i hacked-in an attempt to spectral conversion and print of negative scans
the idea is to use what we know of a film stock (spectroscopic and densitometry data) to have good estimates of the cmy film densities from scene referred scans.
main challenges are:
- fluctuation of the film densities and base due to processing issues or variance in general
- scanning conditions and illuminant
- metameric issues related to scanner ssfs and the specifics of the dye absorptions of the film
the way i plan to tackles some of the problems are:
- process the raw with reference D65
- start by selecting a negative scan illuminant that best neutralize the base
- fit the film base correction by morphing the original data-sheet base with CMY controls (an automatic fitting can be used if unexposed areas are available)
- use a 3x3 correction matrix that tries to solve the metameric issues due to the scanner not being a standard observer (this is fitted from a colorful image if available). the calibration should be reusable for the same scanner/camera and stock
here are a couple of quick and dirty tests from play raws:
play raw link
play raw link
not completely satisfied with them
more tests will follow
(it is fair to say that the play raws only prompted us with very “difficult” negatives, it would be nice to try with easy negatives)
if someone is willing to share some nice and well exposed portrait scans it would be great 
the main theme of the play raws is “here is my messed up negative, can you find a way to fix it”, but to test things we need well behaved negatives first