Dan Margulis Applied Color Theory
The Future of Custom Separators
Future of custom separators?
Posted by: "aliruleweb"
Mon Mar 16, 2009 4:36 pm (PDT)
Hi all,
In doing some research regarding the adoption of G7 and
SWOP3 by some of the printers we send files to, several questions have
started to percolate in my mind regarding the future and need for high end
separators in the printing industry. This is a very disturbing concept for
me as I've been working in a high end separator environment for 25 years. I
think I need a little reality check.
A little back ground about our images:
I work for a prepress house where we convert all our
RGB and provided CMYK images into one basic in-house custom profile using
an action to blend several conversion techniques to achieve the best
separation with the most channel detail targeting the end printing
conditions. This has been a very successful process which has evolved over
the years based off of our experience on press across many projects. We
then use a multitude of techniques to reduce skin texture, achieve critical
sample match and enhance shape and detail. We'll even re-channel mix a
single item in an image to have a long black with in a short black
separation. Additionally, we're creating custom CMYK quad tones starting
with a half tone then using channel mixer and curves to get our final
result. In all cases, we create contract proofs with a Fuji Final Proof
calibrated to the clients printers system that are balanced to on press
checks.
Conditions we separate to and their success:
*In all cases, the client has chosen the printer. We
are never involved in that selection process.
1. Web printers calibrated to SWOP who use our files as
is (no other conversion). These printers can match our contract proofs
fine.
2. Web printers calibrated to SWOP who convert our
files to adjust and control GCR using who knows what input profile as we do
not tag our files. Depending on the printer, we've seen mixed results.
3. Web printers calibrated using G7 who don't convert
our images. Images don't print too poorly but there is a lot of room for
improvement hence my research after the fact.
4. Web printers calibrated using G7 who do convert our
images. They can't match their own in house proofs but they are able to run
closer to our Fuji proofs.
My questions:
- If the printer converts our files, who is then
responsible for how the job prints? Us or them? I know this is a some what
rhetorical question. These tend to be the group who have trouble matching
our proofs not surprisingly.
- If a client has been sold the benefits of a G7
process on press and the printer is going to re-seperate our files any way,
why don't we just send RGB?
- If we are having to use devise link profiles to
simulate the G7 process on our Fuji proofing system, why don't we just go
to an Epson proof. What is the benefit of the laminate system when the
Epson can do a paper simulation?
- First there were camera separations, then scanner
separations now profile separations. In each case the high end separator
always tweaked the separation once it was in CMYK. What is the argument for
this continued practice?
I would appreciate any and all comments.
Thanks,
Alison Walker
___________________________________________________________________________
Re: Future of custom separators?
Posted by: Dan Margulis
Wed Mar 18, 2009 7:31 am (PDT)
Alison writes,
In doing some research regarding the adoption of G7 and
SWOP3 by
some of the printers we send files to, several
questions have
started to percolate in my mind regarding the future
and need for
high end separators in the printing industry. This is a
very
disturbing concept for me as I've been working in a
high end
separator environment for 25 years. I think I need a
little reality
check.
The need for professional separators is far less than
what it was years ago for the obvious reason that clients can now do it
themselves. Those who don't care about quality or (hopefully) are skilled
themselves have no need for somebody to do this work for them.
That still leaves clients for the service, but to keep
them, you have to be *decisively* better quality-wise than they would be if
they tried it themselves. You have to be able to achieve consistent quality
on press, often in spite of the printer the client has chosen. Printers are
generally quite ignorant of separation issues. Allowing them to ruin jobs
at a whim is fatal, so you have to step in where needed.
Conditions we separate to and their success:
*In all cases, the client has chosen the printer. We
are never
involved in that selection process.
This has been normal in the industry since the
beginning of time.
1. Web printers calibrated to SWOP who use our files as
is (no
other conversion). These printers can match our
contract proofs fine.
Ditto.
2. Web printers calibrated to SWOP who convert our
files to adjust
and control GCR using who knows what input profile as
we do not tag
our files. Depending on the printer, we've seen mixed
results.
Not surprising. My suggestion is that your instructions
should include a statement to the effect that the GCR and other separation
settings in the file are critical to print quality, and that reseparation
by the printer will result in the rejection of the printed job. I would
print this on a red label and paste it on the bottom of the print order.
3. Web printers calibrated using G7 who don't convert
our images.
Images don't print too poorly but there is a lot of
room for
improvement hence my research after the fact.
If they are printing to a standard, any standard, as
long as they are printing to it consistently you should be able to adjust
to it.
4. Web printers calibrated using G7 who do convert our
images. They
can't match their own in house proofs but they are able
to run
closer to our Fuji proofs.
An infinite number of monkeys with an infinite number
of typewriters...
My questions:
- If the printer converts our files, who is then
responsible for
how the job prints? Us or them?
Them. They have seen that you clearly indicated, by
your failure to embed a profile, that you did not wish them to reseparate
your file, and they have done so anyway. The situation is exactly the same
as if they had entered your image files and recorrected them. If you happen
to like the result they got, by all means accept it. However, unauthorized
changes like these give you and your client a right to reject the work if
it is not to your satisfaction. The question of whether ink densities are
within proper tolerances is not relevant.
These tend to be the group who have trouble matching
our proofs not
surprisingly.
No, it's not a surprise. The mere fact that they would
guess at and assign a profile, let alone reseparate the file, indicates
that they have little regard for quality.
If a client has been sold the benefits of a G7 process
on press
and the printer is going to re-seperate our files any
way, why
don't we just send RGB?
This is like asking if the client is going to leave a
large amount of cash in plain sight on the front seat of his car, wouldn't
it be a better idea to leave the door unlocked so that anyone wishing to
remove it will not have to break the window to do so.
