Setting up a color management system for multiple presses
Mike Todryk
color management system for multiple presses. We have 2 Komori 40in and
one Komori 28in presses, all less than 3 years old. We have 4 Integersis
proofers (Epsons) and one Veris. We have 2 Brisque 5 RIP's. I have all
of the color tools I need (Spectroscan, X-Rite 530, etc.) and I have in
mind a strategy, but can always use some more expert advice. Thanks.
Mike Todryk
Prepress Tech
Gernot Hoffmann
as far as I know, the FOGRA recommends to use strictly
standard profiles like ISO Coated instead of individual
profiles for each press.
This implies an adjustment of the press in order to force
the machine into a behaviour as standardized by the pre-
defined profile (and an adjustment for the actual task -
ink consuming parts on a page have some influence on fol-
lowing contents).
This strategy is still a matter of discussions.
The underlying idea:
a press has a rather variable behaviour and a today's
profile isn't valid tomorrow.
Relying on standard profiles enables the prepress folks
to prepare their PDF/X1-a files without any reference to
a special printing company.
A link to FOGRA (English available):
http://www.fogra.org/
I'm proof printing by inkjet, using Mutoh Falcon 6100 and
Onyx RIP PosterShop + Profiler
Best/Efi RIP Colorproof + GMB ProfileMaker
Colorgate RIP ProductionServer + integrated Profiler
A comparison is difficult because the software is partly
old and partly new.
By toner printer (LED) OKI C9600, using OKI software +
GMB ProfileMaker
Measuring the targets by X-Rite DTP41.
The toner printer (which might be called a 'laser' for
simplicity) is nowadays a serious alternative to inkjets
for up to A3+. Not as accurate as the inkjet but really
OK for many tasks (much faster).
Best regards --Gernot Hoffmann
Roger Breton
> as far as I know, the FOGRA recommends to use strictly
> standard profiles like ISO Coated instead of individual
> profiles for each press.
This is not such a bad idea, practically.
> This implies an adjustment of the press in order to force
> the machine into a behaviour as standardized by the pre-
> defined profile (and an adjustment for the actual task -
> ink consuming parts on a page have some influence on fol-
> lowing contents).
Right. That is not a bad idea.
> This strategy is still a matter of discussions.
> The underlying idea:
> a press has a rather variable behaviour and a today's
> profile isn't valid tomorrow.
Yes, something like that. To date, I have never "met" a press capable of
completely repeating itself.
> Relying on standard profiles enables the prepress folks
> to prepare their PDF/X1-a files without any reference to
> a special printing company.
Right. That is a sound and wise color management policy.
Roger Breton
Lee Blevins
I doubt you'd have to profile each press individually.
Think about this:
If one press is down can you print the job on the other and match the
color? Sure you can.
You'll have a greater difference in papers than presses for profiling.
To profile your press you should use your house stock and your standard
ink. Do everything you do normally and create a profile of that
condition.
What this profile will provide you with is a representation of your
printing relative to your choice of ink, paper, curve shape, etc.
It's valid for that paper/ink combination only.
The reality of profiling is that it is not a perfect science. After a
perfect profile is built the difference between proof and press with
this profile will be greater than the difference between two of your
presses printing the same job.
Roger Breton
> The reality of profiling is that it is not a perfect science. After a
> perfect profile is built the difference between proof and press with
> this profile will be greater than the difference between two of your
> presses printing the same job.
Why is this so? I mean, why the devil the press could not match it's own
profile?
Roger Breton
Dave Balderstone
<gr...@videotron.ca> wrote:
> Why is this so? I mean, why the devil the press could not match it's own
> profile?
In general because press conditions are variable.
Humidity, paper, ink, fountain solution, plates, blanket conditions,
how hung-over the press crew is, etc...
--
Go read this. Now.
<http://www.ejectejecteject.com/archives/000129.html>
Once you've read that, go read this:
<http://lonestar-mvpa.org/events/2005/05_Katrina.htm>
Roger Breton
What about bringing all those press conditions under tight control.
Then what? Is that still an enterprise doom to failure?
Roger Breton
Neil Gould
>
> What about bringing all those press conditions under tight control.
> Then what? Is that still an enterprise doom to failure?
>
batches differ; ink batches differ; fountain solution differs; press
operators differ; blankets are in constantly degrading. The best color
controls are qualified, experienced pressmen that understand these things
and have a good quality proof to follow. Beyond that, having one's inkjet
restricted to producing colors that can actually be replicated on a press
is as far as "color management" can be expected to go in an open-loop
workflow.
Regards,
Neil
Dave Balderstone
<gr...@videotron.ca> wrote:
> What about bringing all those press conditions under tight control.
> Then what? Is that still an enterprise doom to failure?
I looked seriously at trying to implement a monitor to press ICC-based
workflow a number of years ago. I rather quickly figured out that I'd
need a full time person doing nothing but the profiling, and the press
would be running more tests than paying work.
It's doable, I just can't see the benefit when I look at the
cost/benefit equation. Once ink is hitting paper, I'd rather rely on a
skilled press crew making subjective decisions about color than a
profile.
djb
--
~ Stay Calm... Be Brave... Wait for the Signs ~
------------------------------------------------------
One site: <http://www.balderstone.ca>
The other site, with ww links<http://www.woodenwabbits.com>
Roger Breton
> That's a fantasy.
Hmmh.
> Things change: color and absorbtion properties of paper
> batches differ; ink batches differ; fountain solution differs; press
> operators differ; blankets are in constantly degrading.
Wait. What about maintaining very tight controls on all of the above, then
what?
> The best color
> controls are qualified, experienced pressmen that understand these things
> and have a good quality proof to follow.
Experience pressmen? They are baffled by today's inkjet prints. Talk about
fluorescence in the paper affecting color appearance and the best,
experienced pressmen haven't got a clue as to what's going on, what he has
to match. And, yet, I've seen proofs to press dead nuts matches, without the
pressman actually knowing that he hit the intended colors, because he was
not shown any proof.
> Beyond that, having one's inkjet
> restricted to producing colors that can actually be replicated on a press
> is as far as "color management" can be expected to go in an open-loop
> workflow.
