Re: What from to What
Re: What from to What
- Subject: Re: What from to What
- From: Roger Breton <email@hidden>
- Date: Wed, 16 Jan 2002 08:42:34 -0500
>
I usually suggest rendering intents as follows: Use what looks best!
>
Thankfully with Photoshop 6 we have the ability to look at the effect of the
>
intents in either the Convert to Profile dialog or we can set up a few sets
>
of custom soft proofs for each intent (and I usually have a set with and
>
without black in/paper white).
>
>
If you have to GUESS (that is, apply an intent outside of Photoshop) then
>
I'd probably pick Relative Colorimetric as a default (hold a gun to my
>
head).
>
There are many factors contributing to this decision, indeed.
It could depend on the client application used to apply the profile.
Linocolor, for instance, is notorious for strictly using Perceptal rendering
intent to convert from RGB to CMYK -- there is no UI for selecting RI in
linocolor! Is that bad or good?
IMHO, It depends a lot on the profiling application used to generate the
profile and it's "secret sauce". PO4's Perceptual Rendering, for instance,
encapsulate years of repro experience. So, letting Linocolor convert to Lab
on open using PO4 Perceptual Tag is a good thing. Other profiling packages
perceptual rendering are far from shining, in comparison -- severely
compressing the source into the destination gamut and significantly
desaturating colors on the way in. So what to expect from these
application's profile percepual RI in ANY application : not much. At least
not for very saturated colors. Could still be excellent for skintones.
I agree that a properly calibrated and characterized monitor is a man's best
friend. And not all profiles are created equal.
3D graphical tools where one could hit a button and "see" the position of
Source colors and Destination colors, in Lab or Lch space, would certainly
help make the RI decision easier. But we're not likely to see these soon.
And there are lots of applications where folks are just looking for ONE
universal profile with ONE universal rendering intent. So. But we can be
hopeful that colorimetric research in graphic arts that is presently going
on behind the scene is going to help...in the long run. Hey! Do not forget
that the CIE 1931 XYZ was created in 1931! It took all this time for the
printing industry to catch up to the CIE in a relatively democratic and
accessible way.
It's a forest of factors.
--
Roger Breton
Quibec Institute of Graphic Communications
Montrial (Quibec) Canada T. (514) 389-5061 F. (514) 389-5840
email@hidden
Sorry for my typos. I try to correct myself but I know I always miss some.