Dan Margulis Applied Color Theory
Tagging Files for CTP
Date: Sat, 08 May 2004 07:41:19 -0000
From: "al_qhatani"
Subject: ICC for CTF & CTP
Hi everyone.
I am working as production coordinator for a magazine
publisher in Saudi Arabia. We do not print our magazines but outsource it
instead.
We supply them with a CD with our PageMaker files, all
fonts and hi-res pictures, tagged with our own icc SWOP 21% dotgain.
However, we were informed by our management that on the
next issue we will be switching printers through bidding, ergo we don't
know who's going to win the bid and that if they are using a CTF or a
CTP.
I don't know which direction to go. I hope you could
offer me your 2-cents.
1) will it be safe for us to save all our work in
Adobe RGB, and only convert to final CMYK when we are sure if it's going to
be separated using CTF or CTP.
2) what will happen if I send our job with a
different dotgain setting embedded in it. Does dotgain setting affect the
actual film or plate separation or just our softproof preview.
Many thanks.
Ahmed Saleh
Saudi Arabia
___________________________________________________________________________
Date: Sat, 08 May 2004 23:33:02 -0000
From: Stephen Marsh
Subject: Re: ICC for CTF & CTP
Ahmed writes:
I am working as production coordinator for a magazine
publisher
in Saudi Arabia. We do not print our magazines but
outsource it
instead.
This seems fairly common.
We supply them with a CD with our PageMaker files, all
fonts
and hi-res pictures, tagged with our own icc SWOP 21%
dotgain.
Perhaps less so...
QXP and perhaps AID are often more common, APM
perhaps less so for magazine layouts...
Have you looked into the supply of high res PDF if your
printers prepress section can deal with them hassle free?
There are many advantages, including file sizes for the
pages or and the file being more 'stable/static' than a layout file which
requires fonts and images to be correctly linked and the possibility for
error.
PDF may not always be hassle free, but when it works
well it works very well indeed and there are usually more benefits than
disadvantages.
Of course, this would be tested first and not put
straight into production.
However, we were informed by our management that on the
next
issue we will be switching printers through bidding,
ergo we
don't know who's going to win the bid and that if they
are using a
CTF or a CTP.
Yes, it is always 'fun' when critical aspects of
production suddenly change and results are doubtful or unknown.
I don't know which direction to go. I hope you could
offer me your
2-cents.
I would try not to make any hard decisions until you
know more info, unless production is going to be put out and you will not
be able to meet the press deadline etc.
1) will it be safe for us to save all our work in
Adobe RGB, and
only convert to final CMYK when we are sure if it's
going to be
separated using CTF or CTP.
Is this your normal practise of placing A98 RGB in APM,
or do you convert to CMYK earlier in the process and place CMYK?
If placing RGB, would you then ensure that you convert
every placed image to CMYK or whatever print mode was required, using
Photoshop - or would you attempt to use ICC CM in PageMaker for the placed
images and convert on the fly (not ideal, but it can be done).
What would happen if an A98 RGB file was imaged to
film/plate as if it was a CMYK image (incorrectly handled)...if you miss
some RGB images is this going to be an issue?
If you convert to a CMYK as per normal early in the
production process and then you find that the new condition is very
different with dot gain or stock or inks...is it possible to convert to the
new CMYK from the old CMYK or would you be better served by using the RGB
originals (if they still exist etc).
I think you need to consider these and many other
aspects of production, and how any changes made may affect things down the
line.
I have only been thinking of editorial images, what
about the all important
advertising which pays the bills? What if this is
screwed up? Often
ads are supplied final and are not edited unless one is
explicity told to.
Are the supplied ads going to be the right tone and
colour?
2) what will happen if I send our job with a
different dotgain
setting embedded in it. Does dotgain setting affect the
actual film
or plate separation or just our softproof preview.
When viewing a RGB file via a CMYK softproof the
dotgain is being applied on the fly.
When converting to the destination CMYK profile, the
dotgain is applied to the data - the separations contain altered tonal data
taking the gain into account.
When viewing a CMYK file, the dotgain in the tagged
profile or the assumed profile is used to display the tonality of the
numbers in the file for on-screen preview or for converting back to other
modes.
A CMYK file is device dependent, stock, inks and dot
gain are all
taken into account when the separation is made.
There are two issues in this last question of yours:
a) Tagging a CMYK profile
b) The CMYK values/numbers in the file
The values/numbers describe the colour/tone. If the
numbers are not right for the gain of the process, the file will be too
light or too dark.
In most CMYK PS workflows the CMYK values are simply
passed on unaltered to output with no colour management. The profile means
little once the image is in layout software for many users, and in some
cases may be worse than having no profile (this has been debated many times
in the past) - but the profile is good for Photoshop.
If ICC CM is being used, then the profile in the placed
file may become more of an issue - since the profile may be used for a
conversion at print time.
Stephen Marsh.
___________________________________________________________________________
Date: Mon, 10 May 2004 06:23:34 EDT
From: Dan Margulis
Subject: Re: ICC for CTF & CTP
Ahmed writes,
1) will it be safe for us to save all our work in
Adobe RGB, and
only convert to final CMYK when we are sure if it's
going to be
separated using CTF or CTP.
If this is going to an unknown printer with whom you
have had no previous contact, merely knowing whether they are CTP isn't
enough to alter anything. If they are both film and CTP in the same shop,
they may have modified their CTP curves to match the film. Or, there may be
other process control issues that outweigh the question of how the plates
are made. Unless you get more information before the job begins, you're
basically stuck with just sending them CMYK files as normal, but without an
embedded profile. Then, for the second issue they do, you'll be in a better
shape to react to their conditions.
2) what will happen if I send our job with a
different dotgain
setting embedded in it. Does dotgain setting affect the
actual film
or plate separation or just our softproof preview.
If the dot gain was changed before the file was
converted to CMYK, it affects the separation. If it was changed after the
CMYK conversion, it affects screen preview only. If you don't know
anything about the printer, you shouldn't embed any kind of profile.
Dan Margulis
___________________________________________________________________________
Date: Mon, 10 May 2004 08:21:53 -0400
From:John Castronovo
Subject: Re: ICC for CTF & CTP
We've been through this before, but I'll never
understand how it could hurt if it's the correct profile. It'll probably be
ignored by the printer anyway, but it might help someone too.
john c.
___________________________________________________________________________
Date: Mon, 10 May 2004 10:34:34 -0600
From: Chris Murphy
Subject: Re: ICC for CTF & CTP
On May 10, 2004, at 6:21 AM, jc castronovo wrote:
We've been through this before, but I'll never
understand how it could hurt
if it's the correct profile. It'll probably be ignored
by the printer
anyway, but it might help someone too.
Statistically there is a higher likelihood it will be
ignored. But the other two possibilities are:
1. automatic conversion (i.e. Convert to Working CMYK
type policy behavior. This is something ID 2 does by default and won't not
do so long as color management is turned on, unless you manually turn off
color management on a per image basis)
2. embedded profile preserved but conversion does not
occur
Clearly case #2 is benign. Case #1 could be benign or
it could be a disaster. Really, the recommendation to not embed a profile
in CMYK images is exclusively related to submitting CMYK to people you
haven't had the color management talk with.
In a sense it's like airport security. Treat everyone
as though they're guilty. (That airport security is rather insecure makes
it a not so good analogy but I'm ignoring that for now.) It's not
particularly rational but in most workflows it's better to live with
problems you know than to also have to live with problem you know and don't
know (those having to do with trouble shooting embedded profiles in the off
chance a bad setting will cause a problem).
What I generally recommend, is for people to save CMYK
without embedded profile, but archive the job/project with the profile used
for making the separations. That way it can be used as source for the
sophisticated user who comes across this project at some later date.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Mon, 10 May 2004 15:02:29 -0400
From: "Remaley, Dan"
Subject: RE: ICC for CTF & CTP
Nearly all scans (CMYK) are designed for SWOP.
Including sheetfed or web. - Density and dot gain specs. for (GRACoL #2
(155-175 line) are the same as SWOP (133 line) specs. Traditionally Web's
had more gain but by increasing the line screen in sheetfed, we create more
gain. Most printers are around these specs,. unless they are CTP linear and
print with a low 15% midtone gain. Free pdf's on process control and gray
balance at <www.gain.net> under process control.
Dan Remaley
Process Control Mgr.
Graphic Arts Technical Foundation
412.741.6860x450
___________________________________________________________________________
Date: Mon, 10 May 2004 23:38:09 -0400
From:John Castronovo
Subject: Re: ICC for CTF & CTP
From: "Chris Murphy"
Clearly case #2 is benign. Case #1 could be benign or
it could be a
disaster. Really, the recommendation to not embed a
profile in CMYK
images is exclusively related to submitting CMYK to
people you haven't
had the color management talk with.
If case 2 is benign, then ok. But if case 1 is to be a
disaster, it could only be because the wrong profile was embedded or
profile management hasn't been set up correctly on the rip. The first
instance is the responsibility of the person submitting the file and it's
unlikely that the wrong profile is attached at this point in time. The
chances are very good that it represents the way the guy submitting the
file converted it. If his needs are special and he doesn't use profiles,
fine. Don't use it and no harm done. In the second instance, the idiot
running the rip shouldn't have color management turned on to convert
incorrectly if it was never set up to work in the first place. If he's not
comfortable using it, DON'T - then there's no harm embedding it either if
it's ignored anyway.
I'll tell you this. If I embed the correct profile with
a CMYK file that I submit and it doesn't print correctly because someone
converted it incorrectly, there'll be a shoot out to the death. I'm getting
tired of this B.S. There's no reason for it. I respect other peoples'
profiles all day long and we've been doing it for years. If there's any
doubt, we call. I expect the same from other suppliers who consider
themselves professionals.
Dan, I don't want this to become another flame war, so
just delete this is it offends you, but geez folks, get with it.
john c.
___________________________________________________________________________
Date: Tue, 11 May 2004 12:37:45 -0000
From: Stephen Marsh
Subject: Re: ICC for CTF & CTP
John writes in reply to Chris.
If case 2 is benign, then ok. But if case 1 is to be a
disaster, it could
only be because the wrong profile was embedded or
profile management hasn't
been set up correctly on the rip.
John, when this issue comes up, the list is usually
referring to the supply of CMYK raster files which are used in layout
software and the layout software is not altering the final values of the
CMYK file (in most cases ICC CM may be off or set to not convert etc).
Layout apps when used as many do use them for the common CMYK PostScript
separated film/plate work-flow, do not use ICC CM to do the seps from
layout (seps are done in Photoshop and simply passed through the print
stream).
As discussed in the past, RGB and CMYK are often
handled very differently. What makes sense for one, does not always make
sense for the other.
Users such as myself prepare with care to the target
condition taking all the responsibility of file prep for this condition.
Sorry to shout, but -
WE DO NOT WANT OR NEED A HUMAN OR RIP ALTERING OUR
COLOUR NUMBERS FURTHER DOWN THE LINE (not including proofing, only output
of plate or film).
There is no debate on this point.
If the conditions change, then I change the files to
suit and resupply the data. If this is not possible/realistic, then I am
taken 'out of the loop' and someone else can take the heat or the praise
for the image and file prep etc. If this is happening, then I no longer
want to know or care.
The first instance is the responsibility
of the person submitting the file and it's unlikely
that the wrong profile
is attached at this point in time.
Correct, the person handing off the file is preparing
it for the known target condition then there _is_ no need for altering the
file - it is simply output from layout software with no alteration of the
values in the file.
It does not matter if the profile is correct or not -
that a CMYK image in this scenario has a profile can cause action to be
taken (alteration of colour values) when no profile may mean no action is
taken (no alteration of colour values).
The chances are very good that it
represents the way the guy submitting the file
converted it. If his needs
are special and he doesn't use profiles, fine. Don't
use it and no harm
done.
So now you have no problem with untagged CMYK?
CMYK use is not as easy to pigeon hole as RGB use.
In the second instance, the idiot running the rip
shouldn't have color
management turned on to convert incorrectly if it was
never set up to work
in the first place. If he's not comfortable using it,
DON'T - then there's
no harm embedding it either if it's ignored anyway.
A 'colorimetrically correct' conversion will alter the
colour numbers in the file, to in theory make the same visual response.
This is good on the computer...but a press is very different.
Imagine a file overlapping an element in layout
software, both have the same CMYK colour values for a seamless blend of
colour...but someone converts the image file but not the layout...the new
mix of numbers should equate to the same LAB colour as the old...but back
in the real world a printing press does not behave in a tame fashion and
the two colours will no longer be seamless.
This is only one example.
Another could be a CMYK file with a hand crafted K
plate for fine black
detail, which then get's butchered into CMY with little
black ink etc.
I'll tell you this. If I embed the correct profile with
a CMYK file that I
submit and it doesn't print correctly because someone
converted it
incorrectly, there'll be a shoot out to the death.
That is lots of fun, but I prefer to ensure that the
job works fine first go.
I'm getting tired of this B.S. There's no reason for
it. I respect other peoples' profiles all day
long and we've been doing it for years. If there's any
doubt, we call. I
expect the same from other suppliers who consider
themselves professionals.
Again, bad experience can cause some to take out extra
insurance.
If a profile is meaningless to a certain work-flow,
then I would say that perhaps it is B.S. to include them and mystery meat
is valid?
Yes, if we are talking of handing off images which are
not directly linked/used in a layout and the next user needs to edit/view
them etc and does want to do more than simply output the files
numbers...then yes, a separate profile or even one tagged into every file
may be OK.
It all depends.
Dan, I don't want this to become another flame war, so
just delete this is
it offends you, but geez folks, get with it.
We are 'with it' John - it's just that not everyone has
the same idea of what 'it' is and should be.
As Chris points out, a separate supply of the profile
which represents the aim-point is very workable and is better than tagging
1mb of data into every file, when the profile is redundant to most final
output for CMYK seps...and when a profile may invite 'action' of some sort
where as no profile would not. There is an archived thread about 30,000
images which all use the same profile - is there really a need to tag each
file when/if a work-flow can be used which assumes the supplied profile as
the 'tag'?
Just for the record, I usually do tag - but there are
times when I make a hard decision not to...this could be for internal or
external work-flow reasons...it all depends.
Stephen Marsh.
___________________________________________________________________________
Date: Tue, 11 May 2004 09:54:58 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: "Stephen Marsh"
As Chris points out, a separate supply of the profile
which
represents the aim-point is very workable and is better
than tagging
1mb of data into every file, when the profile is
redundant to most
final output for CMYK seps...and when a profile may
invite 'action'
of some sort where as no profile would not. There is an
archived
thread about 30,000 images which all use the same
profile - is there
really a need to tag each file when/if a work-flow can
be used which
assumes the supplied profile as the 'tag'?
Just for the record, I usually do tag - but there are
times when I
make a hard decision not to...this could be for
internal or external
work-flow reasons...it all depends.
This is all well and good, and I agree with both of you
on these points. What I'll never understand is how simply providing more
information about my file may invite disaster at the hands of some idiot -
and that idiot is defended by this group.
john c.
___________________________________________________________________________
Date: Tue, 11 May 2004 11:52:11 -0600
From: Chris Murphy
Subject: Re: ICC for CTF & CTP
On May 10, 2004, at 9:38 PM, jc castronovo wrote:
If case 2 is benign, then ok. But if case 1 is to be a
disaster, it could
only be because the wrong profile was embedded or
profile management hasn't
been set up correctly on the rip.
Automatic conversion can be a disaster very easily,
even if the source and destination profiles used are the exact right
profiles. Black only drop shadows become four color; black only text
becomes four color; two color text becomes 3 or 4 color text; yellow
anything will become contaminated and is often noticeable; any unique
channel effects applied in CMYK are largely lost when converted such as
specific channel sharpening, or more importantly black generation used
which is both image specific and output device specific.
Most of these problems go away if there is perfect
register and process control. So if the output device is a digital press or
ink jet printer, repurposing can actually be a GOOD thing. Four color black
text with perfect register is more dense than black only. Who wouldn't want
that?
There is more to color reproduction than colorimetry.
Colorimetry can be exactly correct but fail to reproduce the appropriate
results. A simple example of this is total ink limits. If I have 8 inks
available, I have 800% ink possible. But obviously I still have a
conventional ink limit because it's still an ink on paper process. A
profile built without an ink limit would produce up to 800% total area
coverage separations. It will soft and hard proof these separations as
though they will hold detail, but we know that won't happen.
The first instance is the responsibility
of the person submitting the file and it's unlikely
that the wrong profile
is attached at this point in time. The chances are very
good that it
represents the way the guy submitting the file
converted it. If his needs
are special and he doesn't use profiles, fine. Don't
use it and no harm
done. In the second instance, the idiot running the rip
shouldn't have color
management turned on to convert incorrectly if it was
never set up to work
in the first place. If he's not comfortable using it,
DON'T - then there's
no harm embedding it either if it's ignored anyway.
The problem is always a downstream one. And the
upstream person really has only a blunt instrument to prevent most
undesired conversions: not embedding profiles in CMYK images. I don't like
the blunt instrument, but I know that it works with hard headed workflows,
which happen to still be in the majority.
I'll tell you this. If I embed the correct profile with
a CMYK file that I
submit and it doesn't print correctly because someone
converted it
incorrectly, there'll be a shoot out to the death.
There is no doubt that you will be in the right, and
they will be in the wrong. But you are the pedestrian and the printer is a
Mack truck. As you walk into the crosswalk confident you have right of way,
are you really going to let that Mack truck clobber you? After all, you'll
be in the right and the Mack truck driver will be in the wrong. You'll be
dead, but maybe that's beside the point?
It's a case of choosing your battles, and by default
I'm unwilling to recommend other people have this kind of fight with live
jobs while not having the color management talk up front with the printer.
Debating the merits of color management upfront I think is perfectly
valid, and if you're in a position to have a cow in front of the printer,
or even go so far as to pull the job from him because he isn't adequately
color management savvy, I support that also. But to blindly submit a job
with a deadline that has a bunch of embedded profiles in it without having
a conversation with the printer, is just asking for trouble, even if it's
his screw up due to lack of color management expertise.
Ask for trouble, if you want, when you have a way out.
Don't ask for trouble when you don't have the time for it.
I'm getting tired of this
B.S. There's no reason for it. I respect other peoples'
profiles all day
long and we've been doing it for years. If there's any
doubt, we call. I
expect the same from other suppliers who consider
themselves professionals.
Not everyone is in a position to dictate the printer to
be used.
If you do the exact right thing even in InDesign today
it can goose your black generation. Let's take the case of three RGB images
and three CMYK images. Each CMYK image was separated with a different
profile, using different amounts of black generation. Profiles are
embedded. RGB images have a profile embedded as well.
Place into InDesign (or QuarkXPress for that matter).
You have the option to do nothing to the CMYK images in InDesign CS and
QuarkXPress. So their embedded profile is effectively ignored. But lets say
you don't do this, you're going to color manage the whole thing. What you
get is all three RGB images get converted to a single CMYK profile with a
single black generation; and all three CMYK images get reseparated with
that same CMYK profile. So now everything is using the same amount of black
generation. The sophistication of color management in a late or later
binding workflow is risky and what you're proposing is a *possibility* of
late/later binding.
Late binding workflows are flexible, and they can be
very efficient. But they are a lot more complicated. Not everyone wants to
take on that level of complication.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Tue, 11 May 2004 12:02:03 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 11, 2004, at 7:54 AM, jc castronovo wrote:
This is all well and good, and I agree with both of you
on these points.
What I'll never understand is how simply providing more
information about my
file may invite disaster at the hands of some idiot -
and that idiot is
defended by this group.
The idiot is not being defended. I'm simply saying he
exists. If you want to wrestle with an idiot, it's your choice. But I'm not
going to recommend people plan on doing this from the get go, and get
themselves into a potential legal battle to prove they are in the right and
the idiot is in the wrong. Of course they'd win, but what a friggin pain in
the ass.
It's a lot more efficient to negotiate this in advance
with a conversation. If the printer flips out over the idea of color
management, embedded profiles, or whatever, then the efficient, logical,
low stress options are simple: a.) don't use this printer; b.) find out
exactly where he's planted land mines and avoid them so you can get to the
other side.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 12 May 2004 00:53:31 EDT
From: Dan Margulis
Subject: Re: Re: ICC for CTF & CTP
John Castronovo writes,
I'm getting tired of this B.S. There's no reason for
it. I respect other
peoples' profiles all day long and we've been doing it
for years. If there's any
doubt, we call. I expect the same from other suppliers
who consider
themselves professionals. Dan, I don't want this to
become another flame war, so just
delete this is it offends you, but geez folks, get with
it.
This is like the boxer who continues to fight on after
his corner has thrown in the towel. Since the last time this tiresome topic
surfaced here, the subject has become moot. After six years of
resistance-is-futile rhetoric like the above, effective with Photoshop CS,
embedded CMYK profiles are ignored by default. This ends any
possibility of this becoming a standard workflow in anybody's lifetime, and
renders the "get with it" comment above approximately equivalent
to an assurance that Betamax is the format of the future and that we should
forget all this VHS and DVD nonsense.
What I'll never understand is how simply providing more
information about my
file may invite disaster at the hands of some idiot -
and that idiot is
defended by this group.
You have expressed such befuddlement many times in the
past, and in response, several group members have given you example after
example after example after example after example of how embedded CMYK
profiles have caused problems down the line. Adobe's action in pulling the
plug speaks for itself, but I can add yet one more example: at the close of
the most recent thread, both you and another advocate of this workflow
indicated that in a certain job (admittedly not a particularly simple one)
you would have used a correctly embedded CMYK profile to convert, in a
situation where in fact this would have ruined the job. This error in
understanding the purpose of the profile is discussed in
http:
//www.ledet.com/margulis/ACT-postings/DailyLife/ACT-Deliver-in-LAB.htm
(see my message of 7 August 2003 explaining why using
the embedded profile in conversion would have been an error.)
