Re: Laminate profile - Abstract profile?
Re: Laminate profile - Abstract profile?
- Subject: Re: Laminate profile - Abstract profile?
- From: Hanno Hoffstadt <email@hidden>
- Date: Sat, 8 Aug 2009 00:22:37 +0200
Hi,
sorry for not responding earlier. I'm the guy with the presentation at
the 2008 Fogra Color Management Symposium Dan mentioned.
One of the workflow issues is consistency between the laminated and
unlaminated profile. Nils is right that the most consistent way is to
print your unlaminated target, make a pre-LAM profile, laminate the
same target, make a post-LAM profile. Then you've got a consistent
pair. You can adjust your ink zones according to the "wacky" pre-LAM
proof and rightly expect that the laminated print will look like the
post-LAM proof.
But as Terry pointed out, the goal is usually to work with standard
reference profiles and not with custom ones. But there is no post-LAM
version of standard profiles, so you get a discrepancy between your
selfmade post-LAM proof and the standard proof which can ruin the job.
A solution would be to isolate the color shift and apply it to the
standard reference profile. This would give a consistent pair again.
I've tried various approaches since 2001. A first step used dot gain
adjustments and XYZ corrections due to more or less stray light (matte
or glossy), which accounts for the gamut changes and most of the color
shifts. See my paper at CGIV 2004 if you are interested,
http://www.imaging.org/store/epub.cfm?abstrid=30263
The paper also mentions that high-frequency screens and fine
stochastic screens are less affected, as Jon confirmed.
Nils, since you cited my paper, I'd be curious if you came to the same
or different conclusions in your discussion. Your figures for
additional dot gain for matte and glossy finish seem to confirm my
results. In your summary you state that hue changes are negligible.
This is only true for wedges, but overprints with one color high and
another low will change a lot. See your diagram 6, the yellow-green
and orange arrows.
Actually, I've also tried abstract profiles or directly adding the Lab
differences from a custom post-LAM minus pre-LAM to the standard Lab
values, which is more accurate but needs some degree of averaging and
smoothing. Since this works only for process colors, I'm more
interested in a model which can be applied to spot colors, too (like
the CGIV 2004 one).
After a couple of test prints of various substrates with AM and FM
screening, and hundreds of pre/post measurements that have been done
by an ECI working group, I've expanded the empirical model above by
second-order effects, and distributed beta versions of ISOcoated v2
matte and glossy profiles for internal testing (proofing, but also
separation). However, feedback was scarce, so the profiles are still
not officially released.
In my current job, lamination is not a priority, so everything has
slowed down. But if there's enough demand from you all to try beta
profiles, and if you promise to send feedback, I can send the
ISOcoated v2 matte and glossy versions, and I'll surely find the time
to build predicted lamination data sets from GRACoL_coated1 (and maybe
someone at X-Rite would be willing to create reference profiles from
them with the same software and settings as the unlaminated GRACoL
coated1 v2).
Best regards,
Hanno
--
Dr. Hanno Hoffstadt
Color Scientist
------------------------------------------------------------------------
Kirchstrasse 5
D-89081 Ulm
Germany
phone: +49(0)731 38900957
GMG GmbH & Co. KG
Moempelgarder Weg 10
72072 Tuebingen
Germany
Phone: +49 7071 93874-0
Fax: +49 7071 93874-22
mail to: email@hidden
www.gmgcolor.com
------------------------------------------------------------------------
GMG is a leading developer of high-end color management solutions
with more than 9,000 systems in use across the industry spectrum by
ad agencies, prepress houses, offset printers and international
gravure printers.
GMG GmbH & Co. KG. Registered Office: Tuebingen.
Court of Registration: Stuttgart Local Court, No. HRA 381463
General Partner: GMG Verwaltung GmbH. Registered Office: Tuebingen.
Court of Registration: Stuttgart Local Court, No. HRB 382505
Managing Directors: Joerg Weihing, Robert Weihing, Paul Willems.
Date: Mon, 3 Aug 2009 23:01:42 +0200
From: Nils Johansson <email@hidden>
Subject: RE: Laminate profile - Abstract profile?
To: <email@hidden>, <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset="Windows-1252"
Hi
again,
I don't believe that it will be necessary to use the press' ICC-
profile as an
intermediate conversion step I just think that it could reduce the
picture's
gamut in an undesired way, preventing some colours from being
reproduced
correctly in the lamination gamut. Instead I think the best idea
would be to convert directly to the
laminate-profile, trying to keep the original picture's colours
unchanged as
far as possible (our results showed that it appears that some parts
of the
gamut actually gets slightly widened post (glossy) lamination, when
viewed in
the a*b*-plane see diagram on p. 17 in the report).
