Re: DTP41 White Backing
Re: DTP41 White Backing
- Subject: Re: DTP41 White Backing
- From: email@hidden (Bruce Fraser)
- Date: Thu, 28 Feb 2002 17:38:57 -0800
At 4:20 PM +0800 2/28/02, joseph wilhelm wrote:
I've found that it pretty much boils down to, if one backing
doesn't produce good results, try a different one. When you get
results that you can live with, KEEP A RECORD of which backing you
used.
This variable results thing can be really confusing for those of us
(at least me) who don't have experience with a variety of profiling
packages or who are not immersed in profile building. I have
certainly learned that ICC color management is capable of wonderful
results but it can be difficult to troubleshoot when things are less
than wonderful. The people I know seem to weigh the shortcomings
very heavily and when a variable screws things up they readily
trash the whole icc process. The marketing literature often sets
one up for disappointment.
That's more a political than a technological issue...
I would like to hear (very much) how one should go about testing the
behavior and performance of a profile. Are there ways of testing a
profile other than liking or disliking the results or does this
really require broad experience? Liking or disliking is fine but
knowing why is even better and having enough understanding to
prevent poor profile performance would be great.
You have to take considerable care before you make a profile. Is the
measurement data truly representative of the process? (is the process
controlled and optimized, and are you making the measurements
correctly?) This requires some analytical skills, but mostly boils
down to common sense. If the process is shifting, you're chasing a
moving target. If you're measuring translucent material, you need to
pay attention to the backing over which the measurements are made.
I know that one ultimately has to "see" the results (easy for
inkjet/expensive for offset) but what can be learned from a profile
before it is implemented? How do you test?
Carefully? One golden rule is to test one profile at a time. That
means that all your other profiles need to be already tried-and-true.
With press profiles, you can learn a fair bit by cross-rendering
press CMYK to a proofer. If the results from the proofer look
reasonable, it's likely that the results from the press will be
reasonable also. You need experience and understanding of the press
conditions too -- there's just no substitute for that -- which allows
you to look at the numbers and decide if they're reasonable. Is the
profile producing enough ink? Too much ink? You can make rough checks
of black generation by looking to see if it's putting black where you
don't want it -- fleshtones, for example. But ultimately you have to
test on press.
If the answer is trial and error then how can we convince print
shops that ICC color management is better than their closed loop
systems?
I don't think that's ever been the argument. The point is that for
most shops, it's no longer possible to close the loop. Their closed
loop workflows are untenable because they have to deal with
client-supplied files from who-knows-where instead of doing their own
scanning, thus the loop is broken. Some shops try to shoehorn all
outside material into their closed-loop workflow, and usually wind up
complaining that their clients send them crap. The successful ones
have learned to bring outside work into their closed loop (there are
several points at which this can be made to happen), and simply don't
talk about color management because it's their competitive advantage.
Color management is no different from the proprietary lookup tables
on a 1970's Crosfield scanner, EXCEPT that it isn't proprietary. It's
just another way to get from A to B. A large part of the problem in
the US is that printers treat standards as something to be exceeded
rather than met, an attitude which the Europeans sensibly enough
don't share.
So every time I write a book, I have to go and profile the press,
making sure that the run is actually representative of the way the
book will be printed, instead of separating to a press standard with
the reasonably secure knowledge that the press will run to that
standard. When I sense that the shop is feaful of, or hostile to,
color management, I simply deliver final CMYK with no profiles
embedded, and tell them to run the job. As a result, it goes
smoothly, whereas if I embedded profiles they'd find a way to screw
it up. This is not a technology problem...
Bruce
--
email@hidden
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.