Granted that your company is considerably more
sophisticated about these matters than the average printer, and still is
capable of destroying a job by misunderstanding the purpose of an embedded
CMYK profile, the idea of sending such a document to an unknown printer
makes sense only to those who enjoy playing the blame game.
Dan Margulis
___________________________________________________________________________
Date: Tue, 11 May 2004 22:51:36 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: "Chris Murphy"
The idiot is not being defended. I'm simply saying he
exists. If you
want to wrestle with an idiot, it's your choice. But
I'm not going to
recommend people plan on doing this from the get go,
and get themselves
into a potential legal battle to prove they are in the
right and the
idiot is in the wrong. Of course they'd win, but what a
friggin pain in
the ass.
And an unnecessary pain in the ass it is. How can you
be so sure that the same idiot isn't going to one day assign the missing
profile to your image and then convert it to what he feels he should be
using? The real problem is people doing things to your files without
asking, and not whether there's a profile attached to it. If you don't want
it used, tell them to ignore it (they probably will anyway from what you're
saying) or don't embed one at all if that's what the job requires. I
couldn't care much either way. On the other hand, if I put one there,
shouldn't it be easy to prevent it from causing trouble? Is it that hard?
john c.
___________________________________________________________________________
Date: Wed, 12 May 2004 09:50:05 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 11, 2004, at 10:53 PM, Dan Margulis wrote:
This is like the boxer who continues to fight on after
his corner has thrown
in the towel. Since the last time this tiresome topic
surfaced here, the subject
has become moot. After six years of
resistance-is-futile rhetoric like the
above, effective with Photoshop CS, embedded CMYK
profiles are ignored by
default.
Yeah.
Photoshop of all versions have ignored tags in CMYK
images by default. And in Photoshop 6, 7, and 8 when the policy is set to
Off, not only do you get blatant purging of this metadata, but by default
Photoshop doesn't embed a new profile into the image.
Even if the next versions of Adobe applications handle
this better it will take many years before the remnants of Photoshop CS 1
and earlier are effectively irrelevant in this context.
PDF, OS X, InDesign, color management, and other new
technologies fracture the market. They increase the disparity between those
who sharpen the saw on a regular basis, and those who don't have the time
or interest in staying on the bleeding edge, and counting their fingers
every other day.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 12 May 2004 09:40:59 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 11, 2004, at 8:51 PM, jc castronovo wrote:
And an unnecessary pain in the ass it is. How can you
be so sure that the
same idiot isn't going to one day assign the missing
profile to your image
and then convert it to what he feels he should be
using?
I'm reasonably confident in the law of averages that
someone who has already tried doing this has been ejected out of the
nearest airlock already. And it's also unlikely that my project is the
first one he's going to try this on.
I worry about this sort of thing happening like I worry
about a shotgun being pointed at me the instant I walk into a print shop.
It's really quite unlikely. Doing the wrong thing has a funny way of
exposing itself in pretty obvious ways.
The real problem is
people doing things to your files without asking, and
not whether there's a
profile attached to it.
There's no question. But desktop application software
has made it too easy for the unknowledgable individual to plug in settings
that are not at all appropriate for their workflow. The CMYK working space
setting is a very straightforward thing. The Convert to Working CMYK policy
is not. Yet both are equally accessible to users.
If you don't want it used, tell them to ignore it
(they probably will anyway from what you're saying) or
don't embed one at
all if that's what the job requires. I couldn't care
much either way. On the
other hand, if I put one there, shouldn't it be easy to
prevent it from
causing trouble? Is it that hard?
Absolutely they *should* ignore it instead of us
needing to strip valuable metadata. But I'm not going to have this debate
AFTER the fact in the unlikely event the job gets screwed up merely because
there are embedded profiles. I'd rather find out where the bodies are
buried before I get started so my job isn't the one getting hit on the back
of the head with a shovel. I prefer to be pro-active.
Besides, there is functionally little difference for me
to submit my images as untagged CMYK, and then include the CMYK profile
along with the images. Anyone sophisticated enough will know what to do.
Anyone not sophisticated enough will ignore it, including all of their
improperly configured applications.
This is effectively how PDF/X-1a works. All objects are
untagged. They are /DeviceCMYK. But they are implicitly tagged with a
non-binding profile referred to as the OutputIntent. The OutputIntent is
used as the source profile for proofing. By default it even overrides the
working space settings in Acrobat 6. But when it comes time to print, by
default it prints /DeviceCMYK unless you direct your application to produce
a proof.
I am a proponent of metadata. And embedded ICC profiles
are metadata. Stripping them or not including them is a form of data loss.
It's not OK and we should all be working to isolate the problems associated
with using such metadata and work on solving this problems. But hell if I'm
going to stand on a hill and die when the easiest short term solution when
working with unsophisticated printers is to just not give them an excuse to
hose my project.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 12 May 2004 06:26:39 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: Dan Margulis
You have expressed such befuddlement many times in the
past, and in response,
several group members have given you example after
example after example
after example after example of how embedded CMYK
profiles have caused problems
down the line. Adobe's action in pulling the plug
speaks for itself, but I can
add yet one more example: at the close of the most
recent thread, both you and
another advocate of this workflow indicated that in a
certain job (admittedly
not a particularly simple one) you would have used a
correctly embedded CMYK
profile to convert, in a situation where in fact this
would have ruined the job.
First, Adobe didn't pull any plug, they just changed
the default setting in CS to 'not embed'. One can still change that.
Second, you've got it backwards on the strange embedded CMYK profile.
My July 31, 1993 post was as follows:
--
We just got a CMYK file from France that came over the
T1, and thank GOD it had an embedded profile. It was unique, and we never
would've guessed that the photo was the color it was by assigning anything
out of PhotoShop! We checked a few just to make sure that nothing looked
better. Any flavor of SWOP was a mile off. There was no proof and even our
client didn't know what it was supposed to look like. Using their profile,
as it turned out, it was right on. How else can you do that?
--
Of all the reports of embedded profiles causing trouble
that I've been shown, the only complaint that still holds water is the fact
that it increases file size by about 500k. Some fool may see a profile, and
do something stupid because of it, but we stand just as much chance of the
fool doing something to the same file without the profile.
Granted that your company is considerably more
sophisticated about these
matters than the average printer, and still is capable
of destroying a job by
misunderstanding the purpose of an embedded CMYK
profile, the idea of sending such
a document to an unknown printer makes sense only to
those who enjoy playing
the blame game.
I don't think that we ever destroyed a job because of a
profile - a missing one maybe. We can see when the wrong profile is there
and then we change it, call or ignore it, but this is rare now since most
people aren't using PhotoShop 5 any more.
I'm trying to stop the blame game. I never want to hear
that my job was destroyed BECAUSE I used profiles. (BTW, this hasn't
happened yet.) That's no excuse and it's not that hard to turn off
auto-convert. If you have special needs that can be better addressed by not
embedding, I have no problem except that it makes my job harder to guess if
I get that file some time in the future, but there shouldn't be a problem
if I do embed profiles for more routine work. I don't believe in setting
traps for printers to fall into, and I don't play Russian roulette with my
work. Profiles are there to help, and used properly, they do.
john c.
___________________________________________________________________________
Date: Wed, 12 May 2004 14:33:35 EDT
From: Dan Margulis
Subject: Re: Re: ICC for CTF & CTP
John Castronovo writes,
First, Adobe didn't pull any plug, they just changed
the default setting in
CS to 'not embed'. One can still change that.
No. They changed the default to *ignore completely*.
Six years ago, they introduced the concept of the embedded profile as
"the universal language of color", soon to be mandatory in every
file, to which printers and service bureaus had six months to a year to
adapt or die, and without which quality reproduction is impossible, and to
which resistance is futile. Now, no matter how carefully you make and embed
your profile, you can rest assured that the overwhelming majority of
Photoshop CS users will never even know that it's there, let alone make use
of it. Thus, if you want somebody to assume that the embedded profile is
correct, you have no choice but to call that person up and tell him what to
do (just as you would have had to do pre-1998), since the norm is now to
ignore it completely.
Think 8-track tape. Think 9600-baud modems. Think
44-megabyte Syquests. The embedded CMYK profile as a guarantor of
color integrity is deader than any of them.
Second, you've got it backwards on the strange embedded
CMYK profile.
No, you've picked up the wrong quote. I was referring
to the case where I would submit a CMYK file to a printer with instructions
to print it exactly as is, notwithstanding that I did not know what the
CMYK conditions were. The print house hypothetically needs to convert to
RGB in order to pull a proof. Both you and the other advocate of this
workflow indicated that you wanted me to embed a profile so that it could
be used to make an accurate conversion to RGB so that the RGB output would
match what I saw on my screen. But, as the objective is not to match my
screen but rather the final CMYK output, this would have destroyed the job.
The proper decision would have been to treat the file as untagged.
It is precisely to avoid such disastrously incorrect
decisions that sensible people have stayed away from embedding CMYK
profiles unless they're confident that they will be used correctly down the
line. The Photoshop CS decision simply accepts reality. Now if we only get
a few more choices in the Color Settings dialog box, people might actually
be able to configure Photoshop so that it could give meaningful alerts
about incoming tagged files, but that's another story.
Dan Margulis
___________________________________________________________________________
Date: Wed, 12 May 2004 16:11:40 -0400
From: "Lathan, Kim"
Subject: Re: ICC for CTF & CTP
To John C......
So was color management, using ICC profiles, etc easy
for you to understand when you first started? Maybe you are a really smart
guy and picked it up right away or maybe you struggled with it like a lot
of people have or still do.
You should remember that many printers AND designers,
AND photographers, especially a lot of small to medium ones (and even some
big ones) have only recently become true converts/users of color
management as we know it today. And it has only been recently that anyone
has put together a nice book (Real World Color Management) that brought
enough information together in one place for a lot of people to finally
begin to grasp some of its intricacies and interrelationships.
I have been using color management and creating/using
profiles for years now, but I do still remember what it was like to 'wrap
my head' around the concepts/theory. Yes, I do find that embedded profiles
can cause some issues, it depends on the device they are going to and the
applications being used. I also know that sometimes the profiles are there
for a reason and should be treated as such. But in the not so distant past,
Photoshop on a normal install would embed profiles by default if someone
did not realize what was happening. Profiles were being embedded by people
with no particular knowledge of its use or application in a lot of cases.
So I find it very unfair that you call people idiots.
You can do/say what you want, but that is an awful way to portray a group
of people who may or may not have knowledge of a piece of technology that
you find so easy to understand. Are there people using it incorrectly or do
not understand? YES! But education and working together on this is
the key to understanding.
Warm regards,
Kim Lathan
Color/Digital Proofing Department
Manager
___________________________________________________________________________
Date: Wed, 12 May 2004 14:22:57 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 12, 2004, at 12:33 PM, Dan Margulis wrote:
No. They changed the default to *ignore completely*.
Changed from what to what and when? Photoshop 6, 7 and
CS have the same CMYK policy, Off, by default. That means don't embed
profiles, and ignore embedded profiles.
And really what we need are workflows that do a vastly
better job of ignoring CMYK profiles by default with a higher barrier to
turning this feature on. That would actually make it possible to safely
embed profiles in CMYK images, which could then be utilized by those who
know how to leverage them.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 12 May 2004 15:35:18 -0600
From: Jim Donovan
Subject: Re: Re: ICC for CTF & CTP
Good truthful reality call dan,as usual. I sometimes
get the impression that people actually believe that quality pre-profile
4/c reproduction never happened and was not possible.The fact is coffee
table quality is done daily completely ignoring cmyk profiles. I think alot
of people don't see any reason to dive into cmyk profiles when they do top
notch work without them.Especially if they spun color for a living without
cmyk profiles(or mac's for that matter!!) Don't think people resist
change,just have a realistic "if it aint broke don't fix
it",approach. jim donovan
___________________________________________________________________________
Date: Wed, 12 May 2004 17:49:14 -0400
From: Dan Margulis
Subject: Re: Re: ICC for CTF & CTP
Chris Murphy writes,
Changed from what to what and when? Photoshop 6, 7 and
CS have the same
CMYK policy, Off, by default. That means don't embed
profiles, and
ignore embedded profiles.
a) in previous versions this is only true if one sticks
with web defaults, and there is a strong prompt to avoid this. Therefore,
one could call anyone who didn't do so a "fool".
b) in previous versions, even if one used the web
defaults, the embedded profiles were not ignored--there was an alert.
Therefore, anyone who ignored the alert could be called an
"idiot."
In contrast, Photoshop CS, all settings, by default
ignores the CMYK profile altogether, no muss, no bother. Hence, if anyone
embeds a CMYK profile now, it would be nonsensical to assume that anyone
will pay attention to it unless they have been specifically directed to.
And really what we need are workflows that do a vastly
better job of
ignoring CMYK profiles by default with a higher barrier
to turning this
feature on. That would actually make it possible to
safely embed
profiles in CMYK images, which could then be utilized
by those who know
how to leverage them.
Yes. That's the tragedy of the whole thing--it could
have been avoided and we could have been given a most useful tool. The
catastrophic blunder that doomed the whole concept was, as Kim pointed out,
that novices were coerced into embedding tags that they didn't understand,
thereby discrediting the whole workflow.
Even today, there could still be major improvements
made, for example,
*Even if your color management is turned OFF, OFF, OFF,
Photoshop still embeds a profile by default if you have used the Convert to
Profile command, and it has to be disabled manually.
*So many files have wrong profiles attached that
anybody who receives a lot of files from strangers can't realistically ask
for an alert on every file. What is needed is a) separate alert preferences
for CMYK and RGB; and b) A way to respond to the alert saying "OK,
don't alert me to this particular profile any more until the next time I
quit Photoshop."
Dan Margulis
___________________________________________________________________________
Date: Wed, 12 May 2004 18:51:12 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 12, 2004, at 3:35 PM, Jim Donovan wrote:
Good truthful reality call dan,as usual. I sometimes
get the
impression that people actually believe that quality
pre-profile 4/c
reproduction never happened and was not possible.
No one actually believes this. But making color
separations is easier today than it used to be. The gotcha is that today we
have a whole lot of output behaviors compared to the past.
The fact is coffee
table quality is done daily completely ignoring cmyk
profiles.
That's news to me. Someone, somewhere it using a
profile. Either they're using an ICC profile, or they're using some kind of
proprietary table in a scanner, but either way it's a profile.
I think alot of people don't see any reason to dive
into cmyk profiles
when they do top notch work without them.
With how many proof iterations? People do not get good
separations for actual press conditions without some kind of profile.
Especially if they spun
color for a living without cmyk profiles(or mac's for
that matter!!)
Don't think people resist change,just have a realistic
"if it aint
broke don't fix it",approach. jim donovan
The notion that CMYK equates to a specific color is one
that's been broken and discarded for a long time now. CMYK numbers without
a reference point do not equate to specific colors. You make a generic
separation using Photoshop defaults, you end up with a proof that you see
needs to be edited sometimes significantly, before going to press. Yes
that's what a proof is for, but if there had been an accurate profile to
begin with the separation would have been a whole lot better in the first
place.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Thu, 13 May 2004 06:08:11 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: Dan Margulis
No, you've picked up the wrong quote. I was referring
to the case where I
would submit a CMYK file to a printer with instructions
to print it exactly as
is, notwithstanding that I did not know what the CMYK
conditions were. The print
house hypothetically needs to convert to RGB in order
to pull a proof.
Both you and the other advocate of this workflow
indicated that you wanted me to embed a profile so that it could be used to
make an accurate conversion to RGB so that the RGB
output would match what I saw on my screen. But,
as the objective is not to
match my screen but rather the final CMYK output, this
would have destroyed the
job. The proper decision would have been to treat the
file as untagged.
If I treat it as untagged and convert to RGB, doesn't
the process assume whatever default CMYK color space setting is chosen in
PhotoShop's color settings anyway? If you made your file for newsprint, and
I convert to RGB without your file being tagged, I'll get the wrong result.
The conversion to RGB has to assume something about your file. Why not give
it the proper information?
john c.
___________________________________________________________________________
Date: Thu, 13 May 2004 05:12:08 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: "Lathan, Kim"
So I find it very unfair that you call people idiots.
You can do/say what
you want, but that is an awful way to portray a group
of people who may or
may not have knowledge of a piece of technology that
you find so easy to
understand. Are there people using it incorrectly or do
not understand? YES!
But education and working together on this is the key
to understanding.
Kim,
I don't think I called people who don't understand
color management idiots, rather those who would hose my job because they
don't know enough to turn it off if they don't understand it, are. We're
not talking about minimum wage people here. These are professionals who do
this for a living, there are big bucks being spent, and one would assume
that for this degree of responsibility those in charge would know enough by
now to either turn it off or do it right.
john c.
___________________________________________________________________________
Date: Wed, 12 May 2004 19:07:54 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 12, 2004, at 3:49 PM, Dan Margulis wrote:
a) in previous versions this is only true if one sticks
with web
defaults, and there is a strong prompt to avoid this.
Therefore, one
could call anyone who didn't do so a "fool".
b) in previous versions, even if one used the web
defaults, the
embedded profiles were not ignored--there was an alert.
Therefore,
anyone who ignored the alert could be called an
"idiot."
I see what you mean. The profile mismatch warning is
disabled by default in Photoshop CS, but not in previous versions. The
default setting of the warning dialog, however, is still to ignore the
embedded profile and use working CMYK instead.
In contrast, Photoshop CS, all settings, by default
ignores the CMYK
profile altogether, no muss, no bother. Hence, if
anyone embeds a CMYK
profile now, it would be nonsensical to assume that
anyone will pay
attention to it unless they have been specifically
directed to.
Right.
Yes. That's the tragedy of the whole thing--it could
have been avoided
and we could have been given a most useful tool. The
catastrophic
blunder that doomed the whole concept was, as Kim
pointed out, that
novices were coerced into embedding tags that they
didn't understand,
thereby discrediting the whole workflow.
The short term solution of not embedding profiles in
CMYK images makes sense until the root cause of problems associated with
embedded profiles in a given workflow are found and solved. It does not
make sense to blame metadata (embedded ICC profile in this case) for
problems, but rather improper action on this metadata. Metadata itself does
not change data, it merely describes it. If something wrong is happening,
it technically isn't fair to blame the metadata.
Even today, there could still be major improvements
made, for example,
*Even if your color management is turned OFF, OFF, OFF,
Photoshop
still embeds a profile by default if you have used the
Convert to
Profile command, and it has to be disabled manually.
The logic behind it is that it's the conversion
equivalent of assign profile. If you have an untagged document, then
explicitly tag it using assign, the profile is embedded by default. If you
have any image and convert to something other than the working CMYK space,
then the thought is that this image should have its numeric values defined
by embedding a profile by default. It's not something that's going to work
for everyone.
*So many files have wrong profiles attached that
anybody who receives
a lot of files from strangers can't realistically ask
for an alert on
every file.
Wrong profiles attached? Do you have an example of this
and how it's occurring?
What is needed is a) separate alert preferences for
CMYK and RGB; and
b) A way to respond to the alert saying "OK, don't
alert me to this
particular profile any more until the next time I quit
Photoshop."
I'd still consider this a work around for fundamentally
busted behavior in various products. I'd rather see those flaws fixed
rather than find work arounds.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Thu, 13 May 2004 11:18:07 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 13, 2004, at 3:12 AM, jc castronovo wrote:
I don't think I called people who don't understand
color management idiots,
rather those who would hose my job because they don't
know enough to turn it
off if they don't understand it, are. We're not talking
about minimum wage
people here. These are professionals who do this for a
living, there are big
bucks being spent, and one would assume that for this
degree of
responsibility those in charge would know enough by now
to either turn it
off or do it right.
Yeah, I don't disagree with your perspective on this.
And I have a certain appreciation for those willing to die on hills like
this one. But I don't think it's really practical for me to suggest
everyone embed profiles in everything, and if the so-called professional
hoses the job, that people should sue. Maybe lawsuits would be an effective
means of coersion, but it's not merely the pre-press or print professionals
fault when this stuff doesn't work correctly. There are bugs in software,
and B.S. licensing agreements get software developers off the hook.
I think for now it makes more sense to not embed
profiles in CMYK images without having an explicit conversation with the
printer. It also makes sense to include the CMYK profile for those images
as a free floating profile in the image folder for the job. And while this
is a stop gap solution we will have to live with for years, we should be
trying to figure out where the problems are with embedded profiles and get
them fixed so that we don't need silly work arounds.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Thu, 13 May 2004 15:31:09 EDT
From: Dan Margulis
Subject: Re: Re: ICC for CTF & CTP
Chris Murphy writes,
Wrong profiles attached? Do you have an example of this
and how it's occurring?
Anybody using any of the "prepress defaults"
for PS 6-7; almost anybody using a built-in digicam profile; anybody who's
using Convert to Profile to generate a specific kind of black and doesn't
realize that a profile is being embedded by default; most of those who have
tried to custom-profile digicams; anybody using PS 5; anybody using PS7 or
CS who owns a consumer digicam and hasn't figured out how to ignore EXIF
profiles, etc., etc.
Plus, any time that the profile is arguably correct but
yet the end user makes an informed decision to ignore it (and that happens
with most profiled CMYK, and often with custom digicam profiles), then for
workflow purposes the profile is wrong. Or, if the vendor is using a very
slightly modified version of a common profile (say PS5 Adobe RGB vs. PS6
Adobe RGB, or Nikon-sRGB vs. sRGB) one would often like to eliminate the
weird name, and for such a purpose the original profile is
"wrong."
