Dotgain compensation in ICC profiles

[jonasand]

I have a question for the color management specialists on the forum.
When I use ICC profiles for color seperation work, is there a dot gain compensation built into the profile or do I still need to use dot gain curves at the platemaking stage?
If I try to explain a little more what I´m talking about, then lets say I have a RGB file in Photoshop and I convert it to my press profile, send it to rip and make plate. Is it possible for me to run linear plates or do I have to use dotgain compensation curves in the plate making to compensate for the dot gain on press?

 

[meddington]

Re: Dotgain compensation in ICC profiles

>When I use ICC profiles for color seperation work, is there a dot gain compensation built into the profile or do I still need to use dot gain curves at the platemaking stage?

Potentially Both. It depends on who the press was profiled. The ICC profile contains the dot gain characteristics of the characterization data that created the profile (i.e. the conditions of the press at the time f profiling). For example...USWebCoated SWOP was created with measurement data from actual pressruns (TR001 data) and those pressruns had approximately 20% dot gain. If your press using linear plates yeild 15% dot gain, than it might be prudent to use a 5% bump curve to obtain the 20% gain that is anticipated with the US WebCoated SWOP profile.


>lets say I have a RGB file in Photoshop and I convert it to my press profile, send it to rip and make plate. Is it possible for me to run linear plates or do I have to use dotgain compensation curves in the plate making to compensate for the dot gain on press?


You have an ICC press profile that represent the conditions of the press at the time that the characterization data was created. This include plates curves, paper and ink, blankets, inherent dot gain, solid ink density, and any number of other characteristics. So to answer your question, if the press profile was created from measurement data that used linear plates in the calibration process, than you would use linear plates for separations created from that press profile.

 

[Luc St-Pierre]

Re: Dotgain compensation in ICC profiles

Just a suggestion.... I never convert files to press profiles for outputting purposes. Since ICC profiles are nothing else than look-up tables and they depend upon a given situation on a given day of press fingerprinting and carry also forever local press missprints or other press “incidents”, doing so you would create yourself a very narrow pathway to success on press. A press profile is very usefull to “see through the eye of a press” a totally well custom linearized, color balanced file at the proofing step. If the press characterization was printed with a dot gain compensation curve, you will need to re-apply the same when outputting films or plates, not for proofing. If the press characterization was printed linear, then the profile will show a smaller gamut and all your files proofed for that profile will need dot gain compensation within themselves and then be outputted for film or plate in a linear manner (no cutback).

 

[disbellj]

Re: Dotgain compensation in ICC profiles

Most of us get images and don't make them ourselves if we are prepress. Because TR001 profiles are the default CMYK profiles in U.S. applications, we tried to set dot gain on plate to match old film-based workflow (50 in file = 54 on plate), made a custom press profile to make proofs of the incoming SWOP CMYK (showing how that CMYK would print on our press), and have had success for years doing this. But really this is past thinking and the way of the future is standardized profiles.

Now GRACoL has made a specification of the international standard ISO 12647, called GRACoL2006_Coated1 (first GRACoL specification that is within ISO tolerances), made freely available characterization data, and official output profile GRACoL2006_Coated1v2 (made by X-Rite), also freely available.

What I would do (my plan to do eventually, but it's very close to what I do now except changing the CMYK profile from SWOP to GRACoL2006_Coated1v2) is to let the SWOP images come in, do not do color management incoming CMYK in Adobe or Quark, except for display of existing CMYK numbers in files (in Adobe I turn CMYK color management policy to Off), set the GRACoL2006_Coated1v2 profile as CMYK profile in color settings, and soft-proof on-screen what the existing SWOP will look like printed on GRACoL2006_Coated1v2 (SWOP looks "natural" printing on GRACoL2006_Coated1v2, although "specific" colors may need to be scanned with a spectro, Lab values put into Photoshop lab document, and converted to GRACoL2006_Coated1v2 to get new CMYK numbers to get appearance on press as close as possible). So that covers the "before rip" portion.

Then I rip my job and proof it on proofer with GRACoL2006_Coated1v2 as source (what I want to match) and custom Epson profile as destination (where it's actually printing), convert using Absolute Colorimetric Intent. Proof done.

Set up press using G7. Done

Don