Why do you say that? What is wrong with an inkjet proof simulating the
colors that the press should aim for? And why is that so fundamentally
flaud, in your opinion?
> Regards,
>
> Neil
Roger Breton
Roger Breton
> I looked seriously at trying to implement a monitor to press ICC-based
> workflow a number of years ago. I rather quickly figured out that I'd
> need a full time person doing nothing but the profiling, and the press
> would be running more tests than paying work.
OK. I respect your opinion.
> It's doable, I just can't see the benefit when I look at the
> cost/benefit equation. Once ink is hitting paper, I'd rather rely on a
> skilled press crew making subjective decisions about color than a
> profile.
The profile can't make the pressman hit its own press, that's true. It can't
hold its hand down and tell him how much to open or close those fountain ink
keys. But it's providing some color guidance nevertheless. It's a contract,
an expectation. The conditions in which this contract is created are
critical, I agree. But, when it's well done, by a conscienscious person,
can't it have any fighting chance?
> djb
Roger Breton
Dave Balderstone
<gr...@videotron.ca> wrote:
> The profile can't make the pressman hit its own press, that's true. It can't
> hold its hand down and tell him how much to open or close those fountain ink
> keys. But it's providing some color guidance nevertheless. It's a contract,
> an expectation. The conditions in which this contract is created are
> critical, I agree. But, when it's well done, by a conscienscious person,
> can't it have any fighting chance?
As I said, I looked at it and couldn't justiify the ongoing cost. You
need to look at it in your own context.
My *opinion* is that there are too many variables at press to make
profiling a worthwile exercise. Your mileage may vary.
Grid knows I've been wrong before.
djb
--
Life. Nature's way of keeping meat fresh. -- Dr. Who
Neil Gould
Recently, Roger Breton <gr...@videotron.ca> posted:
> Neil,
>
>> That's a fantasy.
>
> Hmmh.
>
>> Things change: color and absorbtion properties of paper
>> batches differ; ink batches differ; fountain solution differs; press
>> operators differ; blankets are in constantly degrading.
>
> Wait. What about maintaining very tight controls on all of the above,
> then what?
>
The best that I've seen this done is on 5 & 6 color Heidelbergs that are
basically closed-loop systems. In that context, tight control is practical
and effective. So, if that's your work environment, then the answer is
"yes... possibly".
>> The best color
>> controls are qualified, experienced pressmen that understand these
>> things and have a good quality proof to follow.
>
> Experience pressmen? They are baffled by today's inkjet prints. Talk
> about fluorescence in the paper affecting color appearance and the
> best, experienced pressmen haven't got a clue as to what's going on,
> what he has to match. And, yet, I've seen proofs to press dead nuts
> matches, without the pressman actually knowing that he hit the
> intended colors, because he was not shown any proof.
>
That should tell you all you need to know about inkjet "proofs". ;-)
>> Beyond that, having one's inkjet
>> restricted to producing colors that can actually be replicated on a
>> press is as far as "color management" can be expected to go in an
>> open-loop workflow.
>
> Why do you say that? What is wrong with an inkjet proof simulating the
> colors that the press should aim for? And why is that so fundamentally
> flaud, in your opinion?
>
The technologies are completely different. What you have is one device
simulating the possibilities of another. Unless one's inkjet is equipped
with built-in densitometer/spectrometer capabilities and the user profiles
his printer for every job using paper that is identical to the color of
the paper that will be on the press, how close can it possibly get to the
final output?
There is a wide range of quality when it comes to color accuracy, and I
suspect that the vast majority of print jobs are closer to the "loose" end
than the "tight" end of that range. So, many if not most jobs can be
matched close enough to an inkjet "proof" to satisfy the customer, even
though a critical look at those -- say with photospectrometers and such --
would expose some pretty significant differences (see the thread on
"Pantone joke" that Lee started, for example of how widely something that
one might think *should* be controllable can vary). That's why an
experienced printer can be spot on without even seeing such a "proof".
Regards,
Neil
Dave Balderstone
<ne...@myplaceofwork.com> wrote:
> There is a wide range of quality when it comes to color accuracy, and I
> suspect that the vast majority of print jobs are closer to the "loose" end
> than the "tight" end of that range. So, many if not most jobs can be
> matched close enough to an inkjet "proof" to satisfy the customer, even
> though a critical look at those -- say with photospectrometers and such --
> would expose some pretty significant differences (see the thread on
> "Pantone joke" that Lee started, for example of how widely something that
> one might think *should* be controllable can vary). That's why an
> experienced printer can be spot on without even seeing such a "proof".
Interestingly, we are now dealing with proofs from the notion that they
represent "pleasing color" that the press crew will use as a guideline.
We're moving one of our magazines from one peinter to another, and are
in the process of establishing print settings for our Canon i9900 so
that we will provide the press proof (at about $2 a page instead of the
$15 that seems to be the norm in this market...).
Comparing proofs from the new printer to printed pages from the old
one, to the proofs we're pulling off the Canon is a fascinating
exercise.
Lee Blevins
>
> > The profile can't make the pressman hit its own press, that's true. It can't
> > hold its hand down and tell him how much to open or close those fountain ink
> > keys. But it's providing some color guidance nevertheless. It's a contract,
> > an expectation. The conditions in which this contract is created are
> > critical, I agree. But, when it's well done, by a conscienscious person,
> > can't it have any fighting chance?
>
> As I said, I looked at it and couldn't justiify the ongoing cost. You
> need to look at it in your own context.
>
> My *opinion* is that there are too many variables at press to make
> profiling a worthwile exercise. Your mileage may vary.
I'd have to agree.
Once on press a reviewer will usually tweak a little to get the best
reproduction.
So you can spend a lot of time and money on profiling and go to press
and tweak or you can not spend the money and go to press and tweak.
But because of human psycology someone going in for an on-press OK will
usually tweak. For whatever reason.
Maybe just because they went to jerk a two million dollar piece of
equipment around. Who knows.
But I don't see anybody coming up with a perfect monitor to proof to
press system that works so well that I don't see a difference.
It's just a question of what kind of business arrangement one has that
determines how critical the match needs to be.
Is anyone claiming a proof that is less than 1 Delta E from press?
Lee Blevins
I'm not sure what you mean by "match it's own profile."
A press can match itself. Hopefully. But that can depend on a myriad of
variables not usually found in proofing systems.
Proofing systems tend to be very tightly controlled environments for the
very reason that they are proofing systems.
But one must agree, they are not the press. If they were, we'd call
their output "press proofs" and they'd match very well.
But proofing systems today tend to be "off-press proofing systems."
That's because it's just too darn expensive to proof on press so we use
cheaper means.
So that is the meaning of my statement that the difference between proof
and press will be greater than the difference between two presses.
I've had the pleasure of spending years in high quality lithographic
printing in shops with more than one 40 inch press.
If the presses were all of the same make, model and manufacturer their
ability to match each other in the daily work of printing is very good.
Granted, a press with a completely different fountain system, blankets,
pressures etc, would have a harder time matching another but even then
with todays presses they do it very well.
So long as one can control the mechanical issues of high speed color
offset lithography repeatability is not impossible.
But even with color management, profiles and God, getting proofs (read
off-press proofs) to be exact clones of press sheets is utopian thinking
at best.
In my experience just getting opinions of the proof to press agreement
is a frustrating exercise in human psychology.
I'm reminded often when entering this discussion of an even that happend
in my younger days.
I was commsioned with the uneviable task of trying to make a positive
Cromalin from a set of final negatives. In those days negative Cromalin
was horrible and we made all of our separations via positives and
proofed the randoms as positives.
Unfortunately we stripped (this is the US) negative. So our final
product was as set of composite negative film. We couldn't get our
negative Cromalins to look like the original positives so it was decided
we'd have to make a set of positives from the final negatives to make
the final proof.
Now came the hard part. It would take two sets of film to keep emulsion
to emulsion contact. That was expensive in both time and money and
resulted in dust spots.
So I had the brilliant idea that if I could make an out of contact
positive from the negatives I'd get them in one generation. The issue
became how much tonal distortion we'd get from the out of contact.
I set up an ortho darkroom (get the idea of how long ago this was?) with
a point source more than twice the diagonal of the frame away. I thought
results were pretty good so I called in the "experts" to view the
resulting proofs.
I had one proof made from postives that were made in emulsion to
emulsion contact and positives that were made via my new "out of
contact" method.
To my dismay the "color experts" who were reviewing the proots were
emphatic that they could see profound differences in the two proofs.
During the discussion I noticed the markings on the proofs and realized
the proof maker had made a mistake and made the two proofs from the same
set of film.
Througout my career in color reproduction I've noticed this same thing
happening over and over again.
People see color though a prism of their emotions and perceptual
prejudices. One mans "spot on" match is anothers "miles apart."
Roger Breton
Call me an idealist. Maybe I'll change opinion soon but I'll try to get to
the bottom of this press profiling thing once and for all. As a matter of
fact, I am not so much after press profiling for the moment as much as I'm
after press "calibrating". I want to see how feasible is the idea of
standardizing a press behavior. I'll study the fountain solutions influence
on color, the paper composition on color, the inks formulation on color, the
blanket's characteristics and the plates makeup. I hope to understand what
variable affect a printing press stability enough that I can get it into a
repeatable state. If that last idea is an utopia I'll throw in the towel.
But until then I'll pursue my ideals. Not because I want to profile presses
but because I believe that to some degree presses are repeatable. Otherwise
offset lithography could have never reached the development they have.
Profiling only comes after fullfilling this important prerequisite.
Roger Breton
Mike Russell
"Roger Breton" <gr...@videotron.ca> wrote in message
news:BF52341B.125EB%gr...@videotron.ca...
> Lee and David,
>
> Call me an idealist. Maybe I'll change opinion soon but I'll try to get to
> the bottom of this press profiling thing once and for all. As a matter of
> fact, I am not so much after press profiling for the moment as much as I'm
> after press "calibrating". I want to see how feasible is the idea of
> standardizing a press behavior. I'll study the fountain solutions
> influence
> on color, the paper composition on color, the inks formulation on color,
> the
> blanket's characteristics and the plates makeup. I hope to understand what
> variable affect a printing press stability enough that I can get it into a
> repeatable state. If that last idea is an utopia I'll throw in the towel.
> But until then I'll pursue my ideals. Not because I want to profile
> presses
> but because I believe that to some degree presses are repeatable.
> Otherwise
> offset lithography could have never reached the development they have.
>
> Profiling only comes after fullfilling this important prerequisite.
>
> Roger Breton
Here's a compilation, made with permission from the author, responding to a
series of questions from Gernot Hoffman, of a relatively recent thread from
sci.engr.color that you may find relevant.
http://www.curvemeister.com/OtherArticles/NeutralOffsetGrays.htm
--
Mike Russell
www.curvemeister.com
Roger Breton
Thank's. It helps for sure.
Allen Wessels
Roger Breton <gr...@videotron.ca> wrote:
> Lee and David,
>
> Call me an idealist. Maybe I'll change opinion soon but I'll try to get to
> the bottom of this press profiling thing once and for all. As a matter of
> fact, I am not so much after press profiling for the moment as much as I'm
> after press "calibrating". I want to see how feasible is the idea of
> standardizing a press behavior. I'll study the fountain solutions influence
> on color, the paper composition on color, the inks formulation on color, the
> blanket's characteristics and the plates makeup. I hope to understand what
> variable affect a printing press stability enough that I can get it into a
> repeatable state. If that last idea is an utopia I'll throw in the towel.
> But until then I'll pursue my ideals. Not because I want to profile presses
> but because I believe that to some degree presses are repeatable. Otherwise
> offset lithography could have never reached the development they have.
>
> Profiling only comes after fullfilling this important prerequisite.
Go to www.printplanet.com and search the ctpp forum for posts by Erik
Nikkanen (spelling is close). He has written volumes there about the
inherent design problems in today's presses that make them inconsistent
by design.
I think you're assuming a level of development not in evidence. Selling
print is more about the people than the press.