If you're in a workflow, as many of us are, where you
receive 50 or more images from the same client, and you open the first
three of them and up pops an sRGB profile which is clearly wrong, or a
Nikon-sRGB profile that seems no different from normal sRGB, the profile
mismatch warning box that pops up every time gets old real fast. So, if you
don't wish to see it 50 more times, you have to turn the warning off and
set RGB color management to ignore, and of course that's too bad if in
addition to the 48 images tagged with the wrong profile they've snuck in
two others from a different photographer who has tagged the images
correctly with a different profile.
Obviously, what is needed is a way to tell Photoshop,
"Temporarily stop bothering me with useless alerts about this
particular tag, but please let me know if you encounter some other tag that
doesn't agree with my settings."
The short term solution of not embedding profiles in
CMYK images makes
sense until the root cause of problems associated with
embedded
profiles in a given workflow are found and solved. It
does not make
sense to blame metadata (embedded ICC profile in this
case) for
problems, but rather improper action on this metadata.
Metadata itself
does not change data, it merely describes it. If
something wrong is
happening, it technically isn't fair to blame the
metadata.
Agreed.
Dan Margulis
___________________________________________________________________________
Date: Thu, 13 May 2004 15:15:59 EDT
From: Dan Margulis
Subject: Re: Re: ICC for CTF & CTP
John Castronovo writes,
If I treat it as untagged and convert to RGB, doesn't
the process assume
whatever default CMYK color space setting is chosen in
PhotoShop's color
settings anyway?
Yes. Whatever the printer's default settings are are
the that should be used in this particular scenario, not the embedded
profile.
If you made your file for newsprint, and I convert to
RGB without your file
being tagged, I'll get the wrong result.
This completely misapprehends the purpose of a proof,
which is to predict the appearance of the final output, whether good, bad,
or indifferent. Instead, you seem to feel that it is your job to make the
proof look as pretty as possible. This job is to be output CMYK down the
line without conversion. If you use the embedded newsprint profile for
proofing, the proof will be vastly lighter than the final output, and the
job will be a total loss.
You can now go back to calling those print shops
"idiots" who have not learned enough about embedded CMYK profiles
to use them to massacre jobs in the way you have just advocated; and those
users "fools" who are unwilling to invite such catastrophes by
putting such files into the hands of those who don't know how to use them
correctly.
Dan Margulis
___________________________________________________________________________
Date: Fri, 14 May 2004 00:31:34 -0400
From:John Castronovo
Subject: Re: Re: ICC for CTF & CTP
From: Dan Margulis
This completely misapprehends the purpose of a proof,
which is to predictthe
appearance of the final output, whether good, bad, or
indifferent. Instead,
you seem to feel that it is your job to make the proof
look as pretty as
possible. This job is to be output CMYK down the line
without conversion.
If you use the embedded newsprint profile for proofing,
the proof will be vastly lighter
than the final output, and the job will be a total
loss.
Yes, now that I read the requirements of this test
again, you are correct. In this case we're trying to see what's in these
plates and not trying to go back to the original image. What I do every day
is the latter, and the inclusion of a profile solves problems around here,
not creates them. Putting the profile info in the image folder as Chris
suggests is a welcome work around, but it would be better if it could
travel more closely with the file where those who need it can get at it,
and those who want to ignore it can.
You can now go back to calling those print shops
"idiots" who have not
learned enough about embedded CMYK profiles . . .
Hold on there, Dan. I don't think I called people
idiots who don't embed profiles. If I did, I didn't mean it that way. My
anger was directed at those who need to ignore them and don't know how to
do that. It's because of them that profiles can become a problem, so then
files don't get tagged, which in turn makes my life more difficult.
john c.
___________________________________________________________________________
Date: Mon, 17 May 2004 13:31:55 -0600
From: Chris Murphy
Subject: Re: Re: ICC for CTF & CTP
On May 13, 2004, at 1:31 PM, Dan Margulis wrote:
Anybody using any of the "prepress defaults"
for PS 6-7;
Why are they wrong?
almost anybody using
a built-in digicam profile;
Fair enough.
anybody who's using Convert to Profile to
generate a specific kind of black and doesn't realize
that a profile is being
embedded by default;
Why is it the wrong profile?
most of those who have tried to custom-profile
digicams;
Maybe.
anybody using PS 5;
That's pretty sketchy. Just because it's easy for the
wrong profile to get embedded doesn't mean they're always wrong.
anybody using PS7 or CS who owns a consumer
digicam and hasn't
figured out how to ignore EXIF profiles, etc., etc.
There are no EXIF profiles. It's an implied profile
only used by things that support that EXIF tag. It doesn't cause profiles
to get embedded at all.
Plus, any time that the profile is arguably correct but
yet the end user
makes an informed decision to ignore it (and that
happens with most profiled CMYK,
and often with custom digicam profiles), then for
workflow purposes the
profile is wrong.
Using that logic, you could argue that embedded
profiles could be at any moment so useless that there is no point. And
that's not something I recall you ever saying. That there are clear
workflow failures regarded embedded profiles notwithstanding, I don't think
it's particularly useful to start calling embedded but ignored profiles
"wrong". They still describe color appearance and bring meaning
to otherwise ambiguous RGB or CMYK values, that's all they do. If someone
doesn't want to go with that suggestion, they have the option to ignore at
their own risk or benefit. The barrier to their usefulness is that they can
be, superficially anyway, seemingly random in their risk/benefit ratio.
Or, if the vendor is using a very slightly modified
version of a
common profile (say PS5 Adobe RGB vs. PS6 Adobe RGB, or
Nikon-sRGB vs. sRGB)
one would often like to eliminate the weird name, and
for such a purpose the
original profile is "wrong."
At least for RGB spaces, ACE will not convert the file
if the source and destination are substantially the same. But this is a
pain in the ass when getting bogus profile mismatch warnings.
Obviously, what is needed is a way to tell Photoshop,
"Temporarily stop
bothering me with useless alerts about this particular
tag, but please let me know
if you encounter some other tag that doesn't agree with
my settings."
And a lawsuit to prevent companies from stealing the
intellectual property of other companies, stripping copyright tags and then
renaming them, which among other things sends them the bill for this
additional development. This is something Adobe could certain do better,
but the origin of the problem is not theirs.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Wed, 19 May 2004 20:40:16 -0000
From: "colormatik"
Subject: ICC for CTP - does it really matter?
This is a question for those of you who knows the
printing house side of the issue: In our (Publishing) dealing with the
printing house the question of the image profile is discussed. I'd like to
know why this is still an issue, since to get onto plate (CTP process)
images go through at least three steps of transformation (layout
appl/Quark, Post-script, PDF), at every one of which the color management
systems is OFF which I know since this is done on our side of the
production process. Is after these three trasformations an embedded
profile (was it "right" or "wrong" -- let leave it out
of the question} still in and can it affect result (CMYK values) of ripping
on a printer side?
Igor Tsiperson.
Digital imaging specialist, prepress operator.
___________________________________________________________________________
Date: Tue, 25 May 2004 07:39:08 -0700
From: Doug Walker
Subject: taking charge of deliverable-workflow tweak
questions
On Wednesday, May 12, 2004, at 08:50 AM, Chris
Murphy wrote:
Photoshop of all versions have ignored tags in CMYK
images by default.
And in Photoshop 6, 7, and 8 when the policy is set to
Off, not only do
you get blatant purging of this metadata, but by
default Photoshop
doesn't embed a new profile into the image.
After following the various threads on this topic, and
being one photographer who has been bitten delivering AdobeRGB being
converted in OSX I have am pondering the benefits and drawbacks of either
delivering sRGB untagged vs. CMYK. The past two covers have been
washed out. Client is counting fingers each job.
A couple of questions.
Client is using a printer who is never quite sure which
of 7 web presses will be used on a job. No paper proofs used, only
PDF online proofing for traps and other glaring misteps....color not being
critically judged.
Soooo, I am taking charge.
File Delivery:
Two options remain since AdobeRGB delivery is being
damaged by someone converting and not honoring profile.
I can choose to deliver converted to sRGB or CMYK.
If I convert to either of these from AdobeRGB would I not wish to tag
at the Save As step as well? Isn't communication about the file
critical? The above statement concerns me.
Sharpening:
And since my client is not savvy with respects to
unsharp masking, and the output is low-coated 133 line catalog cover if I
use PKSharpener and apply output sharpening to the image, and my client
crops and uses a portion of the image (say uses 80% of the pixel array
delivered) will this be problematic? ie, will this then be
under-sharpened? Or would some proper output sharpening be fine no
matter how they crop it?
My client is placing my TIF in Quark 5.x, printing to
Postscript, Distilling through Adobe Distiller 5.x and then sending the
resultant PDF FTP to printer. Not sure if this is considered PDF-X or
not.
Thanks,
Doug Walker, FP
"Specializing in Corporate People on Location in a
Clean, Bold Classic Style!"
website: http://www.walkerphoto.com
Member, PPW, ASMP, APA SF
___________________________________________________________________________
Date: Tue, 25 May 2004 10:20:31 -0500
From: Bob Smith
Subject: Re: taking charge of deliverable-workflow
tweak questions
On May 25, 2004, at 9:39 AM, Doug Walker wrote:
I can choose to deliver converted to sRGB or CMYK.
When I send files into situations similar to what you
describe... in other words I have no clue what's ultimately going to happen
to the images... I supply one folder with tagged sRGB versions of the
images. I see no good reason not to tag RGB images. Another folder
contains untagged CMYK in my best guess of the conditions. I also
supply a gang printed inkjet proof of the CMYK versions as a simple rough
verification of what I intend the images to look like. This gives the
receiving party the option of handling the image most any way they want and
shows them a printed copy of what the image should look like. I've
encountered relatively few surprises when working this way. This is
an option that's all about minimizing the chances for things to go really
wrong. Works for me.
Bob Smith
___________________________________________________________________________
Date: Tue, 25 May 2004 11:53:48 -0400
From: Jim Rich
Subject: Re: taking charge of deliverable-workflow
tweak questions
Bob,
I am just curious does the printer ever choose
the srgb files and if so do they charge you for the conversion to CMYK?
Jim Rich
___________________________________________________________________________
Date: Tue, 25 May 2004 12:09:05 -0500
From: Bob Smith
Subject: Re: taking charge of deliverable-workflow
tweak questions
Jim,
I usually don't know which they use but in the few
cases where I do know, they've never used the RGB file. I'm typically
supplying images to some middleman in the production chain, only rarely to
the printer directly. No, I've never been charged or heard of anyone
being charged because they chose to use the RGB file instead.
Bob Smith
___________________________________________________________________________
Date: Tue, 25 May 2004 13:17:19 -0400
From: john.romano
Subject: Re: taking charge of deliverable-workflow
tweak questions
Hi Jim
As a printer We would use the Tagged RGB files, if they
were supplied. Having an Aim proof is a good practice too.
Oh and for Charges..... We don't charge if there is a
profile but if there isn't we whack em with an AA charge!!
No just kidding, supplied images usually get worked on
before page work. So we usually go one round of Color work before
proofing, after that everything is and AA. Having the profile lets us
quickly softproof and convert to our Profiles, getting proofs that look
like what the customer expects.
On one past job we received over 400 cmyk images, some
were tagged some not. Our policy at the time was, convert tagged
files and leave ones that don't have a tag alone, proof as is.... At page
proof customer marked up 50 % of the images for ColorCorrections, looking
back at all of the images that had profiles proofed Exactly the way the
customer wanted them, untagged didn't. I went back and assigned and
converted all the files that required work and I was done. Job printed very
well, another happy customer.
Regards,
John Romano
___________________________________________________________________________
Date: Tue, 25 May 2004 14:49:49 -0400
From: Jim Rich
Subject: Re: taking charge of deliverable-workflow
tweak questions
Hey John,
I work with a printer here locally and they much prefer
to get CMYK images and if they don1t they charge for converting RGB
files. I think the charge is like $5 to $10.
That is why I asked.
Jim Rich
___________________________________________________________________________
Date: Thu, 27 May 2004 22:10:35 -0400
From: Terry Wyse
Subject: Re: taking charge of deliverable-workflow
tweak questions
I'm sorry John, didn't you get the memo that all you
printers took a vote and unanimously rejected CMYK images with embedded
profiles? Shame on you.
John's example just illustrates to me that
"CMYK-embedded" does indeed work in the hands of a knowledgeable
prepress/scanner operator and that, rather then "CMYK-embedded"
being rejected out of KNOWLEDGE, it's more a case of being rejected out of
fear and ignorance.
"CMYK-embedded": copyright of WyseConsul and
Terry Wyse :-)
All in good fun as usual.
Cheers,
Terry
___________________________________________________________________________
Date: Fri, 28 May 2004 08:24:03 -0000
From: Stephen Marsh
Subject: Re: taking charge of deliverable-workflow
tweak questions
Terry Wyse wrote:
John's example just illustrates to me that
"CMYK-embedded" does indeed
work in the hands of a knowledgeable prepress/scanner
operator
Well, that's one possible conclusion.
The flipside is of course that users who prepare for
the target condition do not want their files being altered, so in the
stated workflow having no profile ensured a 'hands off' approach.
If the supplier of the data knows up front, or that you
state quite clearly as part of your delivery specs that you will alter
tagged CMYK data to suit the press conditions (is the target known to the
supplier?) then fair enough - but I would not like to be caught out blindly
with someone altering my files based on the CMYK tag...If I am not doing
the seps, then I may as well hand off an agreed flavour RGB instead of
sending CMYK to this shop.
Stephen Marsh.
___________________________________________________________________________
Date: Fri, 28 May 2004 09:11:06 -0400
From: Terry Wyse
Subject: Re: Re: taking charge of deliverable-workflow
tweak questions
I don't DISagree with you Stephen but have you 1) got a
hold of their custom press profile and prepared the image accordingly or if
not, 2) are you willing to accept the fact that your job will not reproduce
correctly if you've targeted the incorrect print conditions and haven't
given them a clue (embedded profile) of how the job is SUPPOSED to
reproduce? Even if you supply them a proof of how you expect the job to
print, this really isn't sufficient in my book. I frankly would not accept
any proof that comes through the door unless I could verify that it matches
some standard print condition (SWOP for example). Anybody supplying a proof
to a printer should be prepared to prove that their proof is valid or
supply a profile from that proof system.
The knowledgeable shop would see that your embedded
profile matches their (assumed) CMYK working space and would not touch the
file. If they get a profile mismatch, this same shop should be on the phone
with you asking you how they should proceed. On another list I suggested
that a more "moderate" approach would be to possibly supply
untagged CMYK BUT with the profiles supplied with the images so the
prepress operator would have the option of assessing your images via
soft-proofing. If they are CM-challenged they would likely just ignore your
supplied profile anyway.
I'm really not an "embed profiles at all
cost" kind of guy but I cringe at statements that flatly state
supplying "CMYK-embedded" images should be avoided like the
plague. I have to temper my idealism to today's reality but I'm looking
FORWARD to the day that supplying images with profiles embedded will be
embraced and become the SOP with most printers. Until we get to the day
that everybody agrees on a just a few target print conditions (TR001, TR004
for example) and that supplied images are separated accordingly, I don't
see many other options for getting around the current mess that CtP has
somewhat encouraged.
Regards,
Terry
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
___________________________________________________________________________
Date: Thu, 3 Jun 2004 07:03:31 -0700 (PDT)
From: Kathleen Helgert
Subject: CMYK Profiles
Terry Wyse writes;
I don’t DISagree with you Stephen but have you 1)
got
a hold of their custom press profile and prepared the
image accordingly or if not, 2)are you willing to
accept the fact that your job will not reproduce
correctly if you’ve
targeted the incorrect print conditions and
haven’t
given them a clue (embedded profile) of how the job is
SUPPOSED to reproduce? Even if you supply them a proof
of how you expect the job to print, this really
isn’t
sufficient in my book.”
This is exactly where my confusion about profiles comes
into play. If I receive a cmyk file with an embedded profile, and that
profile does not accurately describe my output conditions, why would I
honor that profile? I think the assumption in the industry is print what
you’ve been sent, do not alter the digital file. If I use the
embedded cmyk profile as “source profile” and convert to my
“correct destination profile”, haven’t I just altered the
actual numbers in the digital file? So wouldn’t it be more accurate
for me to ignore the embedded profile, and show you a proof of what your
digital files numbers depict on my output conditions?
Regards,
D Melem
___________________________________________________________________________
Date: Thu, 3 Jun 2004 11:09:42 -0700
From: Paul D. DeRocco
Subject: RE: CMYK Profiles
The profile describes the meaning of the numbers. If
you discard the profile, then you’re forcing a different arbitrary
interpretation on the numbers, changing the image’s appearance. Hence
the need, as you mention, of sending a proof back to see if the sender
likes the new appearance. It seems to me that if the embedded profile is
real, but the wrong profile, then you should convert, so the image looks as
close as possible to what the sender saw.
But it also seems to me that the whole idea of
transporting images in CMYK form is problematic because CMYK is a family of
device-dependent color spaces, and shouldn’t be applied until it is
known exactly what device the images will be printed on.
Ciao,
Paul D. DeRocco
___________________________________________________________________________
Date: Thu, 3 Jun 2004 17:15:12 -0400
From: Dragonfly Imaging & Printing
Subject: Re: CMYK Profiles
Paul,
Excellent point. I agree completely.
Many others have alluded to the dangers of embedded
cmyk profiles.
Maybe we should be supporting a common
device-independent color space.
However, that would mean the printer would be
responsible for the conversion.
Round and round we go ;-)
John Toles
http://www.dragonflyprinting.com/
http://www.dragonflygallery.ca/
___________________________________________________________________________
Date: Thu, 3 Jun 2004 17:33:02 -0700
From: Paul D. DeRocco
Subject: RE: CMYK Profiles
From: Dragonfly Imaging & Printing
Excellent point. I agree completely.
Many others have alluded to the dangers of embedded
cmyk profiles.
Maybe we should be supporting a common
device-independent color space.
However, that would mean the printer would be
responsible for the
conversion.
That’s what RGB spaces are for.
I do all my own printing on inkjets, so I’ve
never dealt with a press. But it seems to me that part of the job of the
printer (the guy who owns the press) is to know how to get the right colors
out of his equipment, which in the digital domain means having a CMYK
profile for his press/ink/paper and the craftsmanship to tweak the image to
deal with the gamut limitations. But from all I’ve read here, it
doesn’t sound like that’s how the industry works, due to a lot
of historical inertia.
Ciao,
Paul D. DeRocco
___________________________________________________________________________
Date: Thu, 3 Jun 2004 21:24:20 -0700
From: Rick Gordon
Subject: RE: CMYK Profiles
This ignores the fact that there are many aspects of
fine-tuning based on specific image characteristics that must be handled
either at the point of deciding HOW to separate (e.g., the black generation
curve) or after separating (special black plate sharpening, black contrast
boosting, etc.).
Different images require different choices. A
one-size-fits-all conversion process may not bring forth all that might be
brought out of an image, although it may be fine for many circumstances.
But as someone who often works on technical books, I can state
unequivocally that you can’t handle a screen capture the same way you
handle a photograph. And there are many other situations where the
difference may not be as dramatic, but they are there.
Rick Gordon
___________________________________________________________________________
Date: Fri, 4 Jun 2004 01:30:24 -0700
From: Paul D. DeRocco
Subject: RE: CMYK Profiles
What you’re saying is that the conversion from an
RGB image into a CMYK
image is even more specialized than even a single
device-dependent CMYK
profile can describe. I don’t doubt that. But
that merely makes the idea of
using a CMYK color space as a generalized image
interchange space even less
plausible.
Someone needs to have the craftsmanship needed to do
that conversion, and
that person can’t do it without knowledge of the
press/inks/paper, or
without one or more proof runs. So if I’m not the
printing expert, then I
don’t think it makes sense for me to deal in
anything but RGB images.
I can’t imagine that the generic CMYK profiles
that come with Photoshop are
all that useful, except for a rough estimate of the
gamut limits of a
particular printing process. But if I use such a
profile for that purpose, I
would still think it better to send the RGB image to
whomever is doing the
printing, since my CMYK profile isn’t accurate.
Ciao,
Paul D. DeRocco
___________________________________________________________________________
Date: Fri, 4 Jun 2004 02:16:11 -0600
From: Chris Murphy
Subject: Re: CMYK Profiles
On Jun 3, 2004, at 8:03 AM, Kathleen Helgert wrote:
This is exactly where my confusion about profiles
comes into play. If I receive a cmyk file with an
embedded profile, and that profile does not accurately
describe my output conditions, why would I honor that
profile?
The profile describes the numbers in the file. If your
device is not the same as the one the profile describes, then the image
needs to be converting using the embedded profile as source profile and a
profile for your output condition as destination profile.
I think the assumption in the industry is
print what you’ve been sent, do not alter the
digital
file.
Yes and that is what leads to unnecessary work. What we
really need are better applications, and superior color management
implementation to prevent the problems associated with the binary logic we
have today: convert all tagged CMYK or never convert tagged CMYK. There is
an in-between, and some level of artificial intelligence or heuristic
should be employed to selectively color manage files intelligently to give
us the benefit of repurposing without the fallout.
If I use the embedded cmyk profile as
“source
profile” and convert to my “correct
destination
profile”, haven’t I just altered the actual
numbers in
the digital file?
Yes.
So wouldn’t it be more accurate for
me to ignore the embedded profile, and show you a
proof of what your digital files numbers depict on my
output conditions?
And since your output conditions differ, the file will
proof poorly. The end result is color correction WHICH ALTERS THE NUMBERS
IN THE FILE, and multiple proof-correct iterations until the desired result
is obtained.