Concerning delta Eab, it appears to be slightly more complicated.
Let's say
that the press operator gets a proof which he tries to match take
for
instance that the actual colour should be patch nr 7 but he matches
it to patch
nr 8 in the report (see Table on page 36, where the left columns
show the
L*a*b*-values pre matte lamination, and the right columns show the
values post
matte lamination).
Prior to lamination this is a colour difference of about 28 delta
Eab, but
after lamination the difference between two such laminated patches
would be
about 17. And when comparing patches 23 and 28 on that same page,
it's about
6,5 before and 4,0 delta Eab after lamination...
This would actually mean that the colour difference should be
smaller than 3
delta Eab in your case (if the difference is also in the chroma)...
But we have
not discussed this further in the report, and I actually don't know
if this is
a general result.
I still think that the best match would be obtained if using as an
accurate
press-profile and as a stable process as possible. Please let me
know if your
plain, vanilla profile works for you :)
Another way of possibly simulating the colour changes would be to
use a special
transformation which translated the unlaminated L*a*b*-values to
laminated
L*a*b*-values. I don't know if there is a good way of doing this.
Perhaps one
could make an Abstract profile (as specified by ICC - http://www.color.org/faqs.xalter#p5)
performing this? (I've never used one, or really know exactly what
it's capable
of doing) :S Anyone out here knowing if this would be possible?
Also, I don't know how spot colours are affected. As it said in the
report, lamination
seems to cause an optical dot gain. Perhaps the fact that one is using
half-tone dots is the reason for these colour changes (however not
the only
reason, as also the solid colours (where no dot gain can occur) are
affected).
If there is anyone else who would like to share their experiences with
simulating lamination, please post a message :)
Nils
From: email@hidden
To: email@hidden; email@hidden
Subject: Laminate profile
Date: Mon, 3 Aug 2009 12:39:52 -0400
Thank you Nils Johansson,
I think I follow your reasoning. The only step I was forgetting
about, which
you so kindly suggested, was to make a proof of how ugly the original
client's separations would look, once laminated.
So, your steps are:
1) Make an ICC-profile based upon a colour chart not being laminated
(this would be the printing press' ICC-profile)
I wonder to what extent I am shooting myself in the foot by *not*
using the
custom press profile as opposed to plain, vanilla GRACoL2006_C1
profile?
There is an average 3.5 deltaEab difference between the actual
press profile
and GRACoL2006_C1. Maybe I could make a first conversion from
GRACoL2006_C1
to myCustomPress profile, and then make a conversion from
myCustomPress
profile to myLaminateProfile?
2) Laminate the test chart from step 1 and make another ICC-
profile (we
call this the laminate-profile).
Got that step right.
3) Use the laminate-profile as a destination (target) profile when
converting your pictures and when producing your plates.
Yes, this make sense as this is the final color space into which the
original imagery will be viewed in.
Perhaps this sounds
strange, but
this should produce the best colour match between the digital colour
data and
the final, laminated result. Converting to this ICC-profile should
eliminate
the change in colour when the prints are laminated.
I agree there should be a close match between the client's original
and the
final laminated results. And it's also good to convince us that the
whole
procedure will actually produced the expected results.
4) However, to show the printer operator which colours he/she should
aim for,
you cannot give him/her a proof based upon the laminated-profile
as the
produced prints have not undergone any lamination yet.
Right. That wouldn't be smart.
Instead you must print a proof where you show how the colours should
look when
unlaminated. This you do by assigning the press' ICC-profile to your
already separated pictures in step 3, and then make a proof.
You seem, this traces back to my first comment, above. If I proof
using the
GRACoL2006_C1 profile then, knowing that the press is already 3.5
deltaE off
GRACoL2006_C1, maybe once laminated, the sheets will be off my 3.5
deltaE
also?
What do you think?
Similarly, if you want to simulate how the laminated result would be
when using
the press' profile as the destination profile (which I believe most
printing
houses do), you should assign the laminate-profile to your
separation
and proof that.
Yes, that will show what kind of color shift will occur as a
consequence of
lamination.
I hope you could follow my reasoning. As I said, we haven't had the
opportunity
to test this in practice, but we believe this should work pretty
well,
at least
theoretically, for CMYK-data.
Yes, I concur with your reasoning.
If you want, you can download our report as a draft from
http://www.hej.st/Laminate_draft.pdf. It is written in Swedish,
but has
an
English summary. I will try to post an English translation in a very
near future.
I still need to brush up on my Swedish :-)
Try to contact the library at Dalarna University (http://
www.du.se) and
ask for the report, if the link doesn't work.
OK. I won't hesitate.
Nils Johansson
Student at Graphic Technology Dept., Dalarna University, Sweden
Thank's / Roger
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden