G7 Proof to Press Match

[QualityPrint]

How do others rate their proof to press visual match when profiling with the G7 methodology? Other employees say I am too picky and want a perfect match that is unobtainable. I have been trying to get a press sheet and proof match from another printer for years now but it seems like other printing companies don't want to share that info. The match I am talking about is from the actual press sheet that is used to profile and the proof which is made from that sheet using G7 methodology. We did a profile today and after 5 people rated the match it averaged out to an 8. (On a scale from 1 to 10... 1 bad & 10 perfect) I rated it a 7 because flesh tones along with the greys looked a bit warm. Warm enough to where if I was on press I would have to pull back the magenta before running the job. Any info is appreciated

 

[gordo]

How do others rate their proof to press visual match when profiling with the G7 methodology? Other employees say I am too picky and want a perfect match that is unobtainable. I have been trying to get a press sheet and proof match from another printer for years now but it seems like other printing companies don't want to share that info. The match I am talking about is from the actual press sheet that is used to profile and the proof which is made from that sheet using G7 methodology. We did a profile today and after 5 people rated the match it averaged out to an 8. (On a scale from 1 to 10... 1 bad & 10 perfect) I rated it a 7 because flesh tones along with the greys looked a bit warm. Warm enough to where if I was on press I would have to pull back the magenta before running the job. Any info is appreciated

Just for clarity...
It is typically press to proof match (i.e. the proof represents the target color which the press aligns with) rather than proof to press visual match.
G7 is a definition of grayscale appearance, and a calibration method for adjusting any CMYK imaging device to simulate the G7 grayscale definition. The G7 method is for targeting NPDC (Neutral Print Density tone curve) but not for specific colorimetry. So the G7 method is not for "matching" color between press and proof.

To your points...

Having a group of people visually rating what is considered a "match" between proof and presswork is problematic due to the vagaries of individual color perception. If you measure the color you'll have a better basis upon which to decide whether the match is adequate or not.

Keep in mind that the only true proof of presswork is a press proof. All non press proofs are to some degree compromised - if only because of the difference in how color is created in an off-press proof compared with how color is created on press. So yes, you may be too picky and want a perfect match that is unobtainable.

Once upon a time IDEAlliance used to sell press sheets printed to GRACoL 7 which could be used to compare with proofs also made to GRACol 7. These no longer appear to be available which should indicate just how difficult a print to proof "match" can be.

best, gordo

 

[TerryWyse]

How do others rate their proof to press visual match when profiling with the G7 methodology? Other employees say I am too picky and want a perfect match that is unobtainable. I have been trying to get a press sheet and proof match from another printer for years now but it seems like other printing companies don't want to share that info. The match I am talking about is from the actual press sheet that is used to profile and the proof which is made from that sheet using G7 methodology. We did a profile today and after 5 people rated the match it averaged out to an 8. (On a scale from 1 to 10... 1 bad & 10 perfect) I rated it a 7 because flesh tones along with the greys looked a bit warm. Warm enough to where if I was on press I would have to pull back the magenta before running the job. Any info is appreciated

As gordo alluded to, G7 by itself is not for press-to-proof matching....for that, G7 must be tied to specific *colorimetry* such as GRACoL2006_Coated1 or SWOP2006_Coated3 (for example). I suspect what you're talking about is a GRACoL proof (made using either the GRACoL ICC profile or data set) and comparing that to a press *calibrated* using the G7 methodology with GRACoL *targeted* colorimetry. (I hate to be so specific on the terminology but one must be specific so we understand what we're talking about here).

As far achieving a match, sure, it depends on how closely the press in question was able to "hit" the G7 gray balance and the target (GRACoL?) colorimetry....proof to press match will vary depending on how successful the press was at meeting this criteria. In my view, this is still a better approach than going down the road of actually profiling a press and using that in proofing.....now not only do your presses not match each other, any of these G7 or "GRACoL" proofs won't match either! Better to have a standard proof and work to improve the press matching that proof than having all manner of custom profiled presses and proofs running around out there, all calling themselves G7 or GRACoL proofs!

Terry

 

[QualityPrint]

Yes, sorry for the incorrect terminology. Our press sheets colorimetrically are pretty good. We have had to resort to making proofs based on the press sheet and not G7 in the past but in this instance we are talking about the proof was made both ways and looked fairly similar but still not the same as the press sheet. I was told that through Oris the goal is to achieve a match on 80% of the sheet and if you want better than that you have to use devicelink profiles. Is this accurate and if so how well to devicelink profiles work and how hard are they to do? Maybe that 20% that we aren't matching is standing out in my mind... Oris only tells us what the average delta E is on out IT8 targets and the highest delta E. How do you pinpoint those high delta E's and deal with them? Probably another costly software...

 

[TerryWyse]

I've done it both ways at several companies......I would say the "typical" match you'll get with G7 plate curves alone is in the 2.5-3 delta e range to the GRACoL data set....device link profiles will typically cut that number in about half in my experience. It wasn't uncommon to get a 1-1.5 dE match to GRACoL using a combination of G7 curves + device links or even device link profiles made from an uncalibrated (linear) press profile. Besides getting better press stability and a modest amount of ink savings using device link profiles, you could expect a very good match to a GRACoL proof.

Terry

 

[Prepper]

Just out of curiosity, what screening method are you using AM or FM?

Also, in our shop, the print manager and I (prepress) see press sheets and perceive that the yellow needs to go down to match the proof better, the pressman and our designer look at the same the same thing and see it as too much magenta and adjust it by lowering that. Color is subjective and individuals see differently. Measurement will tell you whether you're meeting the standard and it's tolerances or not.

There's a paragraph in the G7 Methodology that gives me both pain and comfort, pain because I too want a perfect match and comfort in that there never is although it's really close most of the tilme, it states something close to this: Even the best press operated by the most experienced workers with ideal conditions can vary quite a bit from run to run, because of even small changes in material, temp, humidity, etc. The chance of a 'perfect' press sheet from an offset press run are very small.

This is what I find when we print profiling targets, very few sheets will have more than 2 good targets on them (out of 4) and many only 1, when I 'm gathering at least 10 to measure and average, with a tolerance of plus or minus 5 and after running 1,000 sheets. So individual press sheets will have some inherent variation from tolerance, which is +/- 10 in production, in some area of the sheet and whch will affect some colors more than others depending on where and what it is.

Color Management does an amazing job of getting us close but when you consider all the variables involved its amazing to me that it matches as well as it does. Of course we want a perfect match but we also have to have those realistic expectations.

Terry
(not the wyse consultant)

 

[QualityPrint]

Just out of curiosity, what screening method are you using AM or FM?

We use both AM and FM. After hearing your replys, I feel that we are getting the best match that we can expect without using devicelink profiles. I will have to look into how do use them because I have no clue. I wonder if it is included in the Oris colortuner software?

 

[rich apollo]

Here is a look at some data I've generated, and what some other folks have been kind enough to share with me. What you're looking at is measurement data from G7 "calibration" sheets being compared back to the GRACoL_2006 data in dE(ab).

 

[Mike Strickler]

Making a custom press profile of any utility is not a trivial exercise; it requires averaging many press runs and "smoothing" the numbers to suppress anomalous data, make the profile more generally useful, etc. This is exactly what what was done to create "standard" profiles such as GRACoL, and one should have a very strong reason to reinvent this particular wheel. If you run with standard inks on a good quality sheet with AM screening you should run "to the numbers" for GRACoL and aim your proof there as well. Then you can make small adjustments on press to get a better match to the proof. Printing to specifications is not only more straightforward but also makes it easier to replicate the result, on your press, and on others. You notice the consensus of the three consultants here?

 

[QualityPrint]

What if our G7 proof seems to have a warm feel to it? Should we change our press sheet to match that even though we feel it is too warm? Or should we tweak the proof? I would love to see other G7 proofs from other printers as i'm sure they will vary greatly based on how wide the tolerances are.

 

[gordo]

What if our G7 proof seems to have a warm feel to it? Should we change our press sheet to match that even though we feel it is too warm? Or should we tweak the proof? I would love to see other G7 proofs from other printers as i'm sure they will vary greatly based on how wide the tolerances are.

G7 is a grey balancing method not a color target.
If your proof appears to "have a warm feel to it" it may be that your proof is not actually grey balanced - i.e. problems with how the G7 process was conducted. Or the color shift could be caused by the lighting conditions under which you are viewing the proof that is causing a color shift.
The spectral composition of press inks are very different than those of a proof. They will likely only form a metameric pair under D5000 lighting. That's what you should be aiming for. Also, compare the OBA content of your proofing paper vs your press paper as a difference in OBA content can cause a color shift difference in the 0% to about 50% tone areas.