Chris Murphy
___________________________________________________________________________
Date: Fri, 04 Jun 2004 08:44:38 -0000
From: Stephen Marsh
Subject: Re: taking charge of deliverable-workflow
tweak questions
Terry Wyse wrote:
I don’t DISagree with you Stephen but have you 1)
got a hold of their
custom press profile and prepared the image accordingly
or if not,
I meet whatever delivery guidelines are set.
are you willing to accept the fact that your job will
not reproduce
correctly if you’ve targeted the incorrect print
conditions
Yes, this is how life has always been.
This issue predates ICC profiles, ICC profiles just
make it more obvious.
The knowledgeable shop would see that your embedded
profile matches
their (assumed) CMYK working space and would not touch
the file.
It sounds like that ‘over there’ it is
common practise to convert supplied CMYK files.
This is not the case here.
When folk supply CMYK here for output (layout files and
linked images or CMYK PDF or pre sep PDF etc), it is with the intention of
it simply being output, it is not up to others to convert the data (apart
from digi proofing purposes - but the final data that goes to film or plat
is still the same as sent by the client).
If they get a profile mismatch, this same shop should
be on the phone with
you asking you how they should proceed.
The layout file is linked to the images and the numbers
are simply output - the linked images ICC profile is meaningless in this
part of the workflow (separated output to film/plate in a CMYK PostScript
workflow). Or call this ‘early binding’ or whatever the vogue
terms are.
I’m really not an “embed profiles at all
cost” kind of guy but I cringe
at statements that flatly state supplying
“CMYK-embedded” images should
be avoided like the plague.
In the context of linked files to layouts for a
workflow that should not know or care - then the previously stated reasons
for not using them can sometimes outweigh any use they may have.
We can both agree that untagged and separate supply of
said profile is workable as an alternative to no information at all (if
folk really need this info).
I have to temper my idealism to today’s
reality but I’m looking FORWARD to the day that
supplying images with
profiles embedded will be embraced and become the SOP
with most printers.
Depends.
If you like your printer converting your files to their
profile, then fine.
If your printer is simply imaging your files numbers,
they could care less - which is perhaps why many printers do not embrace
ICC CM in this part of their production chain. They may be using ICC CM to
produce digi proofs, but they probably do not know it.
Terry, it comes down to this for me -
For many vendors and those sending out files, it is an
accepted practise that CMYK files are hands off and final and that the folk
who prepared them do actually want those numbers output.
All of a sudden, there is talk of folk taking it on
themself to change supplied files values.
This is a big departure from the standard workflow that
I am part of and have been commenting on.
For those who do need to repurpose, then I agree with
the ‘pro’ reasons for ICC CM.
For me, the real debate is what are the expectations of
the file supplier (honour numbers or honour the intent) and the
expectations of the printer (honour numbers or honour the intent).
Stephen Marsh.
___________________________________________________________________________
Date: Fri, 04 Jun 2004 09:07:37 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
Kathleen Helgert wrote:
This is exactly where my confusion about profiles
comes into play. If I receive a cmyk file with an
embedded profile, and that profile does not accurately
describe my output conditions, why would I honor that
profile?
Depends on what the client expects.
Does the client expect you to honour the files numbers,
or honour the visual intent of the numbers when linked to the tagged
profile?
Printing the CMYK image linked from layout software
with no ICC CM, the CMYK numbers are simply passed through (the ICC profile
should be meaningless to the production chain here).
Printing the CMYK image from Photoshop or other ICC
savvy software, one can choose to output the numbers like non ICC layout
software or to convert from the tagged space to your printers space.
I think the assumption in the industry is
print what you’ve been sent, do not alter the
digital
file.
This is my viewpoint.
Others on this list are in different workflows with
different viewpoints and a tagged image and conversions are par for the
course.
The problem is that when this issue comes up, it is not
always clear what workflow is in use and what the various parties
expectations are.
If I use the embedded cmyk profile as “source
profile” and convert to my “correct
destination
profile”, haven’t I just altered the actual
numbers in
the digital file?
Yes, to maintain as close as possible the visual intent
of the original numbers/profile.
So wouldn’t it be more accurate for
me to ignore the embedded profile, and show you a
proof of what your digital files numbers depict on my
output conditions?
Exactly Kathleen, this is a very common scenario and in
my experience the more common one. I agree.
This is not a game to force rounds of costly proofing
on clients.
If clients wish ICC CM to be used, then they need to
implement things correctly in-house and find service providers who can work
with them without hassle.
Or printers need to embrace this stuff to gain a
competitive edge over others who do not use ICC CM and simply output the
files numbers, so that they can cater to the market that does wish to have
their files numbers altered rather than simply output.
Stephen Marsh.
___________________________________________________________________________
Date: Fri, 04 Jun 2004 09:12:38 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
John Toles wrote:
Many others have alluded to the dangers of embedded
cmyk profiles.
Maybe we should be supporting a common
device-independent color space.
However, that would mean the printer would be
responsible for the
conversion. Round and round we go ;-)
No need for this John, the CMYK is good - that’s
the whole point we do not wish to alter it!!!
CMYK is not being used as a random delivery space in
this context, but as film/plate/press ready final data.
The issue is not really profiles - it is workflow and
expectations of client and service provider.
No profile is just a hack to help get the point across
that the file should not be ICC CM.
Stephen Marsh.
___________________________________________________________________________
Date: Fri, 4 Jun 2004 06:25:56 -0400
From: John Castronovo
Subject: Re: CMYK Profiles
Well said Chris. It illustrates the importance of
knowing as much about the file's destination as possible if one is to
provide untagged CMYK, but such information is frequently hidden from
suppliers. The point is one that I feel may have been lost on those who
completely oppose profiling. I agree with you that this is going to remain
a problem until the industry finds a better way to deal with the issue,
possibly as you state in the following:
There is an in-between, and some level of artificial
intelligence or
heuristic should be employed to selectively color
manage files intelligently
to give us the benefit of repurposing without the
fallout.
Thanks for your comments.
john c.
___________________________________________________________________________
Date: Fri, 04 Jun 2004 09:43:16 -0400
From: Thomas P. Kaczmarek
Subject: Re: CMYK Profiles
Hello all,
I've been following these printer profile threads with
interest and I can see logic to both sides of the discussion. As a creative
type, I've ALWAYS discussed the print job with the printer before
submitting the job, especially if I haven’t used that printer before.
Submitting a file blindly, I would have no recourse to question print
quality, and would expect going through rounds of color corrections from
the proof. Color Management, bleed, file format requirements are just some
of the items that need to be discussed prior to me handing that job off to
a printer. If I don’t do this I’m not a professional, period,
end of statement!
If I do speak with a printer who obviously does not
comprehend what I’m talking about I will not use them if possible. If
forced, as I have been by my clients, then I would try to idiot proof the
job as much as possible; no embedded profiles, clear instructions and a
proof that represents what I want the job to look like in CMYK.
If the the printer receives a file blindly, that
printer needs to call the client to verify their intentions and ask how to
handle the project. If the client is not a professional, the printer needs
to explain that extra proofs and or rounds of color corrections may be
required to achieve what the client may view as acceptable. If the printer
does not make that phone call, IMHO, he is not a professional.
Educated clients and vendors can make this process so
much easier and CMS can help us reduce the color correction cycle
drastically if it is implemented properly. To implement it we need to
discuss it, hence this forum, and help others help themselves and help us
do our jobs better, quicker and hopefully more profitably. I can’t
think of one instance where a phone call would not help, at least to
improve client/vendor relationships.
Humbly,
Tom Kaczmarek
___________________________________________________________________________
Date: Fri, 04 Jun 2004 06:54:43 -0600
From: Andrew Rodney
Subject: Re: CMYK Profiles
on 6/4/04 2:30 AM, Paul D. DeRocco wrote:
I can't imagine that the generic CMYK profiles that
come with Photoshop are
all that useful, except for a rough estimate of the
gamut limits of a
particular printing process.
Actually many are specifically produced to fingerprint
a well defined behavior like SWOP TR001. The problem is finding a printer
who first tells you they conform to SWOP (certainly one of the three
biggest lies in the world) and then finding they really do conform as
specified by TR001 (and soon TR004 for sheetfed). That being the case and
not a lie, the SWOP Coated v2 profile from Photoshop will produce a superb
conversion. This was illustrated several years ago at Seybold in a
profiling shootout.
Andrew Rodney
http://digitaldog.net/
___________________________________________________________________________
Date: Fri, 04 Jun 2004 07:07:48 -0600
From: Andrew Rodney
Subject: Re: Reverse engineering
on 6/4/04 6:32 AM, Kathleen Helgert wrote:
In regards to color seperations, I feel that
people on the creative end of the workflow have been
led to believe that a file should be able to be
"repurposed" for any possible use.
That's totally doable with RGB files in a working
space. That's why we have working spaces. Apply the right output profile as
many times as you need to output the file. The alternative is to scan the
sucker each time for each output device which is simply not productive
expect for the person making scans. It also ignores all the direct RGB
capture devices like digital cameras. There is no reason why a RGB file
can’t be output dozens of times and done with excellent color
results.
What is the files actual use? Obviously
"high-end"
printing requires more resolution than the web. So
reverse engineering in the thought process is the key.
Scan (shoot) once, use many. So at least capture enough
resolution for all the uses necessary.
In regards to the file not matching what is seen on
the clients screen if the profile isn't honored, that
seems nonsensical to me. Do I know if the clients
monitor is calibrated? How long ago? Is it a higher
end monitor or the Wal-Mart sale item?
We can't assume this but then we can't assume that
someone could take your nice 50mb RGB file, convert it to a 1mb JPEG with
quality level of 3 while pulling enough levels to Posterize the heck out of
the image.
This notion that someone may not calibrate their
display or use correct color management is a straw man argument since there
are plenty of other ways users hose their files. We don’t go about
suggesting we take Photoshop (or better Quark) away from users because
there are dangerous buttons that a user can push.
Don't get me wrong. In theory I understand the logic
of the numbers in the file being arbitrary until a
definition, a profile, is introduced. I just wonder if
the majority of commercial printers out there feel
comfortable converting the numbers in the file as a
standard practice. I'd like feedback on this from
them.
They don’t feel conformable because it's not in
their best interest. I's a lot easier to ignore the issues. There are
plenty of smart printers who “get it” and understand the above.
They provide sound color management practices and don’t mind working
with or educating their customers. If we agree that a digital image file is
just a big pile of numbers and the color is undefined without a descriptor,
how can we not look at what a profile can do and say “ah, useful in
the right hands.” ?
It just seems that in the past, the photographer,
designer, the creator of the piece worked with the
printer to seperate the art into the best possible
reproduction.
As someone that spent many years as a commercial
photographer, I can only recall one or two cases where this happened and
that was when I was printing my own promo cards (I was the customer).
Otherwise, once I handed the transparencies to the art director or other
client, what happened after wasn’t ever discussed with me. The job
either printed well (I was a hero) or it didn’t (and I would blame
the printer or separator). The film was what it was and past that, I
wasn’t involved at all in the repro side. That may have changed today
(but by and large, I seriously doubt it).
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Fri, 4 Jun 2004 10:00:07 -0700 (PDT)
From: Kathleen Helgert
Subject: RGB files with profiles
Let me clarify. RGB files with profiles make sense to
me. I have to convert this to cmyk anyhow, I might as well know what RGB
color space was used. Using that profile as the source and my profile as
the destination makes sense to me. I guess I feel the threshold is when a
file is converted from RGB to CMYK. The entire reason that conversion must
be made is because a conventional printing press cannot print anything but
CMYK. So if the creator of that image is unsure of who is printing the
file, why convert to CMYK?
In my experience, since computers let a file be changed
effortlessly from RGB to CMYK, designers have confusion about the
difference. Personally I feel this is apples and oranges. (I’ve had
designers mix colors in page layout programs using RGB) Two completely
different color modes with different gamuts. So is it realistic to expect
an RGB file to exactly match a CMYK file, even if profiles are used in the
conversion process? Isn’t “rendering intent” an issue
here?
But I don’t feel comfortable
“converting” a CMYK file so that I “match” what
someone is seeing on their monitor. Would two monitors side by side display
the exact same color even if they were calibrated?
___________________________________________________________________________
Date: Fri, 4 Jun 2004 14:20:14 -0600
From: Chris Murphy
Subject: Re: CMYK Profiles
On Jun 4, 2004, at 2:30 AM, Paul D. DeRocco wrote:
I can’t imagine that the generic CMYK profiles
that come with Photoshop are
all that useful, except for a rough estimate of the
gamut limits of a
particular printing process. But if I use such a
profile for that purpose, I
would still think it better to send the RGB image to
whomever is doing the
printing, since my CMYK profile isn’t accurate.
I’ve submitted jobs run to SWOP and sheetfed
proofed with Matchprints using U.S. Web Coated (SWOP) v2 and U.S. Sheetfed
Coated v2 respectively. They came out quite good. Actually for a Seybold
shootout I did, comparing a bunch of profiling packages, I also tested the
U.S. Sheetfed Coated v2 profile for grins. I gave it a slightly higher
score than any of the custom profiles. In this instance the method of
proofing (Creo Spectrum imaging commercial Matchprint media) paired well
with the behavior in the generic Adobe profile and that’s why it
worked well. It worked better because it had the smoothest results of any
of the profiles I had tested, but it was subtle.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Fri, 4 Jun 2004 14:31:58 -0600
From: Chris Murphy
Subject: Re: CMYK Profiles
On Jun 4, 2004, at 4:25 AM, jc castronovo wrote:
It illustrates the importance of knowing as much about
the file’s
destination as possible if one is to provide untagged
CMYK, but such
information is frequently hidden from suppliers. The
point is one that I
feel may have been lost on those who completely oppose
profiling. I agree
with you that this is going to remain a problem until
the industry finds a
better way to deal with the issue, possibly as you
state in the following:
I think the first step is a mandate that by default all
applications and pre-press equipment ignore embedded profiles in CMYK
images. Key word = by default.
The second step is for all of our applications to no
longer make the embedding of profiles optional. It should be required
information. That way it’s always there for people who want to use
it, but the scaredy cats who don’t understand the technology
can’t sabotage the world for the rest of us. Right now the GOOD IDEA
of tagging images is rejected because application vendors are doing the
wrong thing, and unknowledgeable people scared of profiles force all of us
to live with mediocrity. I do not want my images demoted to lowest common
denominator and untagged CMYK images are pretty close to lowest common
denominator.
In concert it means the data is available but ignored
by default until someone knowledgeable and brave enough wants to use it.
Those who can, benefit. Those who can’t, get the same old workflow
with no corresponding fall out. The lack of metadata in the media rich
world we have today begs for inherent inefficiency.
A big failure in my opinion of the ICC has been their
insistence they stay away from workflow issues. Enabling the technology was
not good enough. We are in the mess we’re in because so many vendors
have implemented the technology in different and sometimes conflicting ways
that causes catastrophic results.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
___________________________________________________________________________
Date: Fri, 4 Jun 2004 22:19:07 -0400
From: john.romano
Subject: Re: Re: CMYK Profiles
Stephen
Any good Prepress Operator is going to take your image
and Softproof it. If it looks good then it wont be touched but 9 out 10
times a suppled file needs work.
With a profile your image will be reproduced
accurately, without the color will be assumed by assigning something that
looks good and then converted.
Im so sick of the “Nobody “ should touch my
images attitude When you stand at the back end of a press on a daily basis
you will understand my point of view. Nobody knows the press conditions
better than the printer, end of story.
Not all printers are CM challenged, but I get the
impression from this list that there are a lot more Photographers and
designers that could use a “Clue “ or two. And they wont get it
here !!!
John
________________________________________________________________________
Date: Sat, 05 Jun 2004 09:46:58 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
Any good Prepress Operator is going to take your image
and Softproof it.
If it looks good then it wont be touched but 9 out 10
times a suppled file
needs work.
And I hope that they are going to leave the colour
numbers unchanged to see how the files numbers compare with what this
profile describes the conditions as being (with no tag this is the likely
outcome).
As previously stated, I am meeting thier delivery
guidelines.
With a profile your image will be reproduced
accurately, without the color
will be assumed by assigning something that looks good
and then converted.
No, the profile is meaningless once the file is in CMYK
to the
CMYK PostScript workflow for film/plate which I am
talking of.
I am only dealing with folk who output the files
numbers to film/plate from non ICC enabled layout software or from final
hands off output ready PDF files (composite or presep).
For digiproofing, they will assume the proofing/press
profile that represents their conditions to CMYK data and convert to the
inkjet space for proofing on the inkjet.
This should show the ‘same’ thing as the
preserve colour numbers softproof.
This profile would have been supplied to me as part of
the delivery requirements. This is what I am separating to. This is what
they are proofing to. This proof is the agreed sign off for colour. This is
what they are matching the press to (their proof which they say they can
match on press).
I am not talking of the odd loose scan here which can
greatly benefit from ICC metadata, but complete layout files.
Im so sick of the “Nobody” should touch my
images attitude
Perhaps you have not had the misfortune to have someone
hose your files when they should not have been touched and simply output as
is.
When you stand at the back end of a press on a daily
basis you will
understand my point of view.
Nobody knows the press conditions better than the
printer, end of story.
I can spend as much time in the pressroom as I like, it
is 30 seconds away from me (as inhouse prepress I have it lucky in this
respect), the issues I am talking of are prepress related and not press
related.
Stephen Marsh.
___________________________________________________________________________
Date: Sat, 5 Jun 2004 07:14:59 -0400
From: john.romano
Subject: Re: Re: CMYK Profiles
Hi Stephen
And I hope that they are going to leave the colour
numbers unchanged
to see how the files numbers compare with what this
profile describes
the conditions as being (with no tag this is the likely
outcome).
This would be how I would softproof, with or with out a
tag.
For digiproofing, they will assume the proofing/press
profile that
represents thier conditions to CMYK data and convert to
the inkjet
space for proofing on the inkjet.
This should show the ‘same’ thing as the
preserve colour numbers
softproof.
Yes it will and if your numbers dont look good then we
just wasted a proof !!
Ok so you are only talking Postscrip level CM, But what
about getting supplied files that are not ready for pages.Customer want to
see loose proofs from printer that match their supplied proofs?
So you work in Prepress ? You dont get supplied files ?
If so how do you handle them ?
Just proof them out ? wasting a proof and then color
correcting till they look good. Atleast another wasted proof and time
for an operator to work on images. Time is money along with the cost of
proofs, not talking Inkjets here but Approvals.
Perhaps you have not had the misfortune to have someone
hose your
files when they should not have been touched and simply
output as is.
No your right I havent had anyone Hose “My”
files, just been on the recieving end of said hosed file.And I would say
hosed file had nothing to do with a profile.
We have different concerns and thats OK, But all Im
saying is I am going to make the assesment, give me the tools to do it
right if I need too convert. Im not trying to make more work for myself but
when the need is there a profile will make my life a little easier.
Communication is key, without a profile I cant know
what your file is supposed to look like.Should I leave it alone when I
softproof it and it looks bad ?
Regards
John Romano
___________________________________________________________________________
Date: Sat, 05 Jun 2004 12:08:52 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
John Romano wrote:
And I hope that they are going to leave the colour
numbers unchanged
to see how the files numbers compare with what this
profile describes
the conditions as being (with no tag this is the likely
outcome).
This would be how I would softproof, with or with out a
tag.
Agreed, with recent talk of honouring the CMYK visual
intent which is described by the tagged profile over the CMYK numerical
values - I was not sure how common it was to convert the colour numbers
based on the tag to the proof space.
The stage of the workflow I have been commenting on
(over the years) is a specific, final one which concerns the separated
output of CMYK data from non ICC enabled layout files.
For digiproofing, they will assume the proofing/press
profile that
represents thier conditions to CMYK data and convert to
the inkjet
space for proofing on the inkjet.
This should show the ‘same’ thing as the
preserve colour numbers
softproof.
Yes it will and if your numbers dont look good then we
just wasted
a proof!!
The earlier use of softproofing should greatly cut down
on this, as does reading/knowing the files numbers and how they relate to
this particular process.
Ok so you are only talking Postscrip level CM, But what
about getting
supplied files that are not ready for pages.Customer
want to see loose
proofs from printer that match their supplied proofs ?
As I have previously stated over the years on this list
in many related threads, when it comes to file exchange and other matters
which are not directly related to the preservation of the numerical intent
of the output of press ready data - then I am all for more information
being supplied...in many forms...and for ICC work tagged is good, but I am
realistic enough to know when the separate supply may be more appropriate
(redundant data etc).
So you work in Prepress ? You dont get supplied files ?
If so how do you handle them ?
Depends on the client, budget, job brief, what I need
to do at the time and many other variables.
Tagged files can sometimes help, sometimes not
depending on variables. For file supply to me, then I am all for them in
general.
Perhaps you have not had
the misfortune to have someone hose your
files when they should not have been touched and simply
output as is.
No your right I havent had anyone Hose “My”
files, just been on the
recieving end of said hosed file.And I would say hosed
file had nothing to
do with a profile.
One job that I am still sour over was most likely an
unwanted conversion, but it could have been an unwanted edit. It does not
really matter, the files values were changed which stuffed up the entire
look of the job. In hindsight I should have discussed the special
relationship between the CMY and K plates on this part of the job (it was
not important enough to the client for a press approval), but I did not
expect them to monkey with the file.
We have different concerns and thats OK, But all Im
saying is I am going to
make the assesment, give me the tools to do it right if
I need too convert.
Im not trying to make more work for myself but when the
need is there a
profile will make my life a little easier.
Agreed, if the supplier wishes you to make this call
then the CMYK profile is a big aid. If the image supplier is intentionally
not using a CMYK profile for the limited insurance reasons it offers for a
specific output orientated workflow, then I see that as fine too.
Communication is key,
I have said this many times on the list over the years
- in verbal, written (hard copy and filenames, folders, read me files etc)
and metadata form where possible.
without a profile I cant know what your file is
supposed to look like.Should I leave it alone when I
softproof it and it
looks bad ?
