What's happening with G7 Master printers

[tmiller_iluvprinting]

Gordon,
I apologize, I should have known, or stated if you have not attended....! We find the G7 Methodology spot-on. I have used G7 to calibrate UV Inks, stochastic screening, coated papers, uncoated papers, Heidelberg presses, Komori presses, UV Inks on plastic, and fade-resistant inks. I have found that G7 achieves the goal of colorimetrically arriving at a shared visual appearence between processes and platforms. It has reduced rework because of color issues and significantly reduced makeready time and materials. Has it solved all the issues in the pressroom? We both know better than that, but I don't think that is the goal of G7. G7 is not that complicated, and makes it easier to target and achieve a specification like SWOP or GRACol. Don does not tout G7 as a save-all end-all solution to everyday printing issues.
Best regards,
Todd

 

[gordo]

Gordon,
I apologize, I should have known, or stated if you have not attended....! We find the G7 Methodology spot-on. I have used G7 to calibrate UV Inks, stochastic screening, coated papers, uncoated papers, Heidelberg presses, Komori presses, UV Inks on plastic, and fade-resistant inks. I have found that G7 achieves the goal of colorimetrically arriving at a shared visual appearence between processes and platforms. It has reduced rework because of color issues and significantly reduced makeready time and materials. Has it solved all the issues in the pressroom? We both know better than that, but I don't think that is the goal of G7. G7 is not that complicated, and makes it easier to target and achieve a specification like SWOP or GRACol. Don does not tout G7 as a save-all end-all solution to everyday printing issues.
Best regards,
Todd

Don't apologize.

I don't disagree that the G7 method achieves grey balance. I do get confused when you write: "I have found that G7 achieves the goal of colorimetrically arriving at a shared visual appearence between processes and platforms." Since, I believe, that is NOT what the G7 method does or promises to do. You need color management on top of G7 to do that.

I have not seen any evidence to support the G7 claim that if all printing types shared a common
gray scale appearance would also have a similar visual “appearance.” (whatever "similar" means). In fact at the one committee pressrun (at ManRolland demo facilities) where G7 press sheets were available from a previous run (Integrity Graphics) the sheets did not have a similar appearance despite the insistence that they did. I would have liked to have seen this theory tested in an objective fashion.

If it has reduced rework because of color issues (which G7 does not address) - I'm guessing that it was because it brought some level of process control to your shop.

There are many other ways to accomplish the goal of calibrating for grey balance on press and proof. Creo/Kodak had tools to do this years before G7. I'm sure other vendors did as well. It would have been interesting to see if one of their tools would have been more effective or not required an onsite certified expert to accomplish the goal. (I know that the Creo/Kodak tool could be used by any printer, did not require an onsite "expert" and did not require them to be equipped with a degree in math or an excel spreadsheet).

Like I said, I agree with the goal, however the implementation was IMHO, poorly executed and as a result I think an important opportunity was missed and a great deal of confusion created. For one example, exploring other methodologies like Isometric curves which replace grey balance curves and which can provide a colormetric match.

best, gordon p

 

[martin.weberg]

exploring other methodologies like Isometric curves which replace grey balance curves and which can provide a colormetric match.

Gordo, do you mean isometric curves like those described in this study? http://www.doplganger.com/resources/TAGA/TAGA_2004.pdf

Or, can you point me to another reference?

Martin Weberg

 

[gordo]

Gordo, do you mean isometric curves like those described in this study? http://www.doplganger.com/resources/TAGA/TAGA_2004.pdf

Or, can you point me to another reference?

Martin Weberg

Yes. They are the only folks who, as far as I know, have formally studied ad published the notion of isometric curves for color calibration and building of tone curves. Other people, including myself, have looked at them in a more informal way. The industry has certain sacred axioms (sacred cows) - grey balance being one of them - that seem to prevent basic research in defining what exactly is the problem that is being addressed as well as alternative methods of solving the problem.