- Allen
Dave Balderstone
Blevins <le...@digitalgraphics.net> wrote:
> had one proof made from postives that were made in emulsion to
> emulsion contact and positives that were made via my new "out of
> contact" method.
>
> To my dismay the "color experts" who were reviewing the proots were
> emphatic that they could see profound differences in the two proofs.
>
> During the discussion I noticed the markings on the proofs and realized
> the proof maker had made a mistake and made the two proofs from the same
> set of film.
Now THAT'S funny!
Color is all in the brain, isn't it?
I attended Dan Margulis's color course a few years back. There were,
coincidentally, three of us in a class of 10 wearing black shirts.
Dan had us stand end - middle - end of one wall, and everyone agreed we
were all wearing "black".
Then he had us stand shoulder to shoulder... Three very different
colors.
Dave Balderstone
Blevins <le...@digitalgraphics.net> wrote:
> 'd have to agree.
>
> Once on press a reviewer will usually tweak a little to get the best
> reproduction.
>
> So you can spend a lot of time and money on profiling and go to press
> and tweak or you can not spend the money and go to press and tweak.
>
> But because of human psycology someone going in for an on-press OK will
> usually tweak. For whatever reason.
>
> Maybe just because they went to jerk a two million dollar piece of
> equipment around. Who knows.
When I do press checks, I like to sit back, have a donut, and wait for
the press crew to wave me over.
Once I'm actually asked to evaluate the press sheets, I'm usually most
interested in the colors in the ads. Id the blue a Ford blue? Is the
green a John Deere green?
Beyond that, I rely on the pressman's opinion for "pleasing color".
It's truly amazing what a pressman will do once he realizes that the
guy standing there at press check actually does appreciate his advice
and knowledge.
I've actually had pressmen argue with me when I've been ready to sign
off on color, because they're not happy yet. Maybe that says something
about my color sense...
But every pressman I know can tell stories about the client that needs
to piss in the soup before it tastes right to them. Me, I figure
"pleasing color" is all I'm after, and whether there's 2% more or less
cyan in that purple really isn't worth anyone's time.
Dave Balderstone
<gr...@videotron.ca> wrote:
> As a matter of
> fact, I am not so much after press profiling for the moment as much as I'm
> after press "calibrating". I want to see how feasible is the idea of
> standardizing a press behavior.
If your goal is to understand how the press you're working on is
behaving, then I say "bravo!"
That's a very good thing, and your business will benefit from your
efforts.
If you're looking for a magic profile bullet, though, you'll be
disappointed.
Working towards understanding how your press behaves under different
conditions will only improve your ability to deliver quality work
efficiently (profitably).
Good on ya. You sound like someone I'd be happy to do business with.
Paul
Dave Balderstone <dave@N_O_T_T_H_I_S.balderstone.ca> wrote:
> When I do press checks, I like to sit back, have a donut, and wait for
> the press crew to wave me over.
> But every pressman I know can tell stories about the client that needs
> to piss in the soup before it tastes right to them. Me, I figure
> "pleasing color" is all I'm after, and whether there's 2% more or less
> cyan in that purple really isn't worth anyone's time.
I worked on a web press for 25 years, your method is perfect.
On a typical press the crew works the color up to their satisfaction.
Then a supervisor comes over and works it to his satisfaction. At that
point they'll call the customer to the press, with the goal being to
have him sign off when he's satisfied too. Then they'll send him on his
way with a few samples to show to his boss or client.
After that it's up to the press crew to maintain the color, matching the
press sheets against the sheet the customer signed. And after that the
color can vary quite a bit, even more so as one shift ends and another
crew takes over. Everyone sees color differently, and it's going to
vary over the length of a press run.
Funny story: On a press check one day my supervisor seemed to be intent
on making the sheet "perfect." The client was an older fellow, and he
seemed to be getting impatient. Finally, after many rounds of very
minor tweaks, the supervisor stepped away for a moment to take care of
something. The client quickly whipped out his pen and signed the sheet.
While signing the sheet he told me that spending any more time tweaking
the color wouldn't sell any more items, he was a business man and he
more important things to tend to--like running his business.
Lee Blevins
> Lee and David,
>
> Call me an idealist. Maybe I'll change opinion soon but I'll try to get to
> the bottom of this press profiling thing once and for all. As a matter of
> fact, I am not so much after press profiling for the moment as much as I'm
> after press "calibrating". I want to see how feasible is the idea of
> standardizing a press behavior. I'll study the fountain solutions influence
> on color, the paper composition on color, the inks formulation on color, the
> blanket's characteristics and the plates makeup. I hope to understand what
> variable affect a printing press stability enough that I can get it into a
> repeatable state. If that last idea is an utopia I'll throw in the towel.
> But until then I'll pursue my ideals. Not because I want to profile presses
> but because I believe that to some degree presses are repeatable. Otherwise
> offset lithography could have never reached the development they have.
>
> Profiling only comes after fullfilling this important prerequisite.
>
> Roger Breton
Without a doubt profiling should not take place on any device until it
has been calibrated. This applies to any device. Not just a press.
I've followed a number of threads on Apple's colorsync list and on rare
occasion here about the validity of profiling presses.
I'm surprised when I see those who would offer that it is a waste of
time or it can't be done relative to profiling a press.
To me it shows a bit of "lack of understanding" or experience with
offset printing.
Of course you should profile your press.
How else are you going to get your inkjet proof to look like your press
sheet?
You're not going to download some profile off the web and plop it in
your rip and as if by magic you suddenly have a simulation of your
ink/paper combination.
And that is the issue.
Very few printers run the same inks and papers on the same type of
machines.
Very few pressmen believe in the same methods of achieving their result.
Printing is not a pure science, it's a hybrid of science and art.
Commonly referred to as a "craft."
There is one thing for sure today and I believe it is inarguable.
That is the old method of mechanical layered proofs is dying and giving
way to inkjet proofs. For one simple reason. Cost.
The ideas presented but the article stated in this thread would leave
one believing that reproducing color at all it nearly impossible via.
offset printing.