If that is the expectation of supplier or job brief,
then yes!
By all means raise your concerns before ink hits paper
with those higher up in the food chain (internally or client) - this is the
beauty of softproofing and proofing.
Stephen Marsh.
___________________________________________________________________________
Date: Sat, 5 Jun 2004 14:51:10 -0700
From: Dennis Dunbar
Subject: Re: CMYK profiles
From: Kathleen Helgert
In regards to color seperations, I feel that
people on the creative end of the workflow have been
led to believe that a file should be able to be
“repurposed” for any possible use.
Personally I feel
that this is like the old “jack-of-all-trades,
master
of none” logic.
Most folks would agree with this. But the reality for
creatives is that clients are going to repurpose the files in many
many cases. If the file were bound for just one use and it were possible to
communicate with the next shop in the workflow chain this whole question
would be easy to answer.
For instance I work in entertainment advertising
creating final art for movie posters, ads, DVD packaging etc. In a recent
ad campaign I created an image to be used for a magazine ad promoting an
upcoming movie. The file was prepared for the expected conditions of the
publications and created at the necessary size for the ad. Anticipating
that this image would be used for a teaser poster I created a higher res
version that would work well at 27x40. When the poster was printed the
client said they just used the first version we gave them and ran the job
without ever asking for a new file. So the lower res version that was
prepped for SWOP was re-purposed for a much larger reproduction printed on
a Sheetfed press.
Often times the clients do not know the difference yet
they expect a high end result no matter what. Fortunately in entertainment
advertising they have the budgets to work with good printers who generally
do a good job of getting it right. But in other industries we’re not
so lucky.
Given this situation there is a BIG need for some
consensus on what to provide in these cases. Since this forum has a wide
range of voices who come from many different levels of experience what
would you all recommend? Tell the client you only give RGB files when they
ask for CMYK? Give them CMYK prepped for something close to what you hope
will be the final conditions and embed a profile? Or give them CMYK blindly
prepped so it looks OK for your set up, not embed a profile and pray the
next one down the line does a good job?
Again it’s easy if we know where and who and what
conditions the job will be printed under. But most of the time in my
experience I only get to know they want “CMYK” and I am certain
the file is going to be repurposed without any control or input on my part.
That’s when it gets tough.
Dennis Dunbar
APA Digital Chair
________________________________________________________________________
Date: Mon, 7 Jun 2004 17:12:31 EDT
From: Dan Margulis
Subject: Re: Re: CMYK profiles
Dennis writes,
In a recent ad campaign I
created an image to be used for a magazine ad promoting
an upcoming
movie. The file was prepared for the expected
conditions of the
publications and created at the necessary size for the
ad. Anticipating
that this image would be used for a teaser poster I
created a higher
res version that would work well at 27x40. When the
poster was printed
the client said they just used the first version we
gave them and ran
the job without ever asking for a new file. So the
lower res version
that was prepped for SWOP was re-purposed for a much
larger
reproduction printed on a Sheetfed press.
You should get rid of this client. They have an
obligation to be better informed than this. I suspect that they did this on
purpose, to sabotage your work. It is a real disgrace how clients have
failed to keep up with the times and how anti-progress they are. If you
dump this client, maybe others will wake up and we can finally get
somewhere in the graphics industry.
Right.
Odd, isn’t it, that we tolerate clients whose
work procedures aren’t to our liking, but the minute people
disapprove of what a printer is doing, they spout pap just as extreme as
the above?
Given this situation there is a BIG need for some
consensus on what to
provide in these cases. Since this forum has a wide
range of voices who
come from many different levels of experience what
would you all
recommend? Tell the client you only give RGB files when
they ask for
CMYK? Give them CMYK prepped for something close to
what you hope will
be the final conditions and embed a profile? Or give
them CMYK blindly
prepped so it looks OK for your set up, not embed a
profile and pray
the next one down the line does a good job?
There *is* a consensus, accepted by all but the most
wild-eyed extremists. You've given three options. #2, the embedded CMYK
profile approach, is obsolete. It died several years ago and the official
funeral service was Photoshop CS. As for #1, submitting RGB, here is the
current conventional wisdom.
"At this stage in the evolution of the industry,
I’d have to say that the only rule of thumb for supplying RGB is,
don’t even think about doing it unless you have an ironclad guarantee
from the recipient that they know how to handle it. If the recipients know
how to handle it, any profiled RGB space should work just fine. If the
recipients open supplied RGB in Photoshop 4 and convert to CMYK on default
settings (which is way more common than one would hope) all RGB spaces are
equally dangerous."
The last sentence of the above is wrong—in the
scenario being described, wide-gamut RGBs (read: Adobe RGB) are very much
worse than other kinds. However, all the rest of the above might reasonably
be taken as something that I had written myself in 1998 or at any point
thereafter. In fact, it's written by Bruce Fraser, just a couple of weeks
ago. As Bruce has not exactly made a name for himself as an opponent
of color management systems, when *he* reaches a position like this,
that’s a consensus.
Since of the three options you presented, two are not
viable, you are left with the third. Yes, knowing the exact CMYK
destination sometimes helps. Admittedly, clients should be more intelligent
than they are. And most assuredly, chocolate ice cream shouldn't contain
carbohydrates. But we work with what we have, and none of these factors are
killers. If you have a good generic CMYK file, it will print well anywhere.
If you don't have a good file, all the knowledge of output conditions in
the world won't help.
Dan Margulis
________________________________________________________________________
Date: Mon, 07 Jun 2004 15:58:04 -0700
From: Peter Figen
Subject: Re: Re: CMYK profiles
I recently had a situation where a sculptor client of
mine had given RGB scans I had given her and 4X5 transparencies to the
printer for a catalog book she is printing.
The first round of test proofs were so horrible that
they were not even fixable. Neither the supplied RGB nor their scans
resembled anything acceptable. I intervened and made a custom profile of
the proofing system and when I picked up the printed target, I peeked into
their Photoshop color preferences. In PS 7, I found they had it set to
ignore any profiles, RGB or CMYK and that their conversion to CMYK was
using the PS4CMYK Default, which was not even close for their new, post
DTP, proofing system. So, in effect, they were making two false
assumptions, one being that any RGB they encountered was Apple RGB,
and the second being the conversion using the wrong setup or profile. Their
Crossfield scans had the same bad color as their PS conversions only
with plugged shadows and 383% ink densities. When I finished the Quark
document last Friday, I realized that I had 44 pages of CMYK tiffs with
embedded profiles that might be a problem. I called the printer and asked
their prepress guy how they treated tagged files and his response was that
sometimes they convert them and sometimes they don’t, but he
didn’t really know what a profile was or what it did or didn’t
do. Worse, was the fact that he didn’t know if their RIP ignored
profiles of did something to the data. I immediately dropped the folder of
images onto the Remove Profile AppleScript and had to re-update the entire
book.
This is not to say that anyone is right or wrong, just
to illustrate one real world situation of one person’s experience
with one printer. I should add that I did simulate their workflow by
assuming the wrong profile and making the same CMYK conversion and then
assigning my new custom profile to the resulting file. Assigning the new
profile gave me an on screen view that was virtually the same as their
original bad proofs. Maybe after the book is printed I’ll go in and
show them what I did.
Peter Figen
________________________________________________________________________
Date: Tue, 8 Jun 2004 10:39:32 +1200
From: Nick T
Subject: Re: Re: CMYK profiles
Hi Dan
So if I’m supplying a good "generic
CMYK" (and I usually supply one of the 4 V2 flavours), Should I go
with the lower TIL of one of the WEB seps and hope that someone will add
more ink later or should I go with a sheetfed sep and run the risk of too
much ink on a low quality stock?
Regards
Nick Tresidder
________________________________________________________________________
Date: Tue, 8 Jun 2004 07:36:13 -0600
From: Les De Moss
Subject: RE: CMYK Profiles
(Dennis or Dan?... unsure who posted this comment)
It is a real disgrace how clients have failed to keep
up with the times
and how anti-progress they are. If you dump this
client, maybe others will
wake up and we can finally get somewhere in the
graphics industry.
*************
To lambaste a client for their lack of knowledge seems
supremely counter-intuitive; a business-model that would fail to stand any
test of time. I doubt that dumping clients for their ignorance would serve
to do anything positive for our industry, or any particular shop.
One of our jobs as leading-edge professionals in the
color industry is to educate our clientele as new technologies cause
fundamental change in all of our workflows.
It was not that long ago that many on this list didn't
know the difference between a color space and a color picker. Many years of
discussion and development have brought us to where we are today...and yet,
even among professionals, there remains many areas of debate and
misunderstanding; many things we have only recently begun to successfully
employ in day to day workflows. It is no wonder that those we serve trail
us in their understanding and implementation.
That said, I fully agree that when a client is
responsible for hosing a job (or 'sabotaging' as Dan suggested in this
case), the responsibility must be appropriately placed on their shoulders.
I would add, however, that this also presents an opportunity to both
educate the client, and review (our part of) the job in retrospect. Did I
fully prepare the client with the information necessary for success down
the road? If not, I'd review my methods and encourage them to do the same.
Les De Moss
________________________________________________________________________
Date: Tue, 8 Jun 2004 10:48:18 EDT
From: Dan Margulis
Subject: Re: Re: CMYK profiles
Nick writes,
Hi Dan
So if I'm supplying a good "generic
CMYK" (and I usually supply one of
the 4 V2 flavours), Should I go with the lower TIL of
one of the WEB
seps and hope that someone will add more ink later or
should I go with
a sheetfed sep and run the risk of too much ink on a
low quality stock?
If you think that a very important destination of the
file is going to be an offset press, and you're not sure of what the
characteristics of that press are, then you guess that its dot gain is
high, meaning, you send a slightly lighter file than probably is needed.
Note: "Lighter" does not mean "washed out". The shadow
is still to be full value; the midtone is what should be slightly lighter.
The reason: it's always far better to have the pressmen
increase ink if necessary than decrease it. If it looks too light to them
when the print it and they start hiking ink densities, it may look bad but
it may equally well look better even than you were hoping. If they start
cutting densities because your ori ginal looks too dark, it may be better
than doing nothing, but the shadow value will die and the picture will be
too flat.
The Photoshop supplied CMYK profiles are not good IMHO
for dealing with an unknown printer. The web profile may give a sep that's
appropriately dark for a quality-oriented web printer, but it will be too
muddy for a shop that isn't under good control. Worse, the black is too
heavy. With unknown printers, black density is often very problematic.
Also, one would like to leave open the possibility that the printer can
increase the black for aesthetic reasons, particularly if it gets printed
sheetfed. With a traditional Light GCR setting, if the printer feels he can
punch up the black, the picture will look better for it. With one of the
built-in profiles, it won't work--it will louse up too many colored areas.
Dan Margulis
________________________________________________________________________
Date: Tue, 8 Jun 2004 10:48:22 EDT
From: Dan Margulis
Subject: Re: Re: CMYK profiles
Peter Figen writes,
I recently had a situation where a sculptor client of
mine had given RGB scans I had
given her and 4X5 transparencies to the printer for a
catalog book she is printing.
The first round of test proofs were so horrible that
they were not even fixable.
Neither the supplied RGB nor their scans resembled
anything acceptable. I intervened
and made a custom profile of the proofing system and
when I picked up the printed
target, I peeked into their Photoshop color
preferences. In PS 7, I found they had it
set to ignore any profiles, RGB or CMYK and that their
conversion to CMYK was using
the PS4CMYK Default, which was not even close for their
new, post DTP, proofing
system.
For those photographers reading the above, let me
suggest that you put yourself in the client's position. She certainly
understands that the proofs are garbage. Whether she can understand the
above explanation as to why is unclear. The printer's explanation will be a
lot simpler, namely, that Peter's work was no good. It's exceedingly
difficult for clients to figure out who is really to blame in a situation
like this. It is, however, easier to fire a photographer than a printer.
I suspect that the list is tipping its hat to Peter for
going so far above and beyond the call of duty. May I respectfully suggest
that sadder but more realistic view is that, if he had not been able to do
these things, Peter might very well have taken the fall for the failure of
this job.
This is not to say that anyone is right or wrong, just
to illustrate one real world
situation of one person's experience with one printer.
I should add that I did
simulate their workflow by assuming the wrong profile
and making the same CMYK
conversion and then assigning my new custom profile to
the resulting file.
It's extremely real-world, although it may not seem so
to readers of the list. In the workflow Peter describes, a disaster is
about to happen. Three parties can realistically prevent it: the client,
the photographer, and the printer. The client is unlikely to know what to
do, leaving it up to photographer and printer. The photographers and
printers who frequent this list have indicated just by subscribing that
they are way more interested in these topics than many of their colleagues.
They may therefore overestimate the ability of said colleagues in the real
world.
No question, a large number of printers, possibly even
a majority, could not deal with this situation. An overwhelming majority of
photographers, IMHO, Peter being the happy exception, could not deal with
it either.
Those photographers and printers who have chosen to
remain in ignorance are not wicked people. I have never met any
photographer or any printer, however incompetent, who wasn't trying to
satisfy the client to the best of their limited abilities. Such people can
be worked with--but only if we ourselves are able to step up to the plate
and compensate for what they don't know.
If the job had blown up as it appeared it was about to,
too many people here would blame the printer. If true, it's irrelevant. Bad
work reflects badly on everybody associated with it, including the client.
Dan Margulis
________________________________________________________________________
Date: Tue, 08 Jun 2004 14:34:25 -0700
From: Peter Figen
Subject: Re: Re: CMYK profiles
Dan Margulis wrote:
For those photographers reading the above, let me
suggest that you put
yourself in the client's position. She certainly
understands that the proofs are
garbage. Whether she can understand the above
explanation as to why is unclear.
The printer's explanation will be a lot simpler,
namely, that Peter's work was
no good. It's exceedingly difficult for clients to
figure out who is really to
blame in a situation like this. It is, however, easier
to fire a photographer
than a printer.
The printer didn't actually offer any explanation as to
why their first round of proofs looked so off, whether from their scans or
mine. The sculptor did understand a simplified explanation of what went
wrong.
I suspect that the list is tipping its hat to Peter for
going so far above
and beyond the call of duty. May I respectfully suggest
that sadder but more
realistic view is that, if he had not been able to do
these things, Peter might
very well have taken the fall for the failure of this
job.
I agree with what you're saying. I began to see,
several years ago, that if I was going to be responsible for color for my
clients that I was going to have to employ every tool I could reasonably
afford and learn as much as I could about the entire process. This has only
been made more difficult in the last few years as more printers adopt DTP
workflows that are not maintained to any color printing standard other than
their own. Having to use close to twenty or so different vendors for
printing services for different clients, not a single of these vendors,
with their new digital proofing systems, match each other, and only one,
Modern Postcard, appears to print to a standard represented by one of the
standard Adobe profiles. Needless to say, this variety of conditions make
it almost impossible for anyone to hit perfect color on their first round
of proofing without employing every available tool.
If the job had blown up as it appeared it was about to,
too many people here
would blame the printer. If true, it's irrelevant. Bad
work reflects badly on
everybody associated with it, including the client.
This is exactly why I stepped in, the printer didn't
appear capable of fixing the mess and both my work, and the work of the
sculptor, represented in the photos, were going to be done a severe
disservice - one which would be very lasting.
Peter Figen
________________________________________________________________________
Date: Tue, 8 Jun 2004 10:40:56 -0700
From: Dennis Dunbar
Subject: Re: Re: CMYK Profiles
On Jun 8, 2004, at 2:51 AM, Dan Margulis wrote:
You should get rid of this client. They have an
obligation to be better
informed than this. I suspect that they did this on
purpose, to sabotage your work.
It is a real disgrace how clients have failed to keep
up with the times and
how anti-progress they are. If you dump this client,
maybe others will wake up
and we can finally get somewhere in the graphics
industry.
Dan,
Thanks for the really good laugh!
You're close in the way you make the contrast between
what we will tolerate from a client and what we'll tolerate from a printer.
But I think it would be a great mistake to forget that clients are who we
are working to serve, and printers, being in the position of being a
vendor, should be working to support that service.
This is not such a subtle difference in the
relationship and therefore the difference in what we will tolerate is
understandable and even justified to a large extent. In providing a service
to a client we should know more than they about what we are doing. And the
printer, in providing a service to us as well as our client, should
similarly be at least as skilled as we are.
Unfortunately this is often not the case. And there
comes the problem.
My point was to say too many times we cannot prepare a
file for a specific press condition. In my experience things change too
often for that to work the majority of the time.
The idea of not embedding a profile in a CMYK file
makes the most sense to me when the file has been specifically prepared for
the anticipated press. In those cases the "Just print the file!"
approach makes sense. When you have to guess at average conditions I think
a strong case can be made for embedding the profile as it is very
foreseeable that the file will need some adjustment before going to the
final printing.
I also feel that your final point "If you have a
good generic CMYK file, it will print well anywhere." works only if we
know what a good generic CMYK space is. It would be most helpful if you
could share with us, (and if the list could share with us), examples of
what you feel would work as a "generic CMYK" space. Do you have
any recommendations for specific profiles or CMYK setups that we should be
using as this generic space?
Dennis Dunbar
APA Digital Chair
________________________________________________________________________
Date: Tue, 8 Jun 2004 12:11:09 -0400
From: "Annette Murray"
Subject: RE: Re: CMYK profiles
Dan and many others write:
"good generic CMYK file, it will print well
anywhere."
This reference to a generic CMYK file could be
dangerous. Supplied with Photoshop is a CMYK profile named: Generic CMYK
Profile. Do you agree that Generic CMYK Profile creates a less than
desirable separation?
I am afraid that inexperienced Photoshop users reading
these posts may think this "Generic CMYK Profile" would be a good
choice to use when converting from RGB to CMYK because of all the
references to "generic CMYK file".
In fact--for whatever reason--many inexperienced
Photoshop users tend to supply images that have been converted using the
Generic CMYK Profile.
I think they have ColorSync Workflow selected in
Photoshop Color Settings.
What this all boils down to is:
Do you think we need to find a another term for
"good generic CMYK file". Any suggestions?
Annette Murray
Prepress and Color Consultant
ANRO Inc.
222 Lancaster Ave.
Devon, PA 19333
Website: ANRO.com
800.355.2676 x241
________________________________________________________________________
Date: Tue, 08 Jun 2004 18:24:06 -0600
From: Andrew Rodney
Subject: Re: Re: Re: CMYK Profiles
on 6/8/04 11:40 AM, Dennis Dunbar wrote:
I also feel that your final point "If you have a
good generic CMYK
file, it will print well anywhere." works only if
we know what a good
generic CMYK space is.
And I have ocean front property in New Mexico I1d like
to have someone make an offer on. <G>
Let1s see, if all the printers who actually tell us
they print to SWOP actually conformed to TR001 SWOP, then the Adobe
supplied U.S. Web Coated (SWOP) v2 would be a killer generic profile. Or
TR004 (when finalized) for Sheetfed. Or course, it1s my opinion that the
saying 3we print SWOP2 heard all across this country is one of the three
biggest lies in the world. Great joke in which the rest is too dirty to
tell on this fine list.
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Wed, 09 Jun 2004 12:06:57 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
Dennis Dunbar wrote:
In providing a service to a client we
should know more than they about what we are doing. And
the printer, in
providing a service to us as well as our client, should
similarly be at
least as skilled as we are.
Hi Dennis, yes this is a problem.
I am not really sure how things are in the trade
outside of my local conditions (Sydney) - but here, things have really
imploded since the early to mid 1990's.
I visited my old trade school where I studied as an
apprentice (in addition to my on the job training), but this time I was
teaching a night Photoshop class - when talking to the full time teachers
who also taught nights, I was depressed to hear how smaller the student
classes are compared to the 'good old days' of phototypesetting and
paste-up, repro cameras etc when I first learnt the trade.
Fewer apprentices is not a good thing.
When typesetting died as a trade, many of the operators
moved in graphic reproduction or into desktop based art production,
leveraging their existing prepress skills and knowledge. But the loss of
the many type shops that were around was a bad thing, less employers means
less apprentices and less opportunities for existing tradesman to work.
Graphic reproduction as a service has pretty much gone
the way of typesetting services, there are few colour houses when compared
to days gone by. Again, less chances for training and employment of key
personnel.
So where have all the skilled prepress operators gone?
They are probably 'in house' somewhere, either at a art
studio, printer or other service provider...or perhaps they are more happy
doing other things and have left the trade.
For some printers (they will remain nameless, but I am
using one real life printer as an example here), they may now have prepress
in- house, but they may not give it as much consideration as their more
expensive press...where as in the old days, you had a separate colour house
which did a very good job...today the printer probably has prepress
in-house and is making a loss on it in order to pull in the print work so
some things may go on that would not in the old scheme.
Stephen Marsh.
________________________________________________________________________
Date: Wed, 9 Jun 2004 10:03:22 -0600
From: Chris Murphy
Subject: Re: Re: CMYK profiles
On Jun 8, 2004, at 10:11 AM, Annette Murray wrote:
This reference to a generic CMYK file could be
dangerous. Supplied
with Photoshop is a CMYK profile named: Generic CMYK
Profile. Do you
agree that Generic CMYK Profile creates a less than
desirable
separation?
Actually it is not supplied with Photoshop. It's
supplied with Mac OS. On Mac OS 9 this profile would be lethal unless your
final destination is an Apple Color Laserwriter, and even then it may be of
little use. On Mac OS X, an identically named profile is based on a subset
of TR 001 data. It's fine for previewing CMYK files, but I would not
consider it high quality for making separations.
And it is a fair point to note the distinction between
"good generic CMYK file" and use of the "Generic CMYK"
profile supplied by Apple.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Wed, 9 Jun 2004 10:09:24 -0700 (PDT)
From: Kathleen Helgert
Subject: cmyk consensus
If the consensus has been reached that providing
generic CMYK is the way to go, I'm curious as to why the output destination
isn't known.
