ICC Profiles, and Press Scenarios, and Color Accurate Proofs... Oh, my!

Red_Right_Arm

Well-known member
Okay, so I'm about 95% of the way there on my own. I think I'm just missing one aspect to color management...

- Following FIRST specs, we eventually get to the press characterization trials (the stage where you produce the crazy multi-colored test chart).
- We print one of these color test charts every time the press variables change.
- We end up with color test charts that represent the process colors for each combination of printing specs (each ink, on each material, on each press, etc...)
- We scan the color test chart into the RIP and create ICC profiles for each of these printing scenarios.
- Now we can see if using Press A, with Material B, with Inks C will have out of gamut colors. Let's say 80% color gamut, for example.
- We can also see if using Press C, with Material A, with Inks F will have 95% color gamut, for example.
- We can also see if using Press B, with Material C, with Inks A will have 75% color gamut, for example.
- Basically, we end up seeing how much of the color gamut we can hit with each different combination that we test.

... Everybody with me so far? Good...

- The proof printer only ever uses one material, with one ink, and is only this one printer.
- So we can do the color test chart on this proof printer.
- We can measure the color test chart from this proof printer.
- We can create the ICC profile for this proof printer.
- We can find out the color gamut of this proof printer. Let's say 99% color gamut, for example.

... Here's where I get stuck...

- How do I get the proof printer to accurately represent what will happen on a certain press with a certain scenario?
- If the ICC profile controlling the proof printer is created by measuring what the proof printer does when it prints on its own material and inks, that ICC profile is only relevant to that one printer.
- If the other ICC profiles are created by measuring what happened on each press with their own materials and inks, those ICC profiles are only relevant to those presses.
- If I go to make a proof for Press A and its printing scenario, and I plug into the proof printer the ICC profile for that printing scenario, isn't the ICC profile making adjustments incorrectly on the proof printer?

... Basically what I'm getting at here is, if I find out my proof printer can print colors well beyond the color gamut of a press, how do I accurately restrict the proof printer to mimic what can be expected from the press?
 
A good proofer with a wide enough gamut can simulate other printing presses. But in order to do that, it must know what exactly to simulate. i.e., it must know how the simulated device produces color. Therefore, you must have ICC profiles for both sides of the proof making - the proofer and the press.

Think of the whole thing as languages instead of color: you want to translate the meaning of a sentence in language B (the press), to language A (the proofer). They use different words, and different grammar, but the meaning will stay the same. But you must have the dictionary (ICC profile) for both languages in order to translate the sentence.

So, in your proofer RIP, you should have an option of simulating another device. Plug the profile of the press into that option, use "absolute colorimetric" rendering intent, and you should get a color proof representing the press.
I don't know if I managed to explain properly. Sorry, it's pretty late here and I need to sleep...
 
So, in your proofer RIP, you should have an option of simulating another device. Plug the profile of the press into that option, use "absolute colorimetric" rendering intent, and you should get a color proof representing the press.

That's certainly a direction worth looking into. Thanks.
 
Just to note:

it's not "...certainly a direction worth looking into."

It's how it's done. If you look into any other direction but this one, you're wasting your time.


Mike Adams
Correct Color

What I mean is, it's worth seeing if the RIP we have can do this. It's old. It lacks some common functions. I can't say off the bat if it is designed to do conversions the way Schnitzel suggests.
 
I am not sure if you have a RIP or software already but some proofers come with free software to create proof profiles. Check to see if your proofer provides this. If not there are options to buy a RIP or profile (curve) software for your proofer so you can create these proof profiles. The way I made my proof profiles is by making plates of a color proof chart (provided by my proof software) using the correct press curve and printing these charts using the correct press, ink, stock, aniloxes, densities, etc. for that press curve. Then I take these printed charts and read them in (using x-rite i1io) and then create the proof profiles from these readings using my proofer curve software. I am usinf Esko software for my proofer but it is quite expensive but it works well to represent what I get on press.
 
What I mean is, it's worth seeing if the RIP we have can do this. It's old. It lacks some common functions. I can't say off the bat if it is designed to do conversions the way Schnitzel suggests.

I'd wager it does. Almost all RIP's do. But if it doesn't, you can convert your files to whatever print space you're going to use (duplicates, of course, just for proofing) in Photoshop at Absolute Colorimetric rendering intent, print the result and accomplish pretty much the same thing.


Mike Adams
Correct Color
 
What I mean is, it's worth seeing if the RIP we have can do this. It's old. It lacks some common functions. I can't say off the bat if it is designed to do conversions the way Schnitzel suggests.

While there are "work-arounds" as Mike Adams suggested, yeah, the proofing RIP needs to KNOW several things.

Like the color of the paper...

YEARS ago, i used to share this very contrived yet simple example.

https://docs.google.com/document/d/1QaHsfs-YHl-NuHBBA-KeedBWU44ibsjV8lLiN9WzI8E/edit?usp=sharing

this is VERY OLD and NONE OF THE LINKS to any of the assets work, but you will get the idea

PROBLEM - I have a PDF and I want to print it on Green paper, I have a proofing system but no green paper to proof on, what do i do ?

APPROACH - add an OUTPUT INTENT icc profile to the PDF, and the RIP will then "simulate" that green paper by adding cyan and yellow to the PDF objects when rendered.

Hope this helps in some small way. PDF 2.0 will help when they support CxF better
 
We found out the best way to suit and match our needs was to create a device-link profile for the proofer and printers. We also use the same paper to proof on that we print on, but agree with Michael Ejahn "Output Intent" works better for color stock.
 