The fact that ink film thickness is not absolute across the plate is to
some extent arguing minutia. Of course it isn't. This is a mechanical
process. It is a high speed process where ink laydown depends on what is
being laid down.
We all (at least those of us who have been around presses) have seen the
effect of how an image reproduces differently depending on what is in
front of it and behind it on the sheet. But that difference is not so
far out there that we have to stop the press, throw up our hands and
claim it can't be printed.
The reason you profile a press is because the choice of ink and paper is
unique to a specific press condition. The ideas of how to linearize the
system by choice of what is acceptable dot gain is not universaly held
from shop to shop.
Each shop I've gone to has it's own specific view of what good printing
is and how to achive it.
The profile you get is for a "standard" condition that the shop agrees
is a starting point only.
As the customer chooses different papers and decides to alter SID at a
press ok, the they are knowingly departing from that starting point.
But you must have a starting point. You can't just shoot in the dark.
From my experience the greatest difficulty I've had with profiling
presses is when I ask the pressman to run the test targets at their
"standard densities."
The numbers they then use all too often are not at all what they
actually use in practice.
They simply don't gather data on an ongoing basis to even know what
their standard procedures are. I'd feel much better getting sample sheet
form the last 50 jobs that ran on that particular paper and seeing what
the real "standard density" was.
The usual process of lithographic color printing involves a pressman
looking at a target proof and fiddling the densites until they think
they've achieched a good match.
For those that don't believe in profiling a press what would they
substitute? A profle made from a Matchprint or Kodak Approval?
Need I remind them that those proofs do not match the hue of the ink nor
do they account for ink trapping nor would they have the dot gain of the
specific ink/paper/press combination.
To profile a press the person managing the act of profiling should be
someone experienced with lithographic printing. A color scientist is
entirely the wrong person to do this unless they have real world
experience in the day to day business of printing.
Roger Breton
> Funny story: On a press check one day my supervisor seemed to be intent
> on making the sheet "perfect." The client was an older fellow, and he
> seemed to be getting impatient. Finally, after many rounds of very
> minor tweaks, the supervisor stepped away for a moment to take care of
> something. The client quickly whipped out his pen and signed the sheet.
> While signing the sheet he told me that spending any more time tweaking
> the color wouldn't sell any more items, he was a business man and he
> more important things to tend to--like running his business.
Moral of the story lies in the Law of diminishing returns, a fundamental
micro-economic concept that says that up to a point, it is worth pouring in
additional resources in a process to produce incremental output, beyond that
the incremental output becomes negative. Transposed onto a printing press, I
think it means that it's worth getting as much of the mechanical and optical
properties of an press under control but beyond a certain point, it's not
worth it. And it's a matter of judgement, I guess, of where to draw the
line. Yes, there is a lot at stake, businesswisem and there are zillion of
variables that influences the process. But I believe it's possible to study
the ones that explain 90%+ of the ensuing color variance, get those under
tight stewardship, and accept the rest as random behavior on the part of the
system and live with it.
The one thing that I still can't operationalize or comprehend in the
behavior of the ink on the sheet is its variability with regards to image
characteristics. If I print the very same printing form ten times on the
same press, the same day, by the same crew of operators, on the same paper,
using the same inks, with a fresh new set of plates, what are my chances
that the 10th run looks remotely like the 1st run? Good, wouldn't you say?
In fact, I would expect all 10 run to be within a few deltaE of each other,
otherwise I think I have a big problem. Is this a fair statement or am I
expecting too much right there?
Erik Nikkanen, I know you're out there lurking. I know I keep coming back to
your posts for help with understand this phenomenon but the way the ink is
actually consumed by the printing form still elludes me. In particular, one
reccurent theme I have difficulty working around, that is tied to this form
to form inking variation is ghosting, the fact that the ink rollers don't
have time to fully replenish by the time they hit the next part of a sheet,
leaving a 'starved' impression. I am just not clear enough, from an image
content point of view, on the conditions that invariably lead to ghosting.
Can experienced people look at a prepress proof and reliably predict
ghosting, and by extension uneven inking pattern?
What I'm getting at is, given a set of forms that are perfectly
interchangeable in terms of ink consumption across the sheet -- bear with
me, I know this is not realistic -- , what can be said about the
predictibility or reproduciblity of the press results? Could I then
substitute a characterization target for a profiled image and expect some
level of color matching to ensue as a consequence?
Thank's for your help David and Paul and Mike and Lee and Allen,
Roger Breton
Lee Blevins
> In particular, one
> reccurent theme I have difficulty working around, that is tied to this form
> to form inking variation is ghosting, the fact that the ink rollers don't
> have time to fully replenish by the time they hit the next part of a sheet,
> leaving a 'starved' impression. I am just not clear enough, from an image
> content point of view, on the conditions that invariably lead to ghosting.
> Can experienced people look at a prepress proof and reliably predict
> ghosting, and by extension uneven inking pattern?
As in most of this trade (printing) terms like "ghosting" can mean one
thing in one part of the country and another in another.
Ghosting to me isn't a reduction in ink, but rather an increase in ink
in an unwanted area. Its' cause is usually the result of a large hole in
a solid area of a spot color.
Because a sufficient amount of ink has to be put on the stock to cover
the large solid the press doesn't really have the ability to lower the
in in the specific zone where the large hole is.
That "hole" creates a need for less ink in that "ink zone." Keep in mind
that a lithographic press today is a cylinder device spinning at high
speed so a "zone" is a band around the circumference of the cylinder.
The ink zones are bands of a certain width. There may be one zone per
inch or smaller but they are not infinitly variable. Even if they were
it's just nearly impossible to cut the ink down to the limit needed over
the sharp distance the transition requires.
So what happens is the area immediately in front of and behind the hole
get more ink than the rest of the sheet.
And yes, as an experienced lithographer if you show me a proof I'll tell
you it's going to ghost. I'll even be able to tell you some tricks of
the trade to lessen the effects of the ghost. I can even tell you that
some colors ghost worse than others.
Now if you want to knock out a large hole in the middle of a process
image I'm going to advise you to revise your design.
That's when I get into trouble. Trying to explain to a designer that not
all things are possible and you actually have to design around the
mechanical limitations of this process.
