix. I'm investigating how this would affect buggy profiles, which in v2.0 are legion.

A quite complicated code, I see. May I ask for a small snippet that demonstrates the issue? I mean, a number across a parametric curve type 4 that results in...

Not sure about which code are debugging. Anyway if you compare 3D LUT interpolation versus a math expression, the results hardly depends on how you build the LUT. Keep in...

Not further comments so I close the issue