[argyllcms] Re: icclink -G and source gamuts /-profiles
- From: Martin Ling <martin-argyll@xxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sun, 11 May 2008 13:27:36 +0100
On Sun, May 11, 2008 at 12:23:54PM +0100, Alastair M. Robinson wrote:
>
> Yes, that's right - so it seems both Argyll and LCMS have issues with a
> gamma 1.0 input profile. The behaviour of both system improves rapidly
> as the gamma increases, though - at a cursory glance, LCMS seems fine at
> 1.1, cctiff without the -p flag seems accurate enough at 1.3 or so (and
> with -p, of course, is fine even at 1.0)
In a spooky coincidence, at exactly the same time this thread has been
unfolding I find I am running into this issue myself while trying to
sort out gamma conversion properly in the ufraw code, using lcms.
It seems that LCMS actually manages better than Argyll here. I am
testing as follows, using input profiles produced from the per-camera
rgb->xyz matrices in dcraw and gamma curves of 1.0 and 2.2.
for g in 1.0 2.2; do
tifficc -i input-$g.icc -o sRGB.icm -t 0 -w raw-$g.tif lcms-$g.tif;
cctiff -ip input-$g.icc -ip sRGB.icm raw-$g.tif argyll-$g.tif;
done
The LCMS output images match, but the Argyll ones do not, unless the -p
option is added.
Unfortunately as soon as I start using the proofing functions of LCMS,
big differences start appearing in the proofs despite the outputs being
identical.
It seems the only way to get accurate proofs is to apply a gamma
function to your linear data before giving it to the CMS. I would really
like to avoid this, as if the final output profile is (near-)linear then
this operation will just get reversed, damaging information in the
process.
Is it possible to replicate the softproofing ability of tifficc using
Argyll? If I just add my display profile to the cctiff chain, I get:
"Possible photometrics for output colorspace Lab are:
1: CIELab
2: ICCLab
cctiff: Error - An output photometric must be selected with the -e parameter"
Martin
Other related posts:
