Re: Display Tweaking
Re: Display Tweaking
- Subject: Re: Display Tweaking
- From: Henrik Holmegaard <email@hidden>
- Date: Tue, 6 Feb 2001 17:57:55 +0100
Intents are the hardest and most confusing part of ICC color management.
If the LAB-CMYK transform is edited (the output transform), then separation
results through the profile are modified, and the return display of CMYK to
the monitor corresponds to the change in the separations data. This is the
more common edit, as users attempt to tune the input to output match, or
impart a shop-preferred subjective "look" to their output reproductions.
I agree that what users want is as a rule to automate color
corrections. This they do by editing the Lab -> CMYK Perceptual table
(: the printing table) in the profile for the production color space
which is the color space of the final press. It is also possible to
edit the RGB -> Lab table in the scanner / camera profile to get
automatic color correction.
Users shouldn't edit the gamut mapping for the production process in
order to get a better proof. The point is that the proof is
colorimetric, so if the space I intend (no pun intended -:)) to use
for proofing doesn't have the right gamut volume, then my first
option isn't to edit the colorimetric table in the production profile
but to increase the gamut volume of my proofing space by changing
paper or changing inks or getting a better monitor. Because ICC
workflows allow any profiled PostScript printer and any profiled
monitor to be used for simulation, it doesn't follow that my devices
are suitable for the job.
Then again the FOGRA folks are saying about contract proofing that
you load the deviceCMYK UGRA/FOGRA Medienkeil into your color server
and tell the server to match the strip along with the deviceCMYK
images and vector graphics colorimetrically from production CMYK to
proof CMYK. If each color studio / shop delivering artwork for a
press run edits the edit the CMYK -> Lab part of the production
profile then I'm not sure this improves the tolerance range of the
proofs when they are forwarded to the press operator.In this scenario
I'd think the press operator would want to disregard the submitted
proofs and use an ICC color server to normalize the incoming data
simply to find out what will actually print on the press.
This doesn't imply that there are no reasons for modifying the back
transform to the monitor and proof printer. It's just that it's tough
to figure out what to edit and when. Editing intents leads to the
same device dependence as device link profiles incarnate. Two devices
may be tuned to each other, but then something else is going to give.
--
Henrik Holmegaard
TechWrite, Denmark