Gordon p

 

[gordo]

Gordo, do you mean isometric curves like those described in this study? http://www.doplganger.com/resources/TAGA/TAGA_2004.pdf

Or, can you point me to another reference?

Martin Weberg

Yes. They are the only folks who, as far as I know, have formally studied and published the notion of isometric curves for color calibration and building of tone curves. Other people, including myself, have looked at them in a more informal way. The industry has certain sacred axioms (sacred cows) - grey balance being one of them - that seem to prevent basic research in defining what exactly is the problem that is being addressed as well as alternative methods of solving the problem.

Gordon p

 

[tmiller_iluvprinting]

Gordon,
I understand that your experience with IDEAlliance and G7 in its infant stages left a bad taste in your mouth. If you are ever in De Pere, WI look me up, maybe I can show you the 'evidence' that you need.
Best regards,
Todd

 

[Deleted member 16349]

Yes. They are the only folks who, as far as I know, have formally studied and published the notion of isometric curves for color calibration and building of tone curves. Other people, including myself, have looked at them in a more informal way. The industry has certain sacred axioms (sacred cows) - grey balance being one of them - that seem to prevent basic research in defining what exactly is the problem that is being addressed as well as alternative methods of solving the problem.

Gordon p

Since the application of the CMYK ink channels in offset printing are non linear and non independent, I really have a hard time thinking that any set of curves will result in predictable colour.

Also some of the experience printers have with using curves and the technical papers tend to evaluate performance on experimental results and often without sufficient logical analysis. This means that results can be more based on anecdotal outcomes as opposed to proofs of the concept.

In the case where one has three curves (CMY) and the colorimetric result of the combination of these curves into one curve is a gray colour all along the curve, this still can not predict colour matching capabilities between different systems.

Let's say that one has two systems and there is a combined set of curves for both systems that measure the same gray at the same percents. The implication is that these two systems will have the same appearance but is this so.

One test that could be done is to only look at two of the three channels. Let's say C and M and not print the Y channel. Then compare the two systems to see if the CM combined curves for both systems produce the same colour.

I expect that they wouldn't match colour and therefore these curves are no guarantee that two systems will result in similar appearance in regions away from the gray curve.

If both systems were in actuality the same system, then one would expect that they would result in the same colour. IMO this is why there is such a push to standardize ink, etc. because the present methods will only work well if the systems are very similar.

If there is an attempt to standardize all materials just because one can not develop an effective method to reproduce colours, then that is an industry problem. In this case, standardization covers up the incapability of the developers to innovate general solutions to colour reproduction.

 

[gordo]

To Erik,

The isometric curves that Dopelganger uses are not based on grey balance. They are based on brown balance. William Birkett and Charles Spontelli have been using their method to do color alignment in printshops since at least 2001 when I met them and saw samples of their method in production presswork. So in that sense it is not an academic exercise.

I'm not smart enough to argue for or against their method. I only used it as an example of the fact that there were other ideas at the time that GRACol7/G7 was being developed that were not considered or tested because, IMHO, of the agenda of a small group within the committee.

In any case, like I said, it's all ink under the press now. No point in crying about it.

gordon p

 

[Deleted member 16349]

To Erik,

The isometric curves that Dopelganger uses are not based on grey balance. They are based on brown balance.

In any case, like I said, it's all ink under the press now. No point in crying about it.

gordon p

Gordon,

Brown balance? OK, same logic applies.

Your right. Past history that can not be easily changed. I was again tempted to comment but I shouldn't have.

 

[meddington]

Hi Erik

In the case where one has three curves (CMY) and the colorimetric result of the combination of these curves into one curve is a gray colour all along the curve, this still can not predict colour matching capabilities between different systems.

You are of course correct, and although I agree there is a misconception regarding this, it is not and has not been an official G7 claim that 1 dimensional curves can replace color management, or predict color on differing systems, and is stated as such on G7 documentation (Spec and How To). What G7 does do is effectively define gray balance and tonality unambiguously relative to substrate, and in similar systems can go a long way in adjusting colorspace on the whole (the latter an admittedly anecdotal comment based on my experience/opinion, but I put it in there anyway ).

