Hi,I've been trying out the new release of Argyll and stumbled upon a problem - not sure if this is a bug, an intended change in behaviour or an unforseen side-effect of some of the CIECAM02 changes - but the viewing conditions flags for colprof and collink seem to behave differently from how they used to in 0.7b8.
When I generate profiles for my inkjet printer on glossy paper, I find I need to use the -cmd -dpp flags to lighten the output of the perceptual table. Without this adjustment, skin-tones in particular come out a little dark and muddy. The -cmd flag now seems to have far less effect than it used to.
To demonstrate, here's a softproof of an RGBXPLORER8 chart, transformed through the perceptual table, and back to sRGB via relative-colorimetric intent.
The four images are proofs using profiles generated using -cmt -dpp and -cmd -dpp, using 0.7b8 and 1.0.0 (patched) respectively.
http://mypicturepalace.com/albums/userpics/10004/7b8_mt_proof.jpg http://mypicturepalace.com/albums/userpics/10004/7b8_md_proof.jpg http://mypicturepalace.com/albums/userpics/10004/mt_proof.jpg http://mypicturepalace.com/albums/userpics/10004/md_proof.jpgNote that the two "mt" images are as near identical as makes no difference, while the two "md" images are quite different.
Is this change intentional? And if so, can I get something approaching the old behaviour with custom parameters to the -c and -d flags?
PS. an autotools version of the new release will follow shortly, as will an updated version of imgtarget, reflecting the new name for collink and new syntax for tiffgamut.
All the best, -- Alastair M. Robinson