1. Is it going to a web press, sheetfed, or desktop
publishing?
2. Is it being printed on newsprint, coated or uncoated
stock?
3. Any ideas about dot gain or total ink coverage?
4. Printed with stochastic or conventional line screen?
Wouldn't all of these things affect the the color
and/or the ability of the color to be moved on press?
I'm just in awe that the concensus seems to be convert
to cmyk without knowing the above. Just my humble opinion. Seems to me that
stochastic isn't able to be moved on the printing press like conventional
line screens, ergo, the "light" cmyk files are going to print
light.
In regards to the photographer that supplied both the
digital file and the transparency to the printer, IMHO, you covered your
butt when you supplied the transparency and let them do the scan.
I'd be interested in hearing opinions on the rgb
colormatch workspace.
And I have to agree, this group is very educational.
Any chance that Dan Margulis would share with us what
his color settings are in Photoshop?
________________________________________________________________________
Date: Wed, 9 Jun 2004 09:14:52 EDT
From: Dan Margulis
Subject: Re: CMYK Profiles
Dennis writes,
This is not such a subtle difference in the
relationship and therefore
the difference in what we will tolerate is
understandable and even
justified to a large extent. In providing a service to
a client we
should know more than they about what we are doing. And
the printer, in
providing a service to us as well as our client, should
similarly be at
least as skilled as we are.
Woulda, shoulda, coulda is irrelevant here. For better
or for worse you and the client and the printer are all partners. If the
client and the printer are fairly knowledgeable, that's the for better
part. If neither one knows much about prepress that's for worse--but
they're both your partners whether you like it or not.
For the time being, anybody who owns a camera can claim
to be a top professional photographer if he wants to, and anybody who owns
a press can call himself a high-quality printer. When I'm Dictator of the
World, there are going to be strict licensing requirements, and anybody
trying to fob himself off as either one of those things without
demonstrating a requisite level of competence will be sentenced to five
years in a cell with a color management vendor.
Until that time, if a good printer is forced to work
with an incompetent photographer, or the other way around, or if either one
is forced to work with an incompetent client, my suggestion is that there
should be less complaining and more realization that all these people who
have not learned what they should have left a major marketing opportunity
for those that have.
When you have to guess at average conditions I think a
strong case can
be made for embedding the profile as it is very
foreseeable that the
file will need some adjustment before going to the
final printing.
This is basically saying that we all should work hard
to make sure that Al Gore gets elected in 2000. There may have been some
merit to the position at one time but there comes a point where we have to
just give up the argument and move on.
I also feel that your final point "If you have a
good generic CMYK
file, it will print well anywhere." works only if
we know what a good
generic CMYK space is. It would be most helpful if you
could share with
us, (and if the list could share with us), examples of
what you feel
would work as a "generic CMYK" space. Do you
have any recommendations
for specific profiles or CMYK setups that we should be
using as this
generic space?
The traditional Photoshop sep method works well
provided it is changed to Light GCR or UCR, and not more than 85% black.
Dot gain should be set between 17% if you feel lucky and 20% if you don't.
If you wish to be extra careful, once you have finished the work and have
what you think is a good CMYK file, you can Image>Mode>Assign
Profile>SWOP v.2 to see whether you can live with that look. Remember,
if you don't know the exact printing condition, fine-tuning is of limited
utility. What you're looking for is a file that will produce *acceptable*
results under a variety of circumstances.
Dan Margulis
P.S. In reply to Annette's point, no, I am not
referring to the "Generic CMYK" profile, which is best avoided.
________________________________________________________________________
Date: Wed, 9 Jun 2004 22:46:08 -0400
From: Terry Wyse
Subject: Re: Re: CMYK profiles
On Jun 7, 2004, at 5:12 PM, Dan Margulis wrote:
There *is* a consensus, accepted by all but the most
wild-eyed extremists.
You've given three options. #2, the embedded CMYK
profile approach, is obsolete.
It died several years ago and the official funeral
service was Photoshop CS.
Silly.
If you have a good generic CMYK
file, it will print well anywhere. If you don't have a
good file, all the
knowledge of output condition in the world won't help.
I can't believe you would even say this. Please,
what/where is this "Generic CMYK" profile that magicially prints
anywhere? On what output device? What press? In what country?
Terry
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
________________________________________________________________________
Date: Wed, 9 Jun 2004 22:46:16 -0400
From: Terry Wyse
Subject: Re: Re: CMYK profiles
On Jun 8, 2004, at 10:48 AM, Dan Margulis wrote:
The Photoshop supplied CMYK profiles are not good IMHO
for dealing with an
unknown printer. The web profile may give a sep that's
appropriately dark for a
quality-oriented web printer, but it will be too muddy
for a shop that isn't
under good control. Worse, the black is too heavy. With
unknown printers, black
density is often very problematic. Also, one would like
to leave open the
possibility that the printer can increase the black for
aesthetic reasons,
particularly if it gets printed sheetfed. With a
traditional Light GCR setting, if
the printer feels he can punch up the black, the
picture will look better for
it. With one of the built-in profiles, it won't
work--it will louse up too many
colored areas.
So...then...if at least one of the profiles shipped
with Photoshop is NOT the magic generic CMYK profile, then where can we
find it?
Terry
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
________________________________________________________________________
Date: Wed, 9 Jun 2004 22:04:54 -0400
From: John Castronovo
Subject: Re: Re: CMYK Profiles
As a matter of practicality, I have to agree with you
Dan. Placing blame isn't where it's at, but we do have a financial
situation to resolve if the job needs to be fixed and re-run, and when
money's been lost as the result of people not communicating or just being
asleep at the wheel, the question of "who is at fault" does come
to mind.
To be sure, there are many more incompetent clients
than photographers or printers, and no, we don't want to fire them all.
However, I believe it's the printer who must be the one to make sure that
everything has been properly prepped for his workflow. It would solve a lot
if more printers would only ask questions and kick jobs back to their
clients instead of just running them as they're submitted and blaming the
client if it doesn't work out.
IMHO
john c.
________________________________________________________________________
Date: Thu, 10 Jun 2004 07:09:57 -0400
From:John Romano
Subject: Re: Re: CMYK profiles
Dan
Where is it in CS that says CMYK Embedded profiles are
obsolete ? I don't think I got that version !!
You've given three options. #2, the embedded CMYK
profile approach, is obsolete.
It died several years ago and the official funeral
service was Photoshop CS.
This statement is ridiculous coming from a list
moderator.
Also in CS under color settings for US Prepress
defaults why would the Preserve Embedded CMYK profiles be an option ?
And there again in the save dialog..... save with
embedded profile ?
Just another Adobe mistake ?
John Romano
________________________________________________________________________
Date: Thu, 10 Jun 2004 12:23:42 -0700
From: Dennis Dunbar
Subject: Re: CMYK Profiles
On Jun 10, 2004, Dan wrote:
For better or for worse you and the client and the
printer are all
partners. If the client and the printer are fairly
knowledgeable,
that's the for better part. If neither one knows much
about prepress that's for worse--but they're both your
partners
whether you like it or not.
You are right with this point. And I very much agree,
there is not much point, future, or profit in finger pointing.
This is basically saying that we all should work hard
to make sure
that Al Gore gets elected in 2000. There may have been
some merit to the
position at onetime but there comes a point where we
have to just give up the
argument and
move on.
Dan, just because you feel this is a dead argument does
not mean it is. I read frequently of your insistence on these points only
to see them come up again and again. Is there a vast
"Calibrationist" conspiracy or is there perhaps a significant
portion of the industry that feels profiles in CMYK files may be very
useful in certain circumstances?
The traditional Photoshop sep method works well
provided it is changed to
Light GCR or UCR, and not more than 85% black. Dot gain
should be set between 17% if you feel lucky and 20% if you don't. If you
wish to be extra careful, once
you have finished the work and have what you think is a
good CMYK file, you
can Image>Mode>Assign Profile>SWOP v.2 to see
whether you can live with that
look. Remember, if you don't know the exact printing
condition, fine-tuning is
of limited utility. What you're looking for is a file
that will produce
*acceptable* results under a variety of circumstances.
Here is what I was looking for. So to be sure I am
clear your recommendation is to use the "SWOP (Coated), 20%, GCR,
Medium" setup in PS, but change the GCR to light, instead of medium,
change the black to no more than 85%, and perhaps adjust the dot gain to
17%. Then do a preview using the SWOP v2 profile just to see how it may
look in order to check for any predictable disasters.
Has anyone else used this method? Do the others on the
list agree that this should be a workable CMYK setup when the specifics of
the final conditions are unknown?
Thanks!
Dennis Dunbar
________________________________________________________________________
Date: Thu, 10 Jun 2004 17:01:39 -0400
From:John Romano
Subject: Re: Re: CMYK Profiles
Glad there is someone else who feel this way, thanks
Dennis.
I don't really know what is so bad about being a
"Calibrationist ".Haven't heard this term in along time, actually
I think it originated here......Dan !
What could be so bad about having all of your devices
calibrated to a standard ?
Here is what I was looking for. So to be sure I am
clear your
recommendation is to use the "SWOP (Coated), 20%,
GCR, Medium" setup in
PS, but change the GCR to light, instead of medium,
change the black to
no more than 85%, and perhaps adjust the dot gain to
17%. Then do a
preview using the SWOP v2 profile just to see how it
may look in order
to check for any predictable disasters.
Who has time to use this in a production
environment ?
The time it takes you to do this you could get the
right info from the printer or supply the image converted to SWOPv2 profile
with it Embedded.
Regards
John Romano
________________________________________________________________________
Date: Thu, 10 Jun 2004 14:57:50 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
There is some funky behavior going on in this 3classic2
CMYK engine.
If I take 0/0/0 and convert to US Coated SWOP v2, the
conversions shows atotal ink of 300% which is what I thought was the
correct value of thatprofile. If I go into Custom, ask for 300% and do the
same conversion, I get266%. If I now alter that custom setting, leave the
value at 300% total inkbut lower the Black Limit to 50 I get a total ink
limit of 278 (higheralthough visually the black looks lighter and the K
value is less).
What1s that all about?
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Thu, 10 Jun 2004 17:07:43 -0500
From: "drhobbes"
Subject: CMYK profiles
Because all of my printing is done on
wide format inkjet printers, I'm not sure how CMYK profiles affect offset
printing.
Suppose I take duplicate images, convert
one to a CMYK profile with 40% dot gain and convert the second to the same
CMYK profile with a 12% dot gain, then convert them back to RGB for
printing on an inkjet.
Both images look the same in RGB, and both print the
same on the inkjet printer.
Now, what result can I expect from the
printed output if these RGB files are sent to a commercial printer for
printing on a sheetfed press? What result would I get if I sent the
original CMYK files to the printer, one with the 40% dot gain and the other
with the 12% dot gain setting. I expect that if the press showed a
dot gain of 17%, one of my CMYK files would print too light and the other
too dark. Does the RGB file contain some kind of information that
alerts the imagesetter that the CMYK versions had different dot gain
values? In other words would the RGB files prepared from these CMYK
files produce the same output on press as their CMYK counterparts did?
Now suppose I sent the two files in two
separate shipments, one in which the files were tagged with the profiles
and the other in which the files were not tagged? What would happen
to the output if all went well (that is, if the printer knew what he was
doing)? What could happen to the printed output if the printer took
the tagged image files and handled them the wrong way? For that
matter, what exactly is the wrong way? What happens when a tagged
file falls into the hands of someone who doesn't know what to do with it?
And the last question: When a
printer receives an untagged file, does he tag it with a profile of his own
choosing or does he just go ahead and send it to the imagesetter? All
the comments about sending in untagged image files makes me wonder why we
bother to use CMYK profiles in the first place, other than because they're
always lurking there when we convert to CMYK. Does the image file
retain the profile information if we strip off the tag, or if we don't even
tag the file? Presumably so, but I don't know and I would like to
find out.
Thanks for any assistance you folks can
offer. Many of you have made a significant contribution to my
knowledge of Photoshop and printing in general.
Howard Smith
________________________________________________________________________
Date: Thu, 10 Jun 2004 18:38:50 -0400
From: Dan Margulis
Subject: Re: Re: CMYK Profiles
Andrew Rodney writes,
If I take 0/0/0 and convert to US Coated SWOP v2, the
conversions shows a total ink of 300% which is what I thought was the
correct value of that profile. If I go into Custom, ask for 300% and do the
same conversion, I get 266%. If I now alter that custom setting, leave the
value at 300% total ink but lower the Black Limit to 50 I get a total ink
limit of 278 (higher although visually the black looks lighter and the K
value is less). What's that all about?
You are using the default of Medium GCR. In comparison
to a traditional separation this gives considerably more black and less
CMY. The CMY maxes out at about 60c50m50y and it is therefore impossible to
achieve 300%. The theory behind this is that with so much detail being
carried in the black channel, adding weight to the CMY would plug the
shadows. If you nevertheless want to do it, you would put in a UCA value.
This would allow 300% with a Medium GCR setting.
With Light GCR or UCR, which are more like traditional
separations, there isn't any such problem because the typical max CMY value
is 80c70m70y.
Dan Margulis
________________________________________________________________________
Date: Thu, 10 Jun 2004 17:36:55 -0700
From: Peter Figen
Subject: Re: Re: CMYK Profiles
Andrew,
This happens when you have Med. black generation
selected. Change it to Light and it works like it should. When I set it to
Light and 90% black, I get 298% total ink. From there you can bump the
total ink up to about 305 if you need to hit an even 300% in the final.
Even when making custom ICC profiles, you sometimes have to enter a
slightly higher total ink value in order to hit your optimal goal. Part of
that is due to the relationships between rendering intents and BPC, and
part is because you almost never have a 0,0,0 actually in the file.
Peter Figen
________________________________________________________________________
Date: Thu, 10 Jun 2004 19:17:04 -0600
From: Chris Murphy
Subject: Re: Re: CMYK profiles
On Jun 10, 2004, at 5:09 AM,John Romano wrote:
Where is it in CS that says CMYK Embedded profiles are
obsolete ? I don't
think I got that version !!
I don't know why it's CS, because the default handling
of CMYK images with embedded profiles is the same since Photoshop 6, which
is to *ignore and strip* embedded CMYK profiles unless they match the
working space. The concept of embedded profiles in CMYK images never got
off the ground really because a.) repurposing was never intelligent enough,
which is necessary to take into account the uniqueness of CMYK images;
color appearance is not the only important information in a tagged CMYK
file. Unique channel content is equally important. b.) Adobe went from one
extreme in Photoshop 5 (always use embedded profiles) to not only ignoring
them but stripping them in subsequent versions by default.
Also in CS under color settings for US Prepress
defaults why would the
Preserve Embedded CMYK profiles be an option ?
And there again in the save dialog..... save with
embedded profile ?
Just another Adobe mistake ?
Yes but not by default. The default behavior of
Photoshop 6, 7, and 8 is very unfriendly to embedded profiles in CMYK
images. And an even worse mistake, which made embedding profiles ipso facto
potentially lethal, is how easy Adobe made it for regular Joe User, and
purported experts with no prepress or printing experience to recommend, to
automatically convert CMYK images unintelligently upon open. This is the
Convert to Working CMYK policy. At best it should be removed and stuck into
the Automate portion of Photoshop. At worst it should be invisible as a
non-option unless the Advanced checkbox is selected.
There are entirely too many people out there who trash
unique aspects of a CMYK file by using this policy. It even happens at
prepress companies where someone used the feature once for a good reason,
left it that way, and without warning a bunch of images would get converted
by a subsequent operator who had no reason to suspect the files he was
working on were being trashed. Live and learn? Of course, but the lack of
intelligent conversion of CMYK, and ensuring null transforms occur when
they should is not currently integrated into any of these products. Yet the
products are in the hands of both professionals and amateurs, and the
options are tricycle mode and space shuttle mode with nothing in between.
And in neither case is a tricycle or space shuttle a good mode of
transportation if all you need to do is get from Denver to Chicago, a need
thousands of people perform every day.
So while there is inherent value in embedding CMYK
profiles, and it should be required, not optional, is that applications and
users too frequently unknowingly collude to trash CMYK images. It's not
BECAUSE a profile is embedded, it's because the applications (and users)
aren't knowledgeable or capable enough. But the easiest work around in the
meantime, even if it is not ideal, is to not embed the profile. That it can
no longer be as easily repurposed next week or next month or next year is a
problem, but less of a problem than the current job 2 days past deadline
getting trashed.
It's a non-ideal solution to a non-ideal problem.
Anecdotes aside as to their usefulness in specific situations, it will not
get any better until the handling of CMYK images with profiles becomes much
safer, and more compelling; and then after that embedding of CMYK profiles
and preserving them EVEN if they need to be ignored becomes default
behavior.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Thu, 10 Jun 2004 21:36:15 -0400
From: Jim Rich
Subject: Re: Re: CMYK Profiles
On 6/10/04 4:57 PM, "Andrew Rodney" wrote:
If I take 0/0/0 and convert to US Coated SWOP v2, the
conversions shows a
total ink of 300% which is what I thought was the
correct value of that
profile. If I go into Custom, ask for 300% and do the
same conversion, I get
266%. If I now alter that custom setting, leave the
value at 300% total ink
but lower the Black Limit to 50 I get a total ink limit
of 278 (higher
although visually the black looks lighter and the K
value is less).
Andrew,
As I recall in the last two beta programs, Adobe
did not want to hear or address the issues with the classic cmyk options. I
reported that when asking for say 345% to 360% I would get an
actual 300%.. After I posted this to Adobe three times with no
answer I let it go.
In some cases the classic cmyk options seems to be
broken.
That1s what I have seen.
However,I don1t depend on the classic cmyk
options too much these days.
My .02
Jim Rich
________________________________________________________________________
Date: Thu, 10 Jun 2004 18:34:23 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/10/04 4:38 PM, Dan Margulis wrote:
You are using the default of Medium GCR. In comparison
to a traditional
separation this gives considerably more black and less
CMY. The CMY maxes out
at about 60c50m50y and it is therefore impossible to
achieve 300%.
Dan, there1s a field called Total Ink Limits. I pop a
value in there and I
don1t get that value. This is unlike every other
software I1ve used to
control total ink. I get what I ask for. Do you not see
something odd about
this?
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Thu, 10 Jun 2004 21:32:33 -0600
From: Chris Murphy
Subject: Re: Re: CMYK Profiles
On Jun 10, 2004, at 6:34 PM, Andrew Rodney wrote:
Dan, there’s a field called Total Ink Limits. I
pop a value in there and I
don’t get that value. This is unlike every other
software I’ve used to
control total ink. I get what I ask for. Do you not see
something odd
about this?
This continues to be a problem for even the top two
profile building applications. Monaco does a better job than GretagMacbeth
in this regard, but even their profiles start to fall apart at lower ink
limits, below about 260%. GretagMacbeth profiles, in both ProfileMaker 4
and 5, do not honor the ink limit settings, rather treating them as a
suggestion. It's a particular problem with respect to the black ink limit
which in ProfileMaker 5 is not honored at all in any rendering intent. When
I ask for 100%K I expect to get 100% K, not 93%. And I don't expect to get
89% just because I've asked for black point compensation, yet that's what I
get.
I don't know what the explanation for these problems
are, but the classic CMYK engine in Photoshop isn't the only one that lack
strict adherence of user selected options. Not that this makes Adobe's
problem OK, what I mean is that it's not OK for anyone to do it, but at the
same time it seems to be somewhat common.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Fri, 11 Jun 2004 10:00:27 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
I don't really know what is so bad about being a
"Calibrationist ".
Haven't heard this term in along time, actually I think
it originated
here......Dan ! What could be so bad about having all
of your devices calibrated to a
standard ?
Dennis, I think Dan uses the term for those who may
stop thinking or using common sense once they start looking at the world
through a spectro.
As with most things in life, it's all about a healthy
balance. <g>
Who has time to use this in a production
environment ?
The time it takes you to do this you could get the
right info from the printer
or supply the image converted to SWOPv2 profile
with it Embedded.
John, this takes little time and it can be quicker than
the alternative (providing a non suitable sep or hand editing a less than
ideal sep more than one should have to edit it).
One profile does not cut it for all seps, even for the
same condition.
And if one has to settle on a single profile, there are
pros/cons to deciding to run with a lower or a higher amount of GCR.
The case at hand, we do not know the printer or the
final conditions.
If the image is 'average' - then I would tag it
correctly.
If the image was not average (four colour grey) then I
would not tag it as I don't want someone messing with the GCR mix - if they
know enough to know that this does not work well for them, then they should
know how easy it is to make a new true neutral sep for their conditions off
this file, with or without any tag on it.
Stephen Marsh.
________________________________________________________________________
Date: Fri, 11 Jun 2004 09:42:24 -0000
From: Stephen Marsh
Subject: Re: CMYK Profiles
Dennis Dunbar wrote in reply to Dan Margulis:
Here is what I was looking for. So to be sure I am
clear your
recommendation is to use the "SWOP (Coated), 20%,
GCR, Medium" setup in
PS, but change the GCR to light, instead of medium,
change the black to
no more than 85%, and perhaps adjust the dot gain to
17%.
Has anyone else used this method?
Yes, prior to APS5.x - this was perhaps the most common
way to separate in Photoshop (unless you had sep tables). Sure many four
colour files were drum seps, but others were often Photoshop ones too.
There are many concerns with this method, but it generally does work as
intended and expected for the task at hand. There is probably more
widespread user use and live testing in jobs of the old sep method over ICC
profiles, considering how long the custom CMYK method has been in use.
