[argyllcms] Re: lightness adaptation algorithms

Roberto Michelena wrote:

So what happens? My source has, for example, min L*=5 and max L*=95 ;
while my proofer (Epson 4000 w/ glossy) has min L*=7 and max L*=94

If I use ICC profiles, a colorimetric intent (either relative or
absolute) results in clipping of highlight detail; it fades to white.
Also, clipping of shadow detail.
A perceptual intent, on the other hand, is not a proof; all colors
vary by a lot.


You seem to be after a mixed intent. For real proofs, absolute colorimetric
is the go, but of course it won't work well if the proofing device has
a smaller gamut than the device you are trying to emulate. One way of
dealing with this would be to make two prints, one using absolute, and
the other using perceptual. You'd use the absolute print for checking
all elements that are in gamut, and the perceptual one to check (as best
you can) out of gamut elements, and overall "look".

Question is, what's the best curve/formula/algorithm do to it? because
linear scaling doesn't sound as the best. Maybe scaling with some
perceptually linear gamma (L* is not perceptually linear, right?); or
maybe in shadows we need to scale a* and b* too, to preserve the
original chroma slope on the gamut edge. Maybe a scaling function that
gets applied only on the ends but merges smoothly into the rest of the
range. Or whatever. Suggestions?


It's really in the realms of research. Most current research in this area
tends to come to the conclusion that each image is optimal with a different
curve. This isn't much help for general purpose, automatic systems.
See <http://www.cis.rit.edu/people/faculty/fairchild/PDFs/PAP07.pdf> for
an example of this type of thing.

Using Argyll, you could try "Luminance matched Appearance" intent, which
may give you something close to what you are after. If you are not using
device links, then in creating the profile you would want to supply the
actual source profile of the device you are emulating, together with
overriding the perceptual table gamut mapping intent to be "Luminance
matched Appearance" intent. You would then apply the profile using its
perceptual intent.

Graeme Gill.


Other related posts: