[argyllcms] Re: gamut mapping

Hi :)

edmund ronald wrote:


I don' know if this is relevant here, but I'm highly sceptical of
using Argyll to achieve adaptive gamut compression. When I tried that,
prints got worse rather than better.
Was the printer profile you were using for this generated using Argyll or something else? Remember that there's no "right" way to do the Perceptual intent - it's part art, part science - so if your printer profile was generated by something other than Argyll, your adaptively compressed results will have discarded the profile's existing perceptual mapping and used Argyll's method. It may simply be that you subjectively prefer the perceptual mapping from your original profile to Argyll's approach.
The other thing that can have a huge bearing is the input profile. For instance, there are several different versions of sRGB floating around which give markedly different results. (Not that I imagine you're using sRGB!) So for a fair comparison you need to use exactly the same version as was used with colprof's -S parameter when building the profile in the first place - assuming it was built using Argyll.
All of that notwithstanding, I have to say I have seen worthwhile improvements from using tiffgamut/collink in this way. The differences are subtle, not night-and-day, but, especially on poorer media such as inkjet printable DVDs, the results have been worthwhile.
If you have a specific test case where the image quality deteriorates it'd be interesting to see the images, profiles and commands you're using. :) 

All the best,
--
Alastair M. Robinson

Other related posts: