I apologize for the late response … it’s because I pass through the forums in a hurry lately … so I forgot this.
My opinion is that expanding the temp/tint range could be dangerous, I thing @jdc Jacques expanded it at the safety limits when we had an issue with underwater photos falling out of range. It would be useful to have Jacques’ input on this …
I any case, as I see it, the tint range expansion is still a workatound because some IR photos will need unknown range (maybe too large).
I think a better solution is to add an option in color management besides “no profile” … i.e. “unitary” or “infrared” checkbox where with this option RT will use the unitary matrix from the start (instead of the normal color correction matrix at Dcraw.cc or camconst.json)
The idea is same with what I used to calculate the above matrixes. Based on the assumption that with IR light all Bayer color filters have same response (i.e we have an almost BW shot) I used a (almost …) unitary RGB matrix and the WB multipliers and then converted to XYZ. I cannot remeber exactly the workflow now … I am away from my home PC, but tomorrow I 'll be back and be able upload an xls with the calculation.
The reason I asked for cc24 shots was to determine the WB multipliers on the grey patches, not to calculate color correction.
BTW is the optimum with IR to have a perfect gray rendering as a start or is it better to be able to distinguish colors according to Bayer CFA response (as we saw in the samples green renders blueish so I suppose the response is not exactly equal) ?.