If there is an attempt to standardize all materials just because one can not develop an effective method to reproduce colours, then that is an industry problem. In this case, standardization covers up the incapability of the developers to innovate general solutions to colour reproduction.

There certainly are methods that more accurately deal with color reproduction on the whole, such as device link profiles. There is debate on whether G7 would add benefit as a baseline calibration on which to build said profiles off of.

 

[meddington]

I only used it as an example of the fact that there were other ideas at the time that GRACol7/G7 was being developed that were not considered or tested because, IMHO, of the agenda of a small group within the committee.

You are likely correct, though I do recall Don Hutcheson often referencing the concept of isometric curves (ala William Birkett and Charles Spontelli) during that era...I think the concept was not fully understood (perhaps completely misunderstood), a bit too far from mainstream to be accepted or both.

 

[Deleted member 16349]

Hi Erik

You are of course correct, and although I agree there is a misconception regarding this, it is not and has not been an official G7 claim that 1 dimensional curves can replace color management, or predict color on differing systems, and is stated as such on G7 documentation (Spec and How To).

Yes, I know they make no claim to colour reproduction. They only claim that the gray curve is similar for different systems. In their literature, they are quite honest about that.

 

[gordo]

Yes, I know they make no claim to colour reproduction. They only claim that the gray curve is similar for different systems. In their literature, they are quite honest about that.

Unfortunately, that fact is very often ignored/missed by most of the printers who talk about their experience and results with G7, which makes me go completely.....oooops, calm down gordo.....think of Daffodils and bunny rabbits........

gordo

 

[Deleted member 16349]

Unfortunately, that fact is very often ignored/missed by most of the printers who talk about their experience and results with G7, which makes me go completely.....oooops, calm down gordo.....think of Daffodils and bunny rabbits........

gordo

...and print buyers. I doubt print buyers have read that part of the spec.

Let's see. G7 insures that only the three colour gray (CMY) matches but we have removed most of the three colour gray with GCR. Hmmm. That does not make a print buyer wonder? ....bunny wabbits.....

 

[gordo]

Let's see. G7 insures that only the three colour gray (CMY) matches but we have removed most of the three colour gray with GCR. Hmmm.

Exactly. There is a great deal of this strangeness in this industry. I had hoped that the CRACoL initiative would have been a great opportunity to test some of the concepts underpinning the specifications before plunging straight into methodologies and specifications. Instead, the Print Properties Group came after the fact. I'm not aware if they have done any testing or reporting yet.

Oh well.

gordo

 

[meddington]

G7 insures that only the three colour gray (CMY) matches but we have removed most of the three colour gray with GCR.

IMO, this does not negate the effectiveness of a 3/c gray patch as a diagnostic tool for process control as it will readily show imbalance, and do so more efficiently/effectively that individual TVI values. We measure SIDs and TVI independently as well, despite the fact that those tones are rarely represented in the imagery in and of themselves, yet they yield useful data to the operator. But we've had these discussions before...I know Gordo's argument that 3/c does not represent the live area, but it is my opinion that deviations of the 3/c gray (in an environment where press and prepress have been finely tuned) shows the direction of press drift of the live image area. Just my opinion based on my experience...meaning I haven't see a case where imagery gets "bluer" when the gray patch gets "redder".

 

[Deleted member 16349]

...meaning I haven't see a case where imagery gets "bluer" when the gray patch gets "redder".

OK,

I am not totally sure of this but here is an example.

If the image is made from dots of M and C and both are increased while the 3C patch, which is made of dots of M, C and Y are also increased but the Y is increased much more than the others, it might show the image get bluer while the 3C patch gets redder.

Just a thought.