Perhaps one of the G7'd shops here or one of the color consultants would be willing to sell you a proof that conforms to GRACoL7.

Normally the proof (that conforms to your industry target) is the "stake in the ground" and you tweak the press condition to align to the proof.

best, gordo

 

[QualityPrint]

But there is a tolerance for the proof as well... right? I mean if we took proofs from 10 different printers that were using the G7 process would they all look and measure the same? Are they supposed to?

 

[gordo]

But there is a tolerance for the proof as well... right? I mean if we took proofs from 10 different printers that were using the G7 process would they all look and measure the same? Are they supposed to?

Yes there are tolerances that you need to be within for both proof and presswork. These are spelled out in the G7 "How to" as well as the ISO 12647-2 documentation.

best, gordo

 

[rich apollo]

What if our G7 proof seems to have a warm feel to it? Should we change our press sheet to match that even though we feel it is too warm? Or should we tweak the proof? I would love to see other G7 proofs from other printers as i'm sure they will vary greatly based on how wide the tolerances are.

Are you targeting the GRACoL colorspace, or one of the SWOP colorspaces. My answer will vary depending on your answer. G7 gray balance is weighted against the paper white.

 

[QualityPrint]

We are targeting the GRACoL colorspace...

 

[mglouis]

Terry,

I buy into the standard proof argument, however I have a tough time using them for Type 4 paper, proofing Fogra47, whether I use C & D curves or G7 on press. Our uncoated paper averages around a -7 b*.

How are you dealing with blue opaques on the market when targeting the latest Type 4 printing specifications and trying to use standard proofs? Do you tweak the press target, the proofing target, both or just leave them be and ship the proofs and press sheets with an apology on behalf of the current state of graphic arts?

Matt Louis

 

[rich apollo]

We are targeting the GRACoL colorspace...

Then match the proof.

I buy into the standard proof argument, however I have a tough time using them for Type 4 paper, proofing Fogra47, whether I use C & D curves or G7 on press. Our uncoated paper averages around a -7 b*.

How are you dealing with blue opaques on the market when targeting the latest Type 4 printing specifications and trying to use standard proofs? Do you tweak the press target, the proofing target, both or just leave them be and ship the proofs and press sheets with an apology on behalf of the current state of graphic arts?

I would alter the white point in the dataset and rebuild the profile. I'm doing that with several "standard" profiles and have had good luck.

 

[mglouis]

Hey Rich,

I have been setting press aimpoints on blue-shade uncoated to hit Fogra47 RGB overprints and editing the proofing reference profile so the white point matches the proofing substrate to eliminate paper simulation with absolute colorimetric. Press to proof results are usually ok with room for improvement. Proofing substrate is close to press substrate of course.

I just followed your advice and noticed the profiles from standard characterization data and edited whitepoint data produce profiles with the same gamut, however making a profile from standard data and then editing the whitepoint in the profile changes the gamut, skewing the outer boundaries towards the ab* values of the paper. (I'd post pics if I knew how) This skewing has been a love/hate relationship for me because it gets overdone in areas such as 100c100m where substrate doesn't and shouldn't take on the ab* paper cast values in proofing.

You have given me plenty of home work since these methods can be hybridized.

Thanks,
Matt Louis

 

[Mike Strickler]

Mr. Apollo is correct. Or to put it differently, if your paper is far from GRACoL spec in color all your colors will be skewed in that direction in proportion to their transparency. If your paper measures -7 for b* your jobs will never pass GRACoL specs no matter what you do. If you have to print on that paper but still want a G7 gray-balanced result you can verify you wil need to edit the white point of the GRACoL profile to match that of your paper. You'll of course want to rename it--for example, GRACoL_bluish.icc, or something similarly descriptive. This will then be your proofer reference profile, with of course absolute colorimetric rendering intent. This is still problematic in practice, if you're using an inkjet printer, as few proofing papers will have the combination of bluishness and brightness to match the white of a bright, blue sheet such as McCoy. With a laminate proofer you would want to use rel col, and probably use a DL to preserve blacks, solids, and pure tints. It's actually not at all a trivial exercise.