Olympus-OM
[Top] [All Lists]

Re: [OM] Colour improvement suggestion using Device Profiles-Reply -Repl

Subject: Re: [OM] Colour improvement suggestion using Device Profiles-Reply -Reply
From: Scott Nelson <SNelson@xxxxxxxx>
Date: Thu, 24 Feb 2000 15:49:07 -0500
Well, I must admit I knew I was flame baiting as your original query about why
scanners don?t all provide ICC profiles begged the question. The short answer is
ICC profiles are too much work and they don?t work. They?re too much work
because you would actually need a profile for each scanner and every emulsion it
scans, much as you need a different profile for every printer/ink/paper
combination. Kodak went down this road in the early 90?s with a unique profile
for every film scanned to a Picture CD. It didn?t work then and it doesn?t work
now. Remember EFICOLOR (aka ?Iffy Color?)? Probably not, because it didn?t work
either. 

ICC profiles don?t work for several reasons. Firstly, they assume consistent
output or for scanners input. Well, scanner CCD?s change with age and printers
like Heidelberg Press vary with humidity or just different pressman. Secondly,
they?re based on absolute metrics where human color perception is relative and
highly influenced by neighboring colors. Furthermore, the Munsell test shows
that each person perceives color differently over a bell curve. Like many senses
this also changes with age. Thirdly, it just isn?t possible to take one master
image and apply variations that will make it come out optimally to say an Iris
printer, a Chromira, a news press, and a high end 4 color press. If it were, I?d
be right there on the bandwagon with you. Finally, the whole concept that we
would want to precisely match color in a slide, monitor and printer is fatally
flawed. The contrast, gamut, and perceptual differences between a
transilluminated slide, a fluorescing monitor and a reflective print can never
be the same. I?m amazed you say you get color matching across these devices
during your workflow. For that matter, no one really wants a perfectly matched
image but rather a perceptually pleasing rendition. This is why I will continue
to do the drudgery of color correction which some call craftsmanship or even
art.

And I still don?t know of any offset press that will do a run without a contract
proof or give one wit how you made the proof. They are the ultimate color
matchers, but rightly refuse to match the unmatchable. They will tell you if
your separations add up to more ink than they can print, however, and for that
reason they, like me, do color correction by the numbers


>>> Jan Steinman <Jan@xxxxxxxxxxxxxx> 02/23/00 09:44am >>>
>From: Scott Nelson <SNelson@xxxxxxxx>
>
>In my experience, ICC profiles are only slightly better than the PS "Auto
>Levels" command.

My experience is quite the opposite. My entire work-flow uses 
profiles, and I can go from light-table, to scanner, to monitor, to 
comping (color laser printer) to final output (50" Hexachrome 
ink-jet), and everything matches. I just don't give it much thought. 
It just works.

Perhaps you have something set up wrong, Scott. It sounds like you 
may be using "absolute colormetric" matching instead of "perceptual." 
Perceptual matching will not cause color shifts the way that "auto 
levels" does. Also, if you think color matching is important, you 
should be using "auto contrast" instead of "auto levels," since it 
works on Photoshop's internal luminance channel rather than 
independently shifting the end points of the individual RGB channels.

>The only thing that has ever worked for me is by-the-numbers
>correction in CMYK mode and test prints form the actual output device.

If you do that, you're manually doing what an ICC profile does, but 
only for one device. You want to go change everything if you have to 
go to a different press or paper? Nah, didn't think so!

>This is
>the method that every 4 color offset press shop uses.

Now I don't believe you. I know of many printers that rely on 
profiles, and at least one that will not guaranty color matching 
unless your source material comes profiled.

So in the least, you should change "every" to "in my experience, 
every" and then I can't argue with your limited experience! :-) But 
certainly, my experience differs.

>ICC profiles are yet
>another attempt to beat human experience and judgement with software 
>automation.

That's not fair. You can adjust things to your heart's content in a 
profiled work-flow, and they'll come out the way you see them. 
Profiles in no way get in the way of tweaking things the way you want 
them.

One could just as easily say that styles for text formatting are also 
a way to "beat human experience." I don't want to have to go through 
a brochure and make sure every little nuance of text formatting is 
consistent, and I don't want to have to correct colors for every 
device I use.

That's what computers are for, handling the boring bits so we humans 
can do the creative stuff unencumbered! I don't especially think 
color-correcting for various devices is an especially creative and 
rewarding way to spend my time.

: Jan Steinman <mailto:Jan@xxxxxxxxxxxxxx>
: Bytesmiths <http://www.bytesmiths.com>

< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >

< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >


<Prev in Thread] Current Thread [Next in Thread>
  • Re: [OM] Colour improvement suggestion using Device Profiles-Reply -Reply, Scott Nelson <=
Sponsored by Tako
Impressum | Datenschutz