Like the time I advised a designer that using 5 pt KO type wasn't a good
idea on their job and they simply said "well if you'd put the dots where
they're supposed to be it wouldn't be a problem."
Offset Lithography today is a Photo/Mechanical process. Many things are
possible in the "photo" part that are impssible in the mechanical part.
Color management may be more effective in photographics processes than
it is in the mechanical process of offset printing but it is growing in
its use and ability to control color reproduction.
With digital photography replacing conventional cmyk drum scanners and
inkjet proofers replacing the mechanical lay downs proofs like
Matchprint and Kodak Approval - color management is going to be a major
player in commercial offset printing. Those printers who are not up to
speed on it would be well advised to start reading.
I can't wait to see an offset press with light cyan and light magenta so
the solid ink density of the process colors can be increased to extend
the color gammut without encountering excessive dot gain at the midtone.
Is any printer doing this? If not, what's taking them so long?
Giordano Galli
> My *opinion* is that there are too many variables at press to make
> profiling a worthwile exercise. Your mileage may vary.
I doubt even this.
Let's take the following 8-up forms as an example:
X A X X
X B X X
X A X X
X C X X
Page A contains a 50C coloured shape.
Page B has a 100C background on it.
Page C is blank.
Imho no profile can take into account the different influence of B and C
on A.
Except for laser printers, I can't think of any proof technologies
sharing this problem with offset presses.
The same holds true for colour shifts caused by sudden transitions in
the ink coverage or local misregistration (paper stretch): I expect most
proof devices to be free from these problems and thus easier to
characterize (ie calibrate and profile).
giordano
Lee Blevins
But...
Consider this.
Wet ink being laid on top of wet ink at high speeds just micro seconds
after the first ink is laid down.
The result? Wet on wet trapping. It affects primary colors such as deep
red, blues and greens or colors that rely on 90+ percent
And...
The hue of the colorant. It's unique to the ink being used.
Without profiling how do you simulate this?
Dave Balderstone
Blevins <le...@digitalgraphics.net> wrote:
> But...
>
> Consider this.
>
> Wet ink being laid on top of wet ink at high speeds just micro seconds
> after the first ink is laid down.
>
> The result? Wet on wet trapping. It affects primary colors such as deep
> red, blues and greens or colors that rely on 90+ percent
>
> And...
>
> The hue of the colorant. It's unique to the ink being used.
>
> Without profiling how do you simulate this?
This is beyond profiling, though, isn't it?
Now we're talking plate profiling rather than press profiling, and
digitally controlled ink keys rather than press profiling, though
understanding the behaviour of the press is important.
N'est ce pas?
Giordano Galli
> For those that don't believe in profiling a press what would they
> substitute? A profle made from a Matchprint or Kodak Approval?
Precisely.
> Need I remind them
No need, thanks. :)
> that those proofs do not match the hue of the ink nor
> do they account for ink trapping nor would they have the dot gain of the
> specific ink/paper/press combination
Very well said. However...
1) Proofs work, and I mean routinely. You sign them, and they get
reproduced as accurately as possible.
2) Imho, the inherent tendency of presses to elude calibration far
outweighs the differences you mention.
> To profile a press the person managing the act of profiling should be
> someone experienced with lithographic printing
I wonder how many ace pressmen would buy the idea of profiling presses.
Have you got any experience to share?
giordano
Lee Blevins
> > To profile a press the person managing the act of profiling should be
> > someone experienced with lithographic printing
>
> I wonder how many ace pressmen would buy the idea of profiling presses.
> Have you got any experience to share?
I have profiled a few presses. In each case I was able to get a better
proof to press match than a Kodak Approval or Matchprint.
The key to profiling a press is instructing the pressman on how to run
the targets. They need supervision or they'll default to thinking it's
just a normal run.
The indicator that something is wrong is when they ask for a proof of
the targets to match.
Lee Blevins
> This is beyond profiling, though, isn't it?
No. It is the very reason a profile is needed.
To capture the actual color space that is produced by the
ink/paper/press combination.
The biggest problem I've had in profiling is getting the pressman to
understand that he's not to try to match a proof when running the
targets.
I'm really surprised at the idea that that a press is not worth
profiling. I've been doing it with great success for a few years.
For those that would suggest profiling a proof they need to be reminded
that laminated contract proofs are dying. Very soon they won't be around
to use as a reference and all proofing will be done on inkjet.
Dave Balderstone
Blevins <le...@digitalgraphics.net> wrote:
> I'm really surprised at the idea that that a press is not worth
> profiling. I've been doing it with great success for a few years.
What conditions cause you to re-profile?
Erik Nikkanen
I have been away for a while and it has taken some time to catch up. A lot
of the things discussed on this thread have been basically true but other
things are quite far off. This is due to the confusing state of knowledge
that even the "experts" have about the process. It is also related to how
one looks at a problem. The usual method viewed by those in the industry
is to look at the symptoms of the problem and not at the fundamental cause.
If one did look at the fundamental causes and correct them, I am quite
sure we would get to your ideal goal.
On the subject of ghosting, we could be talking about mechanical ghosting
or starvation ghosting. These are different and are caused by different
mechanisms. The industry usually blames the printed form for these
problems but the real fundamental cause is the inability of the press to
effectively manage ink films that are applied to the plate. Managing ink
films on press has to do with the overall ink feed, nip locations, roller
sizes, roller positions and oscillation.
A press should be profiled but there are issues with the accuracy of the
profile. Each revolution of the press prints slightly differently due to
the patterns of ink being taken off form rollers by the plate, interacting
with other rollers which then reapply them to the form rollers and then
being reapplied to the plate. So there is some random variation. There
are also some periodic variation which is mainly due to these starved
patterns on the form rollers being reapplied to the plate.
These kind of variations can not be totally removed but they can be reduced
to levels that are below noticeable levels. This is directly related to
the design of the roller train. The solution is not in understanding how
the press works now but in understanding how it must be changed to reduce
variation.
This is a very big problem and it is not a technical problem. The
technical problems are quite easy and much of the work has been done but
needs to be implemented. The real problem is a management problem.
