G7 tolerances

[povertycow]

Am I the only one who misses the good old days of pleasing colour???

 

[ScannerMan]

You'll find that, generally, the SIDs for a number 1 sheet (well, a number 2 - I'm not sure I've ever seen a number 1 sheet) will fall around 1.35c, 1.35-1.4m, 1-1.10y, and 1.8k, BUT THESE ARE NOT HARD AND FAST NUMBERS. YOU WANT TO HIT THE L*a*b* TARGETS. Solid ink density doesn't describe color.

As to TVI - I think that will vary quite a bit depending on the ink package, paper traits, environmental variables, blanket surface, ink-form roller size and condition, plate grain, fountain solution, lunar phase, wind-speed and direction, and whether or not you saw a black cat on the way to the press calibration.

G7 grew out of the inadequacy of ISO 12647 and SWOP to achieve a consistent appearance. TVI and SID do not describe color. They also don't describe your overprints, which I will argue are more important in color reproduction than the solids. In my experience, achieving TVI and SID targets did not yield neutral grays.

It was a funny conversation I had once when I was describing G7 calibration to an old scanner operator. To her, the idea of balancing the grays was not novel at all. That's when it occurred to me that Don Hutcheson is an old scanner operator/engineer/tech and he's an avid photo buff. We're just doing, on press, what scanner operators used to do.

For #1 sheets and a good ink set..ie Toyo you can see SIDs of 140 C, 145-150 M, 110 Y 170-180 K on a good day !

Back to Gordos original question.... Before G7 Calibration existed most shops ran linear plates or slight cutbacks. All 4 colors using the same curve....Run press proof to Density standard of the day or something that the pressroom liked and make proof match that sheet...oh and those were the FUN days.

Now you get gray balance throughout the entire tonal range, from extreme highlights through shadow.CIELab Aims for solids,overprints and gray balance....with the G7 Methedology.

Exactly Rich on the scanner operator analogy.......I remember Dons scanning guide WAY before G7 and it clicked immediately !

And Gordo I implemented G7 Back in 2007, it works better than anything previously done in offset.
Ive used it on Sheetfed and Web presses, Coated and Uncoated stocks,Stochastic and UV inksets.


Just my opinion though !

 

[David Milisock]

>Am I the only one who misses the good old days of pleasing colour???
As 27.5% of G7 certified printers in Pennsylvania have dropped their certification I would say no.

 

[gordo]

Back to Gordos original question.... Before G7 Calibration existed most shops ran linear plates or slight cutbacks. All 4 colors using the same curve....Run press proof to Density standard of the day or something that the pressroom liked and make proof match that sheet...oh and those were the FUN days.

Hmmm...I beg to differ regarding your history. For example, in the film days - linear film was the accepted file exchange format. So all 4 films supplied to the printer were linear. The plates burned from those films were not linear and the resulting presswork certainly wasn't linear and the tone reproduction curves on press were not the same for each color despite all the plates having the same curve. There was never an industry specification (e.g. SWOP) for SIDs - there were suggestions for GRACoL but they weren't specifications. Proofs weren't made to match the presswork condition (except in the few shops that were ignorant about specifications and standards or who fell victim to the sophistry of some color experts). Presswork color was tuned to the certified proof made to the industry specification (e.g. SWOP) from plates burned from the linear film.

Now you get gray balance throughout the entire tonal range, from extreme highlights through shadow.CIELab Aims for solids,overprints and gray balance....with the G7 Methedology.

ISO 12647 x has always referenced CIELab aims for solids, and overprints. Grey balance was the result of a properly set up press condition.

Exactly Rich on the scanner operator analogy.......I remember Dons scanning guide WAY before G7 and it clicked immediately !

Scanners do not mechanically operate like presses - if they did, using just one example - 50c, 50m and 50y would produce neutral grey. Since they don't you end up with a G7 methodology.

And Gordo I implemented G7 Back in 2007, it works better than anything previously done in offset.
Ive used it on Sheetfed and Web presses, Coated and Uncoated stocks,Stochastic and UV inksets.

No offense intended, but, if before G7 you worked according to how you describe the methods used before implementing G7 - then I don't doubt that it works better than anything previously done in [your]offset.

best, gordo

PS, I'd still like to see examples of those pre-G7 press curves and the post-G7 press curves. Also what is the range of SIDs that is represented by a delta E of 5.

 

[rich apollo]

…in the film days - linear film was the accepted file exchange format. So all 4 films supplied to the printer were linear. The plates burned from those films were not linear and the resulting presswork certainly wasn't linear and the tone reproduction curves on press were not the same for each color despite all the plates having the same curve…

I think the difference is the non-linearity in your example was not intentional, or controlled. Plates weren't exposed to achieve a specific TVI; blankets and plates are packed relative to bearer height, not to achieve a particular TVI; impression cylinder pressures are set, generally, to achieve .004" "squeeze", not to hit specific tone reproduction. We have greater control, now; or perhaps it's more accurate to say that we have greater control in PrePress now.

Grey balance was the result of a properly set up press condition.

I will disagree with you on this point, Gordo. Change paper and you change your gray-balance. Print on Balboa Velvet and you get pink, print on Carolina cover and you get yellow, print on Sommerset and you get blue. I have to be ready to run on both newsprint and #5 gloss text. Running plates properly exposed from linear films, should I expect gray-balance on each?

There are too many ink packages available to be able to say that gray-balance is the result of a properly set up press. ISO color conformance is only ONE condition for which an ink might be chosen. What about UV, proximity to food, rub resistance, drying time, chalking, compatibility with coatings, cracking on folds, ability to be run through a laser printer, adhesion to plastics, et cetera.

What about screening? A properly set up press condition established around what screening? When you change the screening, is the press not properly set