 

[Bret Hesler]

Unfortunately, that fact is very often ignored/missed by most of the printers who talk about their experience and results with G7, which makes me go completely.....oooops, calm down gordo.....think of Daffodils and bunny rabbits........

gordo

I am hoping that the combined work of Idealliance and now RIT on developing a standardized process of evaluating and "certifying" a company printing to a standard, (yes, I did just use those two words) similar to the PSO certification from BVDM/ fogra, will close this loop.

G7 on its own will not guarantee a "match" between two dissimilar systems. G7 in combination with ISO/ Gracol solids and overprints has in my experience achieved very good match between different presses with different inks and screening (FM excluded). This was sheetfed and web with four different ink sets at three different plants.

When I have done my G7/ Gracol calibration runs, I have always done them blind (no proof at press), and when I get a press sheet with good G7 tone curve and balance, and good Gracol solids and overprints, I take the press sheet to a light booth and pull out a Gracol 2006 calibrated proof, and it is a pretty good match. Perfect, no, but easily matchable on press with a few adjustments.

So, the confusion and separation of G7 method/ qualification and its underlying use in the Gracol 7 spec and gracol and Swop data sets is understandable. I think there is an assumption that the G7 qualification is based on printing to Gracol and or Swop aims, and I don't think it was made any clearer with the multi- level G7 qualification system.

Bret

 

[gordo]

To Erik:

With your example, I think the grey patch will get Yellower and darker while the Blue (C+M) in the live image area gets more saturated. So the grey patch would tell you to lower the Y SIDs which would have no effect since the color has no Y in it. And if the Blue in the live image are is a saturated blue color then (according to System Brunner and other folks) we likely wouldn't see the difference any way. The press operator would just be wasting time and materials adjusting the press.

As Meddington notes, my ideas on this are fairly well published. What I have yet to see is any published research in support of the efficacy of the grey balance patch in the color bar. It's just taken as an axiom - just like the sun revolving around the earth once was. Is it too much to ask for some hard testing before it gets enshrined in the print gospels?

To Bret;

Thanks for your excellent comments, especially the honesty of "FM excluded".
As I wrote, I've no doubt that the G7 method works. That being said, there are other methods which are capable of delivering the same result which, to my knowledge, were never investigated. Some of those methods do not require an expert consultant on site to implement. I would have preferred it if Idealliance would have just defined the target (e.g. GRACoL7 or better) and allowed folks to achieve the target any way they wanted to. Then make the printer's qualification/certification based on the them delivering the target rather than what method they used. So instead of G7 Master Qualification and G7 Compliance - a blessing of a grey balance/tone curve method being used - the printer would be GRACoL 7 certified - a statement that the printer outputs to the GRACoL7 specification and therefore that the print buyer has a reasonable expectation of the color they're going to get with presswork and proofs coming from different shops. If they're all GRACoL 7 certified then the color should be all similar.

On a sidebar - the G7 Master bug states: "Proof to Print" shouldn't that be "Print to Proof"? It suggests that the proof should target the presswork rather than the presswork target the certified proof.

best, gordon p

 

[meddington]

With your example, I think the grey patch will get Yellower and darker while the Blue (C+M) in the live image area gets more saturated. So the grey patch would tell you to lower the Y SIDs which would have no effect since the color has no Y in it.

True. The gray patch in this scenario would indicate an imbalance, and assuming there is no yellow imagery on the form, the adjustment may be unnecessary. However, you're omitting the potential benefit of an efficient control element to guage the current inking conditions in relation to the target established during calibration. Regardless of what imagery on the form, it would be beneficial to have a consistent starting point on a press run if imagery is based on the same conditions.

Thanks for your excellent comments, especially the honesty of "FM excluded"

Just a comment. We proof toward Gracolcoated and print with FM (20 micron HDS) with an excellent proof/press match on the majority of imagery...possible exceptions to some pastel tones.

So instead of G7 Master Qualification and G7 Compliance - a blessing of a grey balance/tone curve method being used - the printer would be GRACoL 7 certified -

A concept that is being considered (in addition to, not necessarily a replacement of).