Managers in printing companies and in equipment suppliers are not willing
to look at the process in a fundamental way and are scared to death to
innovate because they do not know what to do.
There are a lot of problems that need to be corrected. Many of them are in
the press itself but there are also problems in prepress. While there have
been comments even on this thread about the interest and reality of running
different inks, higher densities or different papers, etc., the standard
organizations are trying to make standards that will only work if
everything is the same. A one profile fits all situation. For many
reasons they are going off in the wrong direction and will not get the
results they desire.
The fact that printing is done every day is evidence that the process is
not that difficult. And when it is analysed properly, it is quite easy to
understand what the systemic requirements need to be to make it very
consistent and predictable.
One very big problem for the person who is new to the process or one who
has seen the process for 30 years is that the level of science in the
industry is very poor. Collectively, there are so many issues that have
not been explained properly that there then is no ability to design
predictable performance with that confused knowledge.
It is very difficult to talk about these issues because in general people
have no imagination of what is possible. They can't imagine because they
have been fed myth as fact and experience as knowledge.
Erik
Giordano Galli
> I have profiled a few presses
Thanks Lee.
If you can bear with me, I'd be most interested in investigating the
outcome.
First of all, how tightly is the profile bound to the press & paper
combination?
What about applying it to a sibling press?
How wide a range of papers will a coated profile cover?
giordano
Dave Balderstone
<g...@skylink.it> wrote:
> Lee Blevins <le...@digitalgraphics.net> wrote:
>
> > I have profiled a few presses
>
> Thanks Lee.
>
> If you can bear with me, I'd be most interested in investigating the
> outcome.
I'm quite interested in this thread as well.
Lee, if you have the time I would very much like to hear more.
Lee Blevins
> Lee Blevins <le...@digitalgraphics.net> wrote:
>
> > I have profiled a few presses
>
> Thanks Lee.
>
> If you can bear with me, I'd be most interested in investigating the
> outcome.
>
> First of all, how tightly is the profile bound to the press & paper
> combination?
Most commerical sheetfed printers have a "house" stock. In my experience
it's usually a #2 gloss coated paper. This stock is used as a basis for
quoting before another stock is chosen. There's a lot of reasons why
this stock is and how it affects the business.
If you're going to profile a printer, it's the first paper you profile.
There are other reasons to profile. In one case the printer has a client
that prints on an uncoated stock. The drop in solid ink density and
increase in midtone gain have a profound effect on the color gamut.
In another case, it's a big job that's color critical and worth 100's of
K's. The client wants a better proof to press match.
Or it might be a magazine that runs on a web on a particular paper. Be
it a weekly or a monthly, it's worth profiling to get a better proof to
press match.
Laminated proofs like a Matchprint or Kodak Approval can't really show
the effects of drops in solid ink density. Their ink trapping is 100%.
In the case of the Approval you can move SID and adjust dot gain but
Kodak will tell you when you're setting it up about the problem with the
yellow. What problem you say? It's about ink trapping.
The yellow goes down last on Approval. Since it's a dry laminated proof
it's a 100% trap. That's not how a press works. Most commercial sheetfed
presses today run KCMY. The yellow goes down last but it doesn't trap
that well to the already coated shadow areas. The effect is less yellow
actualy sticks there than would on a Matchprint or Approval.
When the Approval lays down the yellow last, you get a "bronzing" of the
shadows because it's a 100% trap. Kodak says to counter this you run the
yellow density about 10 points less than press. So much for proof to
press agreement.
Inkjets with color managment don't have these problems. They can nail
every position of the tonal scale as it is rendered on press. The
shadows come out with the same cast as press. The ink trap reduction in
overprints is as it happens on press.
> What about applying it to a sibling press?
Basically you have the same problem you'd have with a Matchprint but to
a lesser degree. Can all the presses in the pressroom match the
Matchprint? No self respecting print salesman would claim anything
othewise but the difference between them is minimal.
The reality is this. If you have 4 four color presses in the same
pressroom being fed the same ink and paper, The plates are being made in
the same plateroom from the same prepress department. There's a common
pressroom supervisor controlling blanket purchasing, fountain solutions,
washup fluid, etc., what you have is a very homogenous enviroment in
which color deviation between presses would be very minimal. Now mind
you if one of the presses sucked a board through or it has gear wear or
slop that produces an anomaly outside normal working conditions, yes,
you'll have a greater deviation. But normal working 4 color presses that
are in good working order can come very close to each other.
Very close.
I have personally withnessed jobs that ran the first time on press #1
and re-run on press #3 and was little if any difference. Many high
quality jobs are printed part on one press and part on another. You
couldn't do business in high volume printing if it weren't true.
On the other hand, if you have one 2 color that does 4 color work as 2+2
and one inline 4 color, God can't make 'em match. So you profile them
separately. By the same token, you can't compare a web to a sheetfed.
And what can I say about SFUPs? There's nothing anybody can say. There's
alot of them and who knows what will happen in their pressrooms. A
poorly managed environment is never going to be worth profiling. Most of
them can't print the same color the next day on the same press with the
same paper.
> How wide a range of papers will a coated profile cover?
It works like this. How close are papers? I have a job printed on this
or that paper, and now it's not made anymore and the paper merchant
tells me that this new paper is identical. We run a color target on it,
evaluate and decide if we have to re-profile.
Case in point, I profile Xerox Xpressions on my Phaser 7750GX. Competing
paper merchant say I don't have to spend the money on that expensive
Xpressions, he's got a paper that's just like it and cheaper.
My point is papers can be very close. One can switch papers and the
diffference may not be worth re-profiling.
But papers can be very different. A profile made on a #1 Gloss coated is
not going to work on a #2 dull coated or offset stock. They're just too
different. They have profoundly different ink absorbtion rates and will
deliver comepletely different color gamuts.
I tell a printer that the profile is good for that ink/paper combo.
The hard part I said is instructing the pressmen on how to print the
targets. From my experince most good process color pressmen like to
paint pictures. They get the plates hung and roll up and start looking
at the proofs they have and the result they're getting out of the
delivery unit.
