G7 Calirbation and Color Management NOT WORKING!

Hello All -

I'm currently running several Epson printers on EFI Fiery XF 4.5.8 (I've paid for the upgrade to V5.0 - just waiting for my EAC code to upgrade). I'm experiencing some major color issues and I can't explain/figure out why. I'm hoping someone here can help solve this problem for me.

In EFI I've set up workflows as follows:

1) Created an output device for one of our S70670 Epson 64" Solvent 10 color printers.
2) Using Color Manager created a Baselin (.epl) file using the EFI Color Manager recommendations for ink limit per channel and total ink limit (this is a CT driver).
3) Hijacked the lin and lin-output device and applied the .epl file I created to that lin/output device that was hijacked
4) Without color management on, I've printed P2P targets, measured, imported measurements into Curves 3 and created a vcc file back in EFI Color Manager to reach G7 calibration.
5) Using the .vcc and .epl file (still without color management) I've printed an IT8.7/4 Random test charge and measured/created profiles (tried both my I1Profiler and the built in profiling software from EFI).
6) Connected the ICC profiles to the EPL files.
7) Turned on the Color Management and sent color settings as follows:

Honor embedded profiles when available (checked - because our art department embeds color profiles into our print files that they want us to use).

CMYK Source: SWOP 2006 3V2
Rendering Intent: Relative Color Metric w/ BPC

RGB Source: Adobe RGB 1998
Rendering Intent: Relative Color Metric w/ BPC

Gray Source: None
Rendering Intent: Relative Color Metric w/ BPC

Multicolor Source: None
Rendering Intent: Relative Color Metric w/ BPC

Use PDF Output intent if available (checked - because again we embed ICC profiles into our print files)
Simulation Profile: SWOP 2006 3v2
Rendering Intent: Relative Color Metric w/ BPC

We print from PDFS that are saved to the 2008 PDF/X-1A standard in Illustrator with output intent set to embed SWOP US Web Coated v2


When I print files I see colors don't match our color calibrated monitors, and customers complain colors are not accurate (which I have to agree with).

We measure the Delta E and see we have some major Delta-e values on colors.

Also, we do a similar process on our older GS6000s and when we print the same file on the different printers, we get different color outputs (even though all the colors in the files are within both print gamuts - SWOP US Web coated v2


Any insight on to what I could be doing wrong would be greatly appreciated!!!

-DS Matt

rich apollo

Well-known member
Why are you trying to interrupt the calibration process? Are you dis-satisfied with the color going through the built-in calibration/linearization process? It's my opinion that you're shooting yourself in the foot - unless I'm missing something.

Go through the built in calibration, set up your color management in the RIP, and check your result.

It might be worth noting that the source profile being set up in the RIP and the output intent being embedded do not reference the same colorspace.

You say you, "...measure the Delta E and see we have some major Delta-e values on colors." Can you tell us what you're measuring, and what you're comparing to?

We should also look at how the soft proofing is set up.
Hi Rich -

Thanks for the response.

The main reason I've added in the G7 Calibration is because I'm not happy with the results we get with just following the calibration/linearization process from EFI. I attended the G7 Expert training with Ideailliance in March and became certified as a G7 Expert in April. I beleive the G7 Training and calibration should work, so I feel like I'm missing something.

I've since gone back and created an ICC profile and then optimized the profile inside EFI with their built in Profile Optimization tool, which has provided me with a profile that:

1) Covers 97.6 % of SWOP 2006 3 v2
2) Has an Average Delta-e of under one when printed the it8.7/4 Random Reference chart and measuring the values

Those two points I'm happy with (because this is a low resolution high speed production profile I'm not surprised it doesn't cover 100% of SWOP 2006), if I had the luxury of running at a slower speed/higher resolution I think I'd get a larger gamut profile that would possibly cover 100% of SWOP 2006 3 v2.

3) Has an max Delta-e of 9.

The third point is the only issue I have, I'd like to see a max delta-e lower than that, but I'm not sure if I'm just setting my expectations to high, or if that is indeed higher than it should be.


In regards to mismatched source/embedded profiles, I know they are different profiles, but thanks to EFI NOT OFFERING/INCLUDING SWOP US Web Coated in their RIP software, and Adobe making the default for their software (which almost all my customers end up just using) SWOP US Web Coated, I can't make them be the same without chaing my Adobe PDF saving settings (which I could, but their really isn't much benefit to it because the EFI RIP uses the Adobe PPE to process PDFS so it does the same thing at the RIP that Illustrator would do, which is properly convert from SWOP US Web Coated to SWOP2006, changing the CMYK values and preserving color appearance (as long as the color falls in both gamuts, otherwise it would remap the color to an in gamut color using the Relative Color Metric rendering intent - which is just a fact of life, I can't fight gamut compression).

So I don't believe that is the cause of the problem.

As for my soft proofing, we have 4 top of the line Eizo monitors that are calibrated weekly with the Eizo propriety software and I1Pros (which I actually feel is too often, and so do some of the Eizo support personal I've talked to - monitors don't fall out of calibration that fast), we then also have a GTI light booth that has light blinders on it and is neutral gray for a background. We proof under daylight (5000K) light bulbs and monitor the lifetime of the bulbs. So I'm fairly confident my proofing set up is second to none (but please let me know If I'm wrong, we want to make sure we have the best set up we can for color accuracy).

I hope that provides you with the information you need to possibly identify where I'm going wrong! I know that Don Hutchinson doesn't feel EFI really buys into the G7 calibration process, but I think that using the Vcc (visual correction curve files) to apply the curves in Curve 3 is the proper method for G7 calibration on the EFI Rip (which we really do like - we also own Caldera V9.1, ColorGate Production Server 7, WaSatch, and ColorBurst).

Thanks for the help!


Correct Color

Well-known member
I beleive the G7 Training and calibration should work, so I feel like I'm missing something.

Well, unfortunately what you're missing is that no, it won't work.

The simple fact is that G7 is just about useless in large format printing.

If you think about the path you're following, what you're doing is replacing the EFI linearization with G7. And that's all you're doing. I myself am not a great fan of EFI, and particularly not of the XF RIP, but its linearization routine is adequate. You're not gaining anything by substituting G7 for it. All that is is just a different type of calibration routine.

And in an ICC workflow, linearization does not tell the printer what dots to produce, an ICC profile does.

An ICC profile reads an input file, pixel by pixel, for L*a*b* values, and then looks for those L*a*b* values in whatever destination color space it is. How well the profile actually describes how the printer prints determines both how much of the machine's capabilities you use, and how closely how it prints matches what was input. And you could say that you make a more accurate profile with a G7 linearization...but honestly you won't, and keep in mind that even if you did, where you would see a difference is in neutrals only. If your issues are anywhere else, G7 is not going to help you in any environment.

Main thing to remember is that once you've made an ICC profile, you're using that profile. You're not using G7 anymore.

And that's true with any RIP.

My guess is that somewhere in your attempt to implement G7, you've broken the machine-state pathway in your RIP, and that's what's creating your problem.

Frankly, it sounds like you're attempting to apply litho answers to a large-format environment, and in the doing, you're really knee-capping your device capabilities. There is absolutely no reason that under any conditions other than a very soft material, that you should not be able to get well over 100% coverage of SWOP out of that machine, with that inkset.

If you're interested, I can help you get every bit of capability out of each of those machines, on every media. First time; every time.

It's what I do for a living.

Mike Adams
Correct Color
Last edited:

Deleted member 16349


I listened to some of the webinar and I have to say that this just adds to my belief that this industry is clueless in what needs to be done to solve colour management problems.

Near the end of the process control presentation (~41:50), they ask the printer who is using this system, how many reprints does he do to satisfy customer complaints. He replies about 35 to 40%. Duh.

The problem is that there needs to be Capability in the process at the press and in prepress. This is not the situation now. Capable processes that don't need consultants and that are general solutions and not Standard specific. Of course this won't happen.

What About Profitability?

What about Profitability?
Offset yields new advantages

Read All About It