The pro reasons for using legacy Custom CMYK have been
discussed in Dan's books and on this list in the past, a list archive can
be found here:
http:
//www.ledet.com/margulis/ACT_postings/SeparationIssues/ACT-
Ancient-Custom-Color.htm
Having easy access to flexible CMYK separation options
is a good thing for users who do more than simply convert every image to
one output profile.
It is very good for making a sep which is close to what
the aim-point may be in terms of specs for dot gain, black gen, K plate or
total ink limits etc. Often a little K plate curve work and or channel
mixing can do wonders here, so one may choose to separate slightly under
the delivery condition with the intent of massaging the sep into much
better shape.
It is also very good for making 'false separations'
where one needs to blend in data to other seps (say blending in a better
custom CMYK K plate data in darken mode into a Adobe v2 ICC CMYK profile
sep etc).
It is great for making four colour greyscales with
heavy GCR.
It is fantastic for doing seps of screen captures, maps
etc.
It is handy for doing many other things, the list could
go on and on.
Yes, if APS had ICC profile editing abilities then
users who need flexibility would not need to resort to the old system which
is scorned now that ICC profiles are the default. Adobe seem to wish to
only offer us an image editor (in essence) - while Binuscan have offered
ICC abilities in 'PhotoRetouchPro', perhaps due to the lack of said
features in APS.
Then do a
preview using the SWOP v2 profile just to see how it
may look in order
to check for any predictable disasters.
Are we going to strip the tag and assume the current
workspace? What is the workspace if this is so, how close to the sep is it?
Are we then going to tag this 'false profile' to the legacy separation?
If there were true LAB/RGB neutral numbers in the file
before conversion, the grey balance of the SWOP v2 profile is not going to
match that of the legacy sep. On press they may both be able to reach a
neutral - but not if going by the numbers in Photoshop and referencing the
numbers with the new profile. Further conversions out of CMYK would then
mean that there is no longer a neutral due to the 'false profile' which
does not accurately describe neutrals.
Yes, I have made heavy GCR four colour grey seps in the
old legacy engine...then assigned SWOP v2 or other target conditions...then
placed colour LAB samplers...read how far off neutral things were...and
made a quick channel mixer adjustment to alter the cyan and or possibly MY
as well to make a true LAB neutral so that both the numbers and the profile
are true to TR001 (easier than one may think).
Do the others on the list agree that this should be a
workable CMYK
setup when the specifics of the final conditions are
unknown?
Depends on the conditions and the image.
Some images I may just use the SWOP v2 profile from
Adobe and I would tag it as such if just handing off the image and not a
complete job.
Others may be legacy CMYK separated and tagged as such
if appropriate to the work-flow at hand.
Sometimes I may use the Chromix TR001 profiles, which
do offer different GCR and UCR profiles as well as breaking the SWOP 300%
total ink with 280% and 330% options as well. I have some issues with these
profiles, so they are not always used but it is good to have lots of choice
when separating (this can't be said enough, one profile and no profile
editing options is not good).
I would target to the region if known (Euro, US, Japan)
using the appropriate v2 or legacy profile if I did not have a good
'Cromalin' or other profile which described the same general conditions
etc.
Stephen Marsh.
________________________________________________________________________
Date: Fri, 11 Jun 2004 10:17:09 -0000
From: Stephen Marsh
Subject: Re: cmyk consensus
Kathleen Helgert wrote:
If the consensus has been reached
Kathleen, I serously doubt than ANY consensus will EVER
be reached on this list.
And that, is not a bad thing. Life is never this
simple, and neither is print.
1. Is it going to a web press, sheetfed, or desktop
publishing?
We may not know.
Sometimes the best answer is CMYK (as if that is the
whole answer).
In this case, the historical practice seems to be to
generally aim to 'SWOP' ink limits or TR001 for more accuracy on the colour
side of things than what SWOP gives us. If in Europe or Japan, a suitable
middle road guess would probably be made too.
If the printer is a newspaper or gravure or some other
non standard operation, they are probably used to making a more 'generic'
sep work, however they decide to do it.
2. Is it being printed on newsprint, coated or uncoated
stock?
When we don't know, coated is the assumption that is
average.
3. Any ideas about dot gain or total ink coverage?
Again, around '20%' in the classic Photoshop engine is
safe, if not a bit too high for some.
4. Printed with stochastic or conventional line screen?
None of this is known.
Not every image delivery situation is perfect.
Wouldn't all of these things affect the the color
and/or the ability of the color to be moved on press?
Yes, but expectations and jobs differ pleasing colour
can be what some are after and not colorimetrically correct colour.
I'm just in awe that the concensus seems to be convert
to cmyk without knowing the above. Just my humble
opinion. Seems to me that stochastic isn't able to be
moved on the printing press like conventional line
screens, ergo, the "light" cmyk files are
going to
print light.
When we are following delivery guidelines, one does
what one can.
Those in the know may also offer RGB, ICC tagged files
or separate profiles, file names or folders which indicate colour space,
read me files, file info, hard copy notes, contact details etc.
Sure, even if we get the general condition right - we
may deliver 300%
TIL when 280% is more appropriate, or we may have been
able to run with 330% or whatever. It is all a crapshoot when we are forced
to deliver
'CMYK' with no other info (or files in general, what
size is needed etc).
Stephen Marsh.
________________________________________________________________________
Date: Fri, 11 Jun 2004 06:01:02 -0700
(PDT)
From: Kathleen Helgert
Subject: color
Gray component replacement and undercolor removal are
various ways that seperations are made to adjust the amount of cyan,
magenta and yellow that will print versus black. Different commercial
printers have different theorys on which prints better on their paper/press
combination. At times to the point of arguing with each other. Which is
another reason why I don't understand why providing cmyk to a commercial
printer is the consensus.
Regardless of who made the RGB to CMYK conversion, a
color proof is supplied by the printer to the client showing what the color
will look like. This to me is the turning point. Either the client is going
to accept the color, tweak the color, or completely reject it and go to a
different printer.
Some of the stories I've read hear sound like the
photograper/designer never got that chance. Why is that? Do things really
get printed that no one signed off on the color?
I thought that proof was the also the commercial
printers safeguard that the client couldn't later complain about the color.
As long as the press sheet matched the proof...
Which makes me ask, do commercial printers out there
actually do a press run when the only thing agreed upon is a soft proof off
of someones monitor? How do you prove after the press run that the color
run was really the color on the monitor?
I appreciate reading everyones opinions. You help me
form my own opinions. Thank you.
________________________________________________________________________
Date: Fri, 11 Jun 2004 06:37:55 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/10/04 6:36 PM, Peter Figen wrote:
This happens when you have Med. black generation
selected. Change it to Light
and it works like it should.
No on considers that a broken engine? What about Black
Ink limits? That1s screwed up if we understand how total ink limits are
supposed to work with reduction of inks.
How can people defend using classic engine when it sure
seems hosed in a few obvious (and perhaps hidden) ways?
Even when making custom ICC profiles, you
sometimes have to enter a slightly higher total ink
value in order to hit your
optimal goal.
Sure, two or three values but 30 or 40?
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Fri, 11 Jun 2004 08:35:04 -0400
From: Dan Margulis
Subject: Re: Re: CMYK Profiles
Andrew Rodney writes,
Dan, there's a field called Total Ink Limits. I pop a
value in there and I
don't get that value. This is unlike every other
software I've used to
control total ink. I get what I ask for. Do you not see
something odd about
this?
No, it's working exactly as it should. There are many
fields in this dialog and they all interrelate. Here, as you can see if you
look at the curves box next to the black generation field, you have chosen
a separation style where the total ink in the shadow is typically 260% or
so. Under those circumstances a higher "total ink" setting isn't
relevant. Similarly, if you choose "None" black generation,
meaning a CMY-only sep, a total ink value of 350% does nothing, and if you
choose "Maximum" black generation, your maximum total ink is
going to be about 225% even if you set "total ink" higher.
If you really want to use the same relatively heavy
style of black generation AND also have a 300% shadow (generally not
recommended), you can easily configure the engine to do so with UCA. And if
you use the basic settings that I recommend, 300% happens by itself.
Dan Margulis
________________________________________________________________________
Date: Fri, 11 Jun 2004 11:35:32 -0400
From: Dan Remaley
Subject: RE: color
Having managed a high quality trade house for many
years, our experience taught us to use around 70% GCR. Our reasoning was
that we spent many hours color correcting an image - we didn't want the
pressman to decide how he wanted the color to look, or select an image that
wasn't as critical for color evaluation.
The other problem is the proof itself! Is it SWOP, SNAP
or some specific color space? What is the density, dot gain characteristics
of the proof? Is there anything to measure on the proof for reference?
(generally not)
Does the proof represent SWOP - go to <swop.org>
and download how to make a certified proof. Use the GATF proof comparator
as a standard reference for color, and output it on the proof - we had one
on every proof! There are SWOP certified 'soft proofers' that match a TR001
reference press sheet.
Dan Remaley/GATF
Process Control Manager
412.741.6860x450
________________________________________________________________________
Date: Fri, 11 Jun 2004 09:10:00 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/11/04 6:35 AM, Dan Margulis wrote:
No, it's working exactly as it should. There are many
fields in this dialog
and they all interrelate.
They do? I don1t see that. Total ink is a fixed value
I1m asking for and not getting. If the field says 300%, I don1t think
getting 266% in the Sep is correct is it?
In other products, when you alter the mix of CMY plus K
to limit an ink, the original field where you1ve specify total ink updates
on the fly. What products do you have that can build a CMYK output profile
other than Photoshop? The ProfileMaker Pro module works correctly in that
if you enter total ink in field and then alter CMY or K sliders up or down,
the field updates on the fly. If you ask for a higher amount of ink, the
limits go up which makes perfect sense.
Here, as you can see if you look at the curves box next
to the black
generation field, you have chosen a separation
style where the total ink in
the shadow is typically 260% or so.
The GUI of the curves change visually, I have no issue
with that. The value doesn1t change and that1s just wrong.
Under those circumstances a higher "total
ink" setting isn't relevant.
Similarly, if you choose "None" black
generation, meaning a CMY-only sep, a
total ink value of 350% does nothing, and if you choose
"Maximum" black
generation, your maximum total ink is going to be about
225% even if you set
"total ink" higher.
But total ink should (correct me if I1m wrong as I1m
sure you will) provide a value in the area you1re asking and in the final
Sep. I1m asking for total ink of 300%. If I lower or raise one or more inks
(depending on the software), that final mix should be what I ask for. What
should be happening here is one should enter a total ink value and get
that. If the use ink limiting (like black limit), the effect should either
be the total ink limit field should update or the software should do what
it needs to do to provide the total ink limit I specify.
There may indeed be ways to fudge the numbers to get
what you ultimately want but this engine is behaving in what still feel is
a very screwy way. Is there any other data entry field in Photoshop that
produces a different value than what you ask for besides here? Off hand I
can think of none.
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Fri, 11 Jun 2004 10:24:13 -0600
From: Andrew Rodney
Subject: Re: color
on 6/11/04 9:35 AM, Remaley, Dan wrote:
There are SWOP certified 'soft proofers' that match a
TR001 reference press
But with regard to setting UCR/GCR and the like, you
don1t see the effects on the proof as that1s a function of the ink mixture
on press, correct?
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Fri, 11 Jun 2004 09:45:46 -0700
From: Stephen Ray
Subject: Re: Re: CMYK Profiles
on 6/11/04 8:10 AM, Andrew Rodney wrote:
Is there any other data entry field in Photoshop that
produces a different
value than what you ask for besides here? Off hand I
can think of none.
Photoshop's Image Size can give incorrect results if
fields are entered by using a screwy sequence, too.
-Stephen Ray
________________________________________________________________________
Date: Fri, 11 Jun 2004 13:35:12 -0400
From: Dan Remaley
Subject: RE: color
If done correctly, there isn't any color change in the
proof, it looks exactly the same (GCR/NON-GCR) as the press sheet.
With digital proofs the black 'appearance' may be
somewhat different than the press sheet. The major problem with GCR use -
is control of the Black, it's the first down ink and hits all the other
blankets, creating more gain than the default 20% of photoshop. With a UCR
or a conventional scan, the black printer is 'ghosted' with few values in
the critical 40-50-60% region.
GCR is a full tone black, using 40-50-60% areas - where
the gain is most on press. Therefore, when using GCR the black gain needs
to compensated for (around 24%). with GCR you have changed the values of
Y-M-C away from 40-50-60% and they can't be changed at press, as easily.
These changes are seen as 'hue shifts' with UCR or traditional scans.
(blue/grays - reddish/grays - yello/grays)
TO ANYONE INTERESTED:
Send me your mailing address and I'll send you a
printed sample of the GCR testing I have done here at GATF on our ManRoland
rotoman 36" press. Dan
Dan Remaley/GATF
Process Control Manager
412.741.6860x450
<www.gain.net>
________________________________________________________________________
Date: Fri, 11 Jun 2004 18:24:55 -0000
From: Dan Margulis
Subject: Re: CMYK Profiles
Terry writes,
So...then...if at least one of the profiles shipped
with
Photoshop is NOT the magic generic CMYK profile, then
where
can we find it?
The question had already been answered the previous
day, message #8398, or if you are looking for something more extensive, the
entirety of Chapter 13 of Professional Photoshop Fourth Edition covers this
topic.
Dan Margulis
________________________________________________________________________
Date: Fri, 11 Jun 2004 18:40:47 -0000
From: Dan Margulis
Subject: Re: CMYK Profiles
Dennis writes,
Dan, just because you feel this is a dead argument does
not mean it is.
No, but it means I am not going to waste further time
discussing the merits of an obviously failed methodology. If others choose
to wallow in the past, that's their affair.
Here is what I was looking for. So to be sure I am
clear your
recommendation is to use the "SWOP (Coated), 20%,
GCR, Medium" setup in
PS, but change the GCR to light, instead of medium,
change the black to
no more than 85%, and perhaps adjust the dot gain to
17%. Then do a
preview using the SWOP v2 profile just to see how it
may look
in order to check for any predictable disasters.
Understanding that we are talking about a situation
where we don't know who the eventual printer is going to be or what their
level of competence is, yes.
Dan Margulis
________________________________________________________________________
Date: Fri, 11 Jun 2004 18:32:04 -0000
From: Dan Margulis
Subject: Re: CMYK profiles
John Romano writes,
Dan
Where is it in CS that says CMYK Embedded profiles are
obsolete ? I don't
think I got that version !!
The last of the many times that this question has been
answered on this list was messages #8219, 8221, 8224,
8232.
Dan Margulis
________________________________________________________________________
Date: Fri, 11 Jun 2004 18:55:10 -0000
From: "dmargulisnj" Dan Margulis
Subject: Re: CMYK Profiles
Andrew Rodney writes,
They do? I don1t see that. Total ink is a fixed value
I1m asking for and not
getting. If the field says 300%, I don1t think getting
266% in the Sep is
correct is it?
Yes, because you have preempted that field with a
different setting.
But total ink should (correct me if I1m wrong as I1m
sure you will) provide
a value in the area you1re asking and in the final Sep.
I1m asking for total
ink of 300%. If I lower or raise one or more inks
(depending on the
software), that final mix should be what I ask for.
What should be happening
here is one should enter a total ink value and get
that. If the use ink
limiting (like black limit), the effect should either
be the total ink limit
field should update or the software should do what it
needs to do to provide
the total ink limit I specify.
No, because this would defeat the purpose of the type
of separation you were asking for. Settings at high GCR values are
*supposed* to have low total inks in comparison to traditional
skeleton-black separations, because the whole idea is to transfer weight
from the CMY into the black.
If you, for example, choose Maximum GCR, you are
explicitly asking for the shadow to be 0c0m0y100k, total ink=100%. Allowing
a total ink setting to override that and introduce CMY would completely
change the character of the separation. Photoshop nevertheless permits you
to do that if you want to, by choosing the UCA option.
If you seriously want a high-GCR, high ink-limit sep,
it's easy to do. But unless you've indicated that you want UCA, higher GCR
implies lower total ink.
Dan Margulis
________________________________________________________________________
Date: Fri, 11 Jun 2004 14:58:00 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/11/04 12:55 PM, dmargulisnj wrote:
If you seriously want a high-GCR, high ink-limit sep,
it's easy to
do. But unless you've indicated that you want UCA,
higher GCR
implies lower total ink.
This isn1t behaving like any of the other products I
have that allow me to make a separation. If I ask for light or heavy GCR
and ask for a specific total ink limit, that1s exactly what I get. In all
other products, the mix of CMY with K which is all different based upon the
GCR settings but always honors the total ink I ask for. I don1t understand
why that wouldn1t be the case. If I want 300% total ink, telling the
separation to use more or less black in substitute for CMY shouldn1t alter
the total ink should it? I don1t see what one has to do with the other. Or
to put it another way, IF CMY and K substitution is going to alter
total ink, the bloody field should update to what it1s going to give me,
not stay fixed on what I asked for and then not give it to me.
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Fri, 11 Jun 2004 22:25:43 -0400
From: Terry Wyse
Subject: Re: Re: CMYK Profiles
on 6/11/04 12:55 PM, dmargulisnj wrote:
If you seriously want a high-GCR, high ink-limit sep,
it's easy to
do. But unless you've indicated that you want UCA,
higher GCR
implies lower total ink.
I just don't think that is the case. There's nothing
about GCR that implies higher or lower ink limits. You may PREFER to run a
higher amount of GCR when lower ink limits are required but, by the same
token, there's nothing to stop you from creating a "Maximum GCR"
("MaxK" in ProfileMaker) and set a total ink limit of 320-340%. I
do it all the time when I'm taking raw press measurement data and giving my
customer several variations of GCR/UCR and ink limits. I will typically
give them a "ShortK" (light GCR), "MediumK" (medium
GCR), "LongK" (heavy GCR) and a "MaxK" ICC profile and
then explain to them under what circumstances and image characteristics
they may choose one profile over the other. If the press can handle 320%
total ink, there's no compelling reason to reduce this amount just because
I'm using increasing amounts of GCR in the press profiles.
While I consider Photoshop a professional separation
tool, I don't consider it a professional CMYK profile creation tool for the
reasons Sir Rodney and others have explained.
Terry
_____________________________
WyseConsul
Color Management Consulting
704.843.0858
________________________________________________________________________
Date: Sat, 12 Jun 2004 09:21:24 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/11/04 8:25 PM, Terry Wyse wrote:
I just don't think that is the case. There's nothing
about GCR that
implies higher or lower ink limits.
Thank you! That1s my understanding but not being an
original CMYK kid, I was hoping my understanding was correct.
Andrew Rodney
Http://digitaldog.net/
________________________________________________________________________
Date: Sat, 12 Jun 2004 13:41:48 -0000
From: Dan Margulis
Subject: Re: CMYK Profiles
Terry writes,
I just don't think that is the case. There's nothing
about GCR
that implies higher or lower ink limits.
That's like saying that winter doesn't imply colder
temperatures than summer. The entire theory of GCR is to substitute smaller
amounts of black ink for larger amounts of CMY. It most certainly *does*
imply less ink coverage, except in the extreme highlights. Total ink in
shadows in high-GCR seps should be definitely lower than in skeleton-black
seps--sometimes by as much as 100 points.
You may PREFER to run a higher
amount of GCR when lower ink limits are required but,
by the same
token, there's nothing to stop you from creating a
"Maximum GCR"
("MaxK" in ProfileMaker) and set a total ink
limit of 320-340%.
Correct. There's nothing to stop you from doing this in
the Custom CMYK dialog box, either, as I explained to Andrew.
If the press can handle 320% total ink, there's
no compelling reason to reduce this amount just because
I'm using
increasing amounts of GCR in the press profiles.
Some of us consider retention of shadow detail a
compelling reason. In high-GCR seps, so much shadow detail has been
transferred into the black that artificially adding ink to the CMY will
hinder rather than help it. This effect is commented on several times in my
books, and in books and classes I show pictures (example for class
attendees: Animals) where the black contains so much detail that the image
is drastically improved by use of a lower total ink, using Selective
ColorBlacks to reduce CMY. A higher total ink is appropriate when you are
deliberately attempting to suppress shadow detail. The example I have used
in books previously is a photograph of a playing card, the Ace of Clubs,
where any detail in the black is undesirable.
Unless your clients are printing gravure, they'll be
better served if your GCR profiles have lower total ink in the shadows.
While I consider Photoshop a professional separation
tool, I don't
consider it a professional CMYK profile creation tool
for the reasons
Sir Rodney and others have explained.
Since the Custom CMYK dialog lets you do it *either*
way--the way you describe, or the way experienced people prefer--I cannot
imagine what the criticism is, unless of course by professional you mean
requiring the payment to others for delivery of a profile. Personally, I
think the absolute minimum standard for a "professional" CMYK
profile creation tool is that the profiles should be editable on the fly
within Photoshop. At the moment, there's only one method that meets that
minimum standard. If you can talk your friends at Adobe into adding others,
that would be a big service.
Dan Margulis
________________________________________________________________________
Date: Sat, 12 Jun 2004 13:28:21 -0400
From: John Castronovo
Subject: Re: Re: CMYK Profiles
This probably isn't within your day-to-day experience
Dan, but would you care to comment on your best guess for CMYK seps that
are destined to wind up being printed inkjet on translucent media
(simulated Duratrans). Since the image will be illuminated from behind,
wouldn't we want far more color in the shadows than would be customary for
press?
john c.
________________________________________________________________________
Date: Sat, 12 Jun 2004 11:19:51 -0700
From: Lee Varis
Subject: Re: Re: CMYK Profiles
Andrew & Terry,
re: TIL, your confusion results from the assumption
that the ink limit applies to areas that are black. As Dan has repeatedly
pointed out, black areas in images sometimes contain some texture which
would be plugged up if the sep allowed 300% coverage in the black area. Med
GCR seps handle this situation by reducing the amount of CMY under the
black ink! On the other hand, if you have a deep, saturated chocolate brown
the sep might allow a higher percentage of CMY (up to the total ink limit)
to get the appropriate color! Also, if too much ink goes under the black
you can get a kind of tonal reversal where the black ends up being not as
dark as it could be – surely you've seen this effect with inkjet
prints when the profile puts too much ink on a particular paper.