They pride themselves in their sharp eyes and ability to see minute
color differences and know what unit to key up or down to pull in a
match.
To tell them that you just want them to balance the sheet evenly and run
to a specific number makes them feel like they're being treated as a
press monkey. A robot.
To them, they are the color geniuses that save the job by their mastery
of adjusting the color.
Most of them will see this mechanical approach as a threat to their
worth and not be very inclined to do a good job on printing the targets.
It has to be supervised by someone who understands what has to be done.
I remember many years ago (I'm too old now) when scanning densitomers
first came out on presses. I was talking to the second shift pressman on
a new 6 color that had a scanning densitomer and he told me it didn't
work because when he put the powder on it shifted the density.
Duh Sir. Genius. He couldn't deduce that to maintain run consistency he
should take the readings AFTER HE PUT THE POWDER ON.
My point is that pressmen (most that I've met) are not much interested
in optical theory of light or color, they're pretty much mechanics.
Printing a set of profile targets on a press has to be done with not
only care, but with the understanding of how it applies to the work that
is to follow.
You can't choose solid ink densities just because they were written in a
book by somebody in the past. If you run the targets with less density
than you print normally, the restricted color gamut will never let you
show the what actual print will look like.
You have to setup the profile sheet to print with enough ink takeoff to
make the test resemble normal conditions. You can't stick an 8 x 10
target in the middle of a 25 x 38 sheet with nothing but a color bar and
expect that to work. This is a point where if you can't make the
decision yourslelf you might have to get the pressman to tell you if
there's issues with the layout with respect to ink takeoff.
After you have a profile and have gone through that war you have to
install it in a color management aware rip and set it up for use.
For my money, only relative colorimetric rendering works. Forget about
perceptual and absolute colorimetric.
Since relative colorimetric rendering skews the image to the paper white
it is imperative that you have a proofing paper with a similar (very
close) white point to the paper you print on.
This has been my biggest challenge in inkjet proofing.
Trying to have the inkjet "dirty" the paper white by putting in a cast
just doesn't work. We're talking about a very subtle light pastel color
that just can't be hit. As I write this I'm sure somebody is making a
proof this way and loves it. For me I don't like it.
It's too bad Iris wouldn't make their Iris Pro Glossy for Epson or HP.
It has a white point very close to offset printing paper.
Last week I tried HP's Graphic Arts Proofing paper and was very pleased
with the results.
I've tried the Mitsubishi Series 1307 and had to give it the thumbs
down. To dark a white point.
Some fine art papers actually have white points that match offset
printing papers better.
At Print05 I was very impressed with the papers Epson was showing for
graphic arts proofing. They had a commercial and a publication that
seemed very good. I haven't tried them but I decided I need an Epson
9800.
Point being that without a matched white point on the paper, you're
going to have a hard time getting a good proof to press match. So that
means you're not going ot be able to proof every ink/paper combo out
there.
So what do you tell the printer that prints every job on a different
paper?
You tell him "Yer screwed. Live with it."
In Mr. Balderstone's case where he is using a magazine printer I'd make
a profile run a condition of doing business. I wouldn't trust them to do
it, I'd setup the targests and make the profile myself. I'd also get a
bunch of blank sheets of the stock and start researching inkjet papers
with similar white points.
Then I'd get an Epson with a good rip like Onyx Postershop.
I don't think you can get a good result from a CLC or Xerox. At least
not the ones I've tested. The inkjet has a wider gamut and can come
closer to press but you have to deal with the inkjet coated paper issue
of finding a good white point.
My whole point is that I believe that you can get a better proof to
press match with an inkjet and a good profile than any other system I've
seen. And I own those other systems and sell their output so I'm not
ignorant of what their capabilities are.
>
>
> giordano
Lee Blevins
> In article <1h396ku.1yot4h911rohgyN%le...@digitalgraphics.net>, Lee
> Blevins <le...@digitalgraphics.net> wrote:
>
> > I'm really surprised at the idea that that a press is not worth
> > profiling. I've been doing it with great success for a few years.
>
> What conditions cause you to re-profile?
The key to managing proofing is to have a starndard test object.
We have a test target that we run on everything we print on. By using
the same target over and over we get so used to seeing it I can tell if
profiling is needed by just looking at it. But we actually take it in a
light booth and compare it to a previous run on the same media. We save
and date runs of the target and store them for future comparing.
After we profile any new media the first thing we do is run this test
target and evaluate it. If we get an accurate reproduction of the
target, profiling was a success.
When we load a new roll of paper we run the target to make sure we're
still getting the same color.
So the answer to your question is when we see a visible difference in
the reproduction of our standard test object.
For same media/ink combos that show a difference it usually doesn't
require re-profling. We just recheck SID and linearity. Once an
ink/paper combo is profiled you don't have to re-profile but you do have
to watch the solid ink density and the tone reproduction. That can
change from batch to batch of paper.
The test form has a variety of images and test targets.
For example.
We have 21 step scales of each process color.
Neutral gray bars in stepped and continuous tone.
A set of gray balance targets for highlight, quartertone, midtone and
three quarter tone. These are squares that are a flat tint of cyan with
a yellow and magenta gradation running over them at 90 degrees to each
other. It creates a chart that enables you to find the neutral spot and
locate the densities on the x and y axis.
A Q60 Target.
Caucasian and African flesh. We humans seem to have an uncannly ability
to determine when our flesh tones are casted.
A high key photo. White rabbit in a snow storm. Ours isn't that but
something that is very light in the tonal scale. It's a picture of a
white adobe building.
A low key photo. Midnight in a coal mine.
An average photo with a lot of saturated color. We have a fruit bowl
with apples, oranges, grapes, watermellon, etc.
By using this same target over and over it becomes a very quick
evaluation.
I suggest anybody serious about color have a standard test object that
they use repeatedly.
We built ours as a two page spread. One of the pages has just the
technical charts on it and the other has the images.
We can run it as an 11 x17 with both or as a single page with just the
scales for metering with a densitometer.
Dave Balderstone
Blevins <le...@digitalgraphics.net> wrote:
> So the answer to your question is when we see a visible difference in
> the reproduction of our standard test object.
Thank you.