If we are having to use devise link profiles to
simulate the G7
process on our Fuji proofing system, why don't we just
go to an
Epson proof. What is the benefit of the laminate system
when the
Epson can do a paper simulation?
The quality of modern large-format printers has gotten
so good, and the speed improved so much, that it's hard to justify
something more expensive. There are many ways of adjusting such printers to
achieve results good enough for contract proofs.
First there were camera separations, then scanner
separations now
profile separations. In each case the high end
separator always
tweaked the separation once it was in CMYK. What is the
argument
for this continued practice?
Camera separations don't produce a digital file, so
there was no electronic tweaking. Scanner separations were
pre-corrected--that is, the scanner operator should have set endpoints and
fixed obvious color deficiencies. So the game is different today. CMYK
correction doesn't do well with bad originals. Where it shines is in small
contrast enhancements, as a recipient of channel blends, and in adjustment
of shadow values. If the file has to end up in CMYK anyway, these tweaks
are valuable.
Dan Margulis
___________________________________________________________________________
.
Re: Future of custom separators?
Posted by: Stephen Marsh
Mon Mar 23, 2009 4:06 am (PDT)
Alison Walker wrote:
- If the printer converts our files, who is then
responsible for how
the job prints? Us or them? I know this is a some what
rhetorical
question. These tend to be the group who have trouble
matching our
proofs not surprisingly.
As you know Alison, it depends on who you are asking!
The customer of the printer would say that the printer is responsible. The
printer would say that they are not responsible, however you may not be
given the chance to take over from them and have to work around the issue
(as has always been the case).
It all depends on your clout, being a customer of a
printer does not always mean that you are always right, despite that old
saying.
"Us or them?" sadly it often comes down to
this, rather than a "partnership" where both parties are
communicating (this is not an attack on you Alison, just a general
comment).
- If a client has been sold the benefits of a G7
process on press and
the printer is going to re-seperate our files any way,
why don't we just
send RGB?
Totally reseparate as in ICC, or regenerate as in DLP
(perhaps honouring or repurposing the K channel data and or other areas of
the original separation values)?
Most of the advantages are lost if they totally
reseparate the image and seriously alter the K plate and relation to CMY
etc. If you are lucky they may have somebody clued in enough to repurpose
the original K plate into the new seps. The gamut compression/mapping has
taken place, so this aspect you can control rather than relying on their
conversion. Even if handing off RGB, one could dupe the file and convert to
CMYK, before going back to an accepted tagged RGB file (perhaps accounting
for any density loss in the final RGB file).
- If we are having to use devise link profiles to
simulate the G7
process on our Fuji proofing system, why don't we just
go to an Epson
proof. What is the benefit of the laminate system when
the Epson can do
a paper simulation?
I am not familiar with this device. Does it produce
halftones similar to the final press? Is this critical for you?
- First there were camera separations, then scanner
separations now
profile separations. In each case the high end
separator always tweaked
the separation once it was in CMYK. What is the
argument for this
continued practice?
The argument is crafted colour vs. commodity colour.
The ability to outdo your competitors (or to fail by being an "unlucky
expert") vs. average uniformity. Again, many of these direct benefits
are lost if the CMYK is reseparated via a simple ICC transform.
The technology and the print market decided this issue
long ago (it just took a little longer than with typesetting), however
there are still some that use the services of folk such as yourself, or
acquire the knowledge to create above average separations for a given
condition (usually with post separation CMYK edits).
Stephen Marsh
___________________________________________________________________________
Re: Future of custom separators?
Posted by: J Walton
Mon Mar 23, 2009 1:32 pm (PDT)
If we are having to use devise link profiles to
simulate the G7
process on our Fuji proofing system, why don't we just
go to an Epson
proof. What is the benefit of the laminate system when
the Epson can do
a paper simulation?
Sorry to be jumping in on this so late, but if I can
add a little bit I'll try.
I assume by Fuji proofing system you mean the
FinalProof. I have some experience with that system (although it's been a
few years since I used one, and more than 5 since I helped set one up).
We purchased the FinalProof to replace the 4-up proofs
we were doing with the Creo. Much better than the Creholes, and better
overall than the Epsons.
The advantages over inkjet?
: : Well, you don't need to do paper simulation since
you can laminate to the exact stock you will be running through press.
: : You will get a higher ink density using a laminate
system - at least a few years ago the Epsons couldn't get quite dark
enough.
: : Halftone dots are an important part of some
proofing systems, since it's hard to catch moires on an inkjet proof
without halftones.
: : If you set things up right you can shoot 1-bit
tiffs that have already been screened, so you have the exact same rosettes
on your FinalProof as you do on your press run. This is, of course, if you
are the printer and can control the screening.
The disadvantages are mainly cost-oriented - inket
proofs are *much* cheaper, but don't look *much* worse. And the big systems
(GMG, EFI, etc.) do some method of halftone screening, although you'll
often find it's not that important anymore unless you are doing a lot of
product shots.
We actually were not using Device Link profiles with
the FinalProofs. We used their calibration utility and used curves to tweak
the color until it matched our press *exactly*. It was not easy, and it was
incredibly expensive (in material usage), but we managed to get it right on
the money without any profile tweaks. I like that method because we didn't
have to worry about blacks getting remade or anything like that. The
downside to that is the ink traps didn't look quite right, as they never
seem to look right without a profile.
We brought in Eric Magnusson (Left Dakota) to help us
with the set up of our inkjet proofers, and while he was going to use his
own product (Link-O-Later) on the FinalProofs even he acknowledged we were
so close with the curves it wouldn't do us any good.
Hope this helps...
J Walton