Just because other products allow you do do something
stupid with the separation doesn't mean that Photoshop is less professional
with the limits it places on custom separations.
Again, one of the reasons you use GCR in the first
place is to reduce the amount of ink used in neutral colored areas even if
more saturated, deep colored areas require more ink coverage.
regards,
Lee Varis
http://www.varis.com
888-964-0024
________________________________________________________________________
Date: Sat, 12 Jun 2004 14:01:33 -0600
From: Chris Murphy
Subject: Re: Re: CMYK Profiles
On Jun 11, 2004, at 2:58 PM, Andrew Rodney wrote:
This isn’t behaving like any of the other
products I have that allow me to
make a separation. If I ask for light or heavy GCR and
ask for a specific
total ink limit, that’s exactly what I get.
First, do they have a UCA option? Most have removed UCA
and compute it automatically based on the total ink limit and black
generation amount.
What Custom CMYK could use is a user interface feature
that doesn't essentially lie to the end user, like the current interface
does. If you set medium GCR with a 300% ink limit, then the software should
automatically insert 30% UCA in order to achieve the set ink limit. If the
user takes UCR down to 15%, then the software should automatically change
total ink limit to 285%.
Second, ProfileMaker 4 and 5 do not work the way you
describe, Andrew. When you set total ink limit and use Relative
Colorimetric rendering, you do not get what you asked for. It's been a
complaint ever since ProfileMaker 4 shipped. And in the most recently
release ProfileMaker 5 it's an even bigger problem in that the separation
settings are not honored for any of the rendering intents.
In the case of Monaco profiles, settings are honored if
you do NOT use black point compensation. If you do use black point
compensation the black ink limit is not honored. 95% black ink becomes 75%
black ink. Just because Black Point Compensation is called for. It's not a
subtle difference, and that's not the way it's supposed to work. We really
need to use black point compensation especially with images with a lot of
important shadow detail or it gets clipped. The quandary then is to either
a.) use Black Point Compensation and get 20% less black ink, but retain
shadow detail; or b.) don't use black point compensation, killing shadow
detail, but get the right amount of black ink.
In some cases a.) might be acceptable, in others b.)
might be acceptable, but in certainly a non-insignificant minority of
images neither will be acceptable.
In all other products, the mix
of CMY with K which is all different based upon the GCR
settings but always
honors the total ink I ask for.
What product? ProfileMaker 5 hasn't with any CMYK data
set I've given it. It's always off by at least a few percent. If it were
still getting me an equivalent L* but with lower ink, I'd live with that.
But it doesn't work that way. I'm losing up to 3 L*, so I have to go back
into Profilemaker and rebuild with a higher ink limit than I want, in order
to end up with the ink limit I want (and then they are still different
between perceptual and relative colorimetric, and black point compensation
takes 100%K and shifts it to 95% K.) Now in litho, perhaps 100%K vs 95% K
is not a big deal. But in large format printing, 100% K is always required.
I want everything to have 100%K before UCA takes effect. None of these
packages appear to have such explicit control as Adobe's Custom CMYK
(despite the misleading interface).
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Sat, 12 Jun 2004 14:23:38 -0600
From: Chris Murphy
Subject: Re: Re: CMYK Profiles
On Jun 12, 2004, at 7:41 AM, dmargulisnj wrote:
That's like saying that winter doesn't imply colder
temperatures
than summer. The entire theory of GCR is to substitute
smaller
amounts of black ink for larger amounts of CMY. It most
certainly
*does* imply less ink coverage, except in the extreme
highlights.
Yes. I think the assumption being made by some people
is that increasing GCR should not imply decreasing reproduction L*. I think
that's a matter of opinion if that is right or wrong to assume.
A good separation engine user interface would allow for
either a free floating total ink limit tied to GCR; or allow total ink
limit to be anchored to achieve a non-floating black point. A corresponding
automatic computation of UCA in Photoshop's UI would be nice regardless,
because as it's currently working it does lie about the nature of the
separations. Can you have Heavy GCR, 325% ink limit, 100% black ink limit,
and no UCA? No. Well the user interface implies it can, so that's a
problem.
But at least when I ask for 100%K I get it. And when I
get UCA set correctly I get a consistent total ink limit and black ink
limit with and without black point compensation. I have yet to track down a
3rd party profile that behaves as consistently.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Sat, 12 Jun 2004 16:05:18 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/12/04 12:19 PM, Lee Varis wrote:
TIL, your confusion results from the assumption that
the ink limit
applies to areas that are black.
That's not what I'm saying. It applies to all areas
(more or less).
Just because other products allow you do do something
stupid with the
separation doesn't mean that Photoshop is less
professional with the
limits it places on custom separations.
No, what Photoshop is doing is funky and undocumented.
How I decide to mix CMY and K and in what proportions throughout the tone
curve has (or should have) no bearing on getting a total ink limit you ask
for. I may be asking for less CMY to be replaced with K but I still get to
tell the separation how much total ink I'm willing to take when all are
going full blast.
If you type in 300% and you get 266%, something is
funky.
Andrew
________________________________________________________________________
Date: Sat, 12 Jun 2004 17:00:29 -0600
From: Chris Murphy
Subject: Re: Re: CMYK Profiles
On Jun 12, 2004, at 4:16 PM, Andrew Rodney wrote:
The last two profiles I built (for Mo) I asked for
300%. One give me 300% on
the money and the other 297% (which isn't right but not
266%). Both using
RelCol.
It definitely makes a difference on the measurement
data it receives and the particular separation settings you use. I've been
testing primarily with DTR 004 and TR 001 since those data sets are
considered reasonable behaviors for the press conditions they are meant to
describe. The difference I'm getting between the requested TAC, and K
limit, and what I get with Relative Colorimetric + Black Point Compensation
is on average 14 points TAC and 5 points in K. It's not OK, regardless of
whose fault it is.
But if you happen to have your profile handy and could
send it to me, I'd like to inspect it.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
________________________________________________________________________
Date: Sat, 12 Jun 2004 16:16:03 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/12/04 2:01 PM, Chris Murphy wrote:
Second, ProfileMaker 4 and 5 do not work the way you
describe, Andrew.
When you set total ink limit and use Relative
Colorimetric rendering,
you do not get what you asked for.
The last two profiles I built (for Mo) I asked for
300%. One give me 300% on the money and the other 297% (which isn't right
but not 266%). Both using RelCol.
Andrew
________________________________________________________________________
Date: Sat, 12 Jun 2004 16:17:49 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/12/04 7:41 AM, dmargulisnj wrote:
The entire theory of GCR is to substitute smaller
amounts of black ink for larger amounts of CMY. It most
certainly
*does* imply less ink coverage, except in the extreme
highlights.
Individually per ink, yes. In total, no. I agree there
is a most significant difference individually per ink (you can see this in
the color channels after conversion). But the total ink from a 0/0/0 RGB
source should read the total ink limit in Photoshop that you specify. It
does this with other applications which I feel behave correctly and
Photoshop is not. That1s why I asked what else you1ve used to produce
similar conversions and if they behave as Photoshop or as I think is
correct.
Personally, I think the absolute minimum standard for a
"professional" CMYK profile creation tool is
that the profiles
should be editable on the fly within Photoshop. At the
moment,
there's only one method that meets that minimum
standard. If
you can talk your friends at Adobe into adding others,
that would
be a big service.
On this point we are in total agreement. The tool
supplied is however if anything, funky.
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Sat, 12 Jun 2004 21:53:51 -0400
From: Terry Wyse
Subject: Re: Re: CMYK Profiles
"In general" ProfileMaker will give you the
total ink you ask for (+/- 5%) IF you're dealing with relatively
well-behaved devices like an offset press. But try this with an inkjet
device and you're liable to get WILDLY different total ink settings than
what you ask for, sometimes by as much as 20-40% (usually lower). When this
happens, it can sometimes point to either a poorly linearized/limited
inkjet or a bad TIL request in the first place. I should note that I'm
using ColorShop X to view the neutral gray curves and total ink amount.
So you're both right. :-)
Terry
________________________________________________________________________
Date: Sun, 13 Jun 2004 10:12:06 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
on 6/12/04 7:53 PM, Terry Wyse wrote:
But try this with an inkjet device and you're
liable to
get WILDLY different total ink settings than what you
ask for,
sometimes by as much as 20-40% (usually lower).
Ah, interesting. I don1t know I1ve ever built a CMYK
profile to an inkjet so that1s an issue I wasn1t aware of.
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Mon, 14 Jun 2004 12:43:29 -0400
From: Henry Davis
Subject: Re: Re: CMYK Profiles
I think that maybe a root of the confusion can be found
in the notion that many RGB users haven't had a historical concern for the
need to save ink - $$$.
Henry Davis
________________________________________________________________________
Date: Mon, 14 Jun 2004 17:50:51 -0400
From: Jim Rich
Subject: Re: Re: CMYK Profiles
As a rule of thumb and in 99.99% of cases, end-user
don1t use UCR or GCR to save ink for inkjet printing. GCR is used to set up
conditions to make better prints and if there is some ink savings (which I
am skeptical of) then the end-user gets a very small bonus.
And in the case of commerical printing ( gravure for
example), I have only ever seen one printer save money by using GCR and
that was using a special program they developed to take the final (CMYK)
images and apply GCR before engraving. They do save substantial sums of
money every year. But they don1t use GCR options that are built into
Photoshop or ProfileMaker to save money on ink costs.
Jim Rich
________________________________________________________________________
Date: Mon, 14 Jun 2004 20:19:20 -0400
From: Henry Davis
Subject: Re: Re: CMYK Profiles
Jim,
Could you explain a little more about how UCR or GCR is
used to make conditions that make for better prints made on inkjets?
I have heard that such devices are very stable, and unlike a press,
don't have an operator working to maintain neutrals and such. With a
stable, grey balanced and profiled inkjet, it would seem that the
replacement or removal amount would be very small, almost nill for an
inkjet, even with imagery that may have a fair amount of neutrals.
Black would hold off until very near the shadow end. Please
correct me if I'm off base here. Is this why there might be numerous
profiles for a single printer/stock combination, all of them right, but one
being better because of image content? And, if it is a CYMK file that
is to be converted using a different profile, doesn't this make the ensuing
GRC/UCR scheme real squirrely?
I agree with your point that many inkjet users are not
concerned with saving ink, which, I think, was the same point that I had
made. I suspect that a fair number of such users are working in RGB,
and relying on a profile to make the conversion, not understanding very
much about GRC/UCR. It is still my belief, though, that UCR may help
to save ink, and that GCR may help to save even more. As inkjet, and
photographic print makers are less concerned with setoff and drying time,
it was my point that an RGB workflow was a possible root of the confusion
that Andrew had mentioned with regard to RGB centric users.
This thread has on occasion, had some very interesting
observations on the reasoning behind and the need for GCR/UCR. That
there may be more art involved in profiling, making basic decisions such as
how black is to be used, leaves me thinking that we have not gotten it down
to a push button solution just yet.
Lee Varis offered a simple, yet interesting observation
that I believe might be even profound:
"Again, one of the reasons you use GCR in the
first place is to reduce
the amount of ink used in neutral colored areas even if
more saturated,
deep colored areas require more ink coverage."
Henry Davis
________________________________________________________________________
Date: Tue, 15 Jun 2004 09:11:58 -0400
From: Jim Rich
Subject: Re: Re: CMYK Profiles
Henry,
I find that when you use the maximum amount of GCR you
remove most if not all of the 3 colors from the 3 color gray areas and
replace it with black while maintaining a neutral gray look. I find this
makes nicer looking neutral grays on inkjets for example.
So when there is a gray image area and the
maximum amount of GCR is used there is less chance the neutral
grays will have less color shifts. But this is only a rule of thumb. It of
course depends on things like paper, how well made your profile is and the
rip.
As for saving ink with UCR. Sure it might?? But
it is only theory and there are no numbers on this anywhere I have
seen.
An analogy for this issue is keeping
your gas tank full in the winter in snowy conditions. You do it so you
don1t run out of gas and become stranded, but the secondary benefit is that
it gives the car more weight for better traction. So why do you fill
your car up with gas?
In the inkjet context; why do you use UCR/GCR? To
save ink? Nope!
You use UCR/GCR in some situations to make good looking
prints.
Jim Rich
________________________________________________________________________
Date: Tue, 15 Jun 2004 06:44:09 -0600
From: Andrew Rodney
Subject: Re: Re: CMYK Profiles
GCR affects more than just neutrals (UCR is pretty much
just affecting neutrals).
Andrew Rodney
http://digitaldog.net/
________________________________________________________________________
Date: Tue, 15 Jun 2004 17:10:34 -0400
From: Henry
Subject: Re: Re: CMYK Profiles
Jim,
Thanks for the reply. The car is probably most
efficient when it is motoring with only just enough gas to reach the gas
station that is nearest the destination. The same goes for flying, the
plane hardly carries a full load of fuel. GATF may have some data on
ink consumption, but I am now more interested in what makes for the best,
most pleasing, printmaking.
Except in areas where more density is desired, I would
think that inkjet prints might look better when black is not allowed where
colors are nearing neutral. As these kinds of printers are described
as being very stable, and since after being linearized they are grey
balanced, it would seem that introducing black would be a fudge factor for
which there is not a need.
Because decisions are made for the structure of the
black plate when making a profile, it would seem that this would entail a
number of profiles that have to be made to suit for particular kinds of
images. Same goes for the other plates. Would one pass an image
through a bunch of different profiles until the results were just right?
I don't understand how one could noodle the image file with any
confidence if it is going to be altered by the profile yet again, and in
particular, noodles that involved the black plate.
In a workflow where a lot of different kinds of imagery
might pass through the same profile, it would seem that some would not
print as well as others. On the other hand, if the target conditions
were known, wouldn't it seem to make sense to avoid profiles, delivering
CMYK that is not to be further processed by a profile? Is it that
profiling software is terrifically better, and intuitive, that one should
busy himself making and noodling the profile rather than the actual image?
I feel that I am either very close to understanding
something about this, or I'm even further away, it's hard to say.
Is it that the profile route is okay for general expediency,
but of less use when it comes to hand crafting an image? Or is
hand crafting the profile the best approach for making the best
print?
Henry Davis
________________________________________________________________________
Date: Tue, 15 Jun 2004 18:48:53 -0400
From: Jim Rich
Subject: Re: Re: CMYK Profiles
Henry,
I think most will agree that using one
setting or one Photoshop adjustment for printing images will
not cut it especially if you want to create high quality prints. Tweaking
an images tones and colors is necessary.
Jim Rich
________________________________________________________________________
Date: Tue, 15 Jun 2004 16:55:50 -0700
From: Peter Figen
Subject: Re: Re: CMYK Profiles
I was at the first press check this afternoon for the
book I had mentioned last week, and the first form looked pretty darned
nice printed on McCoy Silk at 175 l/s.
I also wanted to add to the discussion regarding the PS
CMYK Setup. In my experience, the main deficiency in it is not the total
ink discrepancy, it's that the definitions of custom inks are not very
representative of real inks. I recently had a situation where I had emailed
a ProfileMaker target to a printer in Singapore. They apparently didn't
know exactly what I was going to do with it and folded and creased it,
causing many color patches to lose their ink coverage in the crease. I
opened up the ProfileMaker target file in PS and found the exact patches
that corresponded to the nine printing inks in PS, made single spot
measurements of just those patches and then plugged those Lab numbers into
the Custom Inks setup. The only thing left to do was figure out an
approximate dot gain, which I did by comparing the on screen image to the
printed target and adjusting until the screen matched the print. 14% dot
gain, in this case gave a perfect match and that's what I used. Given the
option, I would always choose the full ProfileMaker route, but this is just
another way to save your butt.
Regarding ProfileMaker honoring requested ink limits: I
too, have had CMYK inkjet profile be off by large amounts, and also
profiles of inkjet based DTP proofing systems. It seems that ProfileMaker
has a mechanism in it that wants to keep you from specifying an ink limit
that is too high, and will, if the data measurements warrant, limit the
total ink levels.
Peter Figen
________________________________________________________________________
Date: Tue, 15 Jun 2004 21:48:58 EDT
From: Dan Margulis
Subject: Re: Re: CMYK Profiles
John Castronovo writes,
This probably isn't within your day-to-day experience
Dan, but would you
care to comment on your best guess for CMYK seps that
are destined to wind
up being printed inkjet on translucent media (simulated
Duratrans). Since
the image will be illuminated from behind, wouldn't we
want far more color
in the shadows than would be customary for press?
Only if it's a flat area of black, in which case I'd go
to 400% if the device allowed it. In a halftone, however, one sticks by the
same rule regardless of output device: the shadow is to be set to the
darkest possible value consistent with retention of detail. This type of
media typically doesn't hold shadow detail well in comparison to quality
sheetfed printing--it's more comparable to web, if that.
The point about more ink coverage is certainly valid,
though. One would want more ink where possible to reduce the amount of
blank translucent substrate; therefore one would *not* be using GCR in
these circumstances. Remember, GCR is usually recommended as a
defensive measure against certain press defects, like poor registration and
unusual fluctuations in ink density. As these problems don't usually exist
in inkjet printing, there's no reason to compensate for them.
Dan Margulis
________________________________________________________________________
Date: Wed, 16 Jun 2004 07:41:09 -0400
From: "John"
Subject: Re: CMYK Profiles/ UCR
Jim Rich Wrote:
As for saving ink with UCR. Sure it might?? But
it is only theory and there
are no numbers on this anywhere I have seen.
UCR has been around since letterpress days and carried
over into offset. Back when only film was provided to publications for ads,
the printers had very specific max. UCD specifications for black seperation
areas. Most were in the 260-280 range, some as low as 220. The printers
(web offset) were non-bending on this and checked every set of film coming
in. If it exceeded, it was kicked back to the agency or provider to be
redone. UCR was purely an economic issue. It allowed higher press speeds,
less heat in the oven required to dry the ink, and less total ink run on
the job. When dealing with normal sheetfed offset run quantities, or a very
shortrun web job, measuring the ink saved on a job with UCR and without
would be a bit tough... probably impossible. Now take a 2,000,000 cutoff
run on a full coverage form, on a full web press, that is going to use up
about 9,000 Lbs of ink... or about $35,000 worth. The difference in UCR
versus No UCR would be measureable in dollars... a ballpark guess would be
about $1,200 for a run like this. Couple presses side by side... 2 or 3
jobs like this a week and it adds up.
John Rawlins
________________________________________________________________________
Date: Wed, 16 Jun 2004 10:15:51 -0400
From: Jim Rich
Subject: Re: Re: CMYK Profiles/ UCR
Hi John,
I hear ya. You are the second person I have ever heard
of who has a hard facts about this. I mean you would think that ( and ther
might be I just have not found it) someone like GATF would have those
numbers. And I used to go looking for this information at scores of prints
shops I have visited. Also, I think I just saw that the average
offset press run is under 10,000 impressions, and that is one reason I made
that statement and I thing we agree on that. Now that I have heard
this let me revise what I said. There are VERY FEW numbers on
saving ink ucr or gcr anywhere I have seen.
Jim Rich
________________________________________________________________________
Date: Wed, 16 Jun 2004 10:59:15 EDT
From: Miles Southworth
Subject: GCR
In the begining, use of GCR and UCR was sold to save
ink and to get better ink trapping and more stabilized neutral grays
when printing. When I asked several sheet fed printers if they saw
savings in ink, their response was that since the grays were neutral
and the color balance was in control with GCR, their pressmen ran the
CMY densities higher for more saturated colors. The
results was better quality, but no ink savings.
Miles Southworth
________________________________________________________________________
Date: Thu, 17 Jun 2004 07:37:46 -0400
From: John Rawlins
Subject: Re: GCR
Miles,
Just like I remember you teaching this back in 1969.
In fact I think this might have been an exam question. Good to hear
from you in the group.
John Rawlins (PR72)
________________________________________________________________________7
Date: Thu, 17 Jun 2004 12:16:32 -0600
From: Chris Murphy
Subject: Re: Re: CMYK Profiles
On Jun 15, 2004, at 7:48 PM, Dan Margulis wrote:
The point about more ink coverage is certainly valid,
though. One would want
more ink where possible to reduce the amount of blank
translucent substrate;
therefore one would *not* be using GCR in these
circumstances. Remember, GCR
is usually recommended as a defensive measure against
certain press defects,
like poor registration and unusual fluctuations in ink
density. As these
problems don't usually exist in inkjet printing,
there's no reason to compensate for
them.
I can attest to the fact maximum GCR saves ink in large
format printing. That one has to also use UCA to get a good dark black is
also true, but you'd need the resulting amount of ink regardless. The
benefit of heavy GCR comes in tones equal to or lighter than 300% where
instead you can print with more black ink and far less of the other three
colors. It also substantially reduces the effect of metamerism, when
viewing such output under different lighting conditions.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.
Receive Schedule
and Pricing updates.
If you let us know how to contact you, we will enter you into our monthly drawing for a free 2-day class (A $895 value).
- Get free video training
- Get free tips, tricks and shortcuts
- Get special sale pricing and free seminar invitations
Sterling Ledet & Associates, Inc. 4100 Perimeter Park South Atlanta, GA 30341
work sales@ledet.com work administration@ledet.com work webmaster@ledet.com
P work 770.414.5007 P work 877.819.2665 F fax 312.896.9094
- Authorized Training Provider
Copyright 2012. Sterling Ledet & Associates. All Rights Reserved.
Design by Digital Scientists, a Conversion Optimization Company.