Okay, so the RIP does have the ability to assign ICC profiles from the presses into the proofer. But now I have another question. Once again, following FIRST specs...
  • You run your optimization trials... before color management.
  • You eventually figure out the best settings/parameters on the press to give you whatever you'd call the best results... before color management.
  • You document these settings/parameters... before color management.
  • You do a press fingerprint using these same settings/parameters... before color management.
  • The press fingerprint produces print samples that are used as the standard for how the press is expected to perform... before color management.
  • It's got your gray balances... before color management.
  • It's got your overprinting... before color management.
  • It's got your ink densities... before color management.
  • It's got your gradients... before color management.
  • It's got your flat color tints... before color management.
  • Basically, it has everything you'd need to determine what this press is supposed to do when the same settings/parameters are used again... before color management.
Alright, so now we do all of that and we say, "I deem thee press fit for thy color management." Then we...
  • Run the press under the same settings/parameters as the fingerprint... before color management.
  • This time we are printing the standard IT8.7/4 characterization target. We print this color target... before color management.
  • Now we measure the IT8.7/4 characterization target that came off press. The press that was under control. Because we followed the same settings/parameters as the fingerprint... before color management.
  • The measured IT8.7/4 characterization target gives us the ICC profile for this press under these specific conditions... before color management.
Okay, right here is where I get screwed...
  • We use the ICC profile to correct for color on future jobs... now with color management.
  • This ICC profile adjusts and balances colors. It alters them... now with color management.
  • So if I have a new job, and I send it to the RIP, my RIP will make adjustments to the plates to control the colors... now with color management.
  • These adjustments will be applied to everything. Including the color control blocks we use in the bearer bars... now with color management.
  • The plates are created with these color adjustments... now with color management.
  • The plates are put on press... now with color management.
  • The pressmen run the press under the same settings/parameters as the fingerprint, because the fingerprint was deemed the standard to measure by... before color management.
Won't their measurements be off now? Gray balance, dot gains, tint percentages, everything will be effected. And thus, different from the standard.

Say the ICC profile corrects 70% magenta to be 60% magenta to compensate for dot gain. The dot gain we know to be standard for 70% might be, let's say, 80% when you measure the fingerprint sample... from before color management. Because the fingerprint sample was made before we used the ICC profile, it has that 80% measurement due to dot gain. But now the ICC profile is compensating for the dot gain and pulling it back to 60%.

You see what I'm saying? The standard was before color management. Now with color management a change has been made. Whenever something changes which effects colors we need to re-optimize, re-fingerprint, and re-characterize the press. Which will result in another ICC profile. Which will make another change. Which means another re-optimize, re-fingerprint, and re-characterize the press.

What am I missing? Where did I go wrong? How do you get Jane to stop this crazy thing?

P.S. – I looked into seeing if our RIP can produce control targets linearly, while making adjustments to the layout to be plated. It can't. It can only do one or the other. And since the RIP generates all of our control targets at the same time that it processes the layout, that means it either adjusts everything or we don't use it and it adjusts nothing.

P.P.S. – I value all of the feedback you've given me.
 
First of all, tell us where and how you are using the ICC profile. It appears you are using it in your RIP to convert incoming files from their source profile to the press profile. In doing this, you are attempting to match the color (lab) of the source using device counts that reproduce that color on your press.

I.e. If your incoming file is tagged with the GRACoL2006 profile and there is a patch of 70% magenta (with an associated color, or L*a*b* value), the profile is converting that 70% magenta to 60% magenta because 60% magenta is what the ICC profile deems the best color match (closest L*a*b*) on your press

Other ways to use a Custom Press Profile:

Design - Set the CMYK profile in the Color Settings of your design application software to your custom press profile. Now, files will be tagged with your custom press profile and no conversions will be required at the RIP.

Custom Proofing - Set the source and simulation profile in your proofing rip to your custom press profile. This would produce proofs that simulate what will happen if you send a CMYK image to press and do not do any conversions at the RIP.

Softproofing - "View - Proof Setup" in Photoshop and set your "Device to Simulate" to your custom press profile.

DeviceLink conversions and Ink Saving - If you are attempting to match to a standard, a DeviceLink profile is a better way to go than a straight conversion. You would create an additional ICC profile (a DeviceLink ICC profile) that captures the relationship between a source and a destination profile (i.e. GRACoL2006 and your press profile). A DeviceLink profile gives you more control over the parameters of the conversion (i.e. you can choose to exclude primaries from conversion, so that your 70% magenta will stay 70%). Many of these parameters are necessary to control, as a straight conversion will introduce printability issues on press (i.e. introducing 1% of an ink into a color because it theoretically provides a better color match). You can also incorporate GCR (Gray Component Removal) into your DeviceLink to save ink (one of the biggest benefits of DeviceLink).

Before you can implement the profile in any of the ways I just mentioned, you need to ask yourself what you are looking to achieve with Color Management. One of the first things you should ask yourself is whether you want a standards based workflow. If so, which standards/specifications are you looking to meet?
 

PressWise

A 30-day Fix for Managed Chaos

As any print professional knows, printing can be managed chaos. Software that solves multiple problems and provides measurable and monetizable value has a direct impact on the bottom-line.

“We reduced order entry costs by about 40%.” Significant savings in a shop that turns about 500 jobs a month.


Learn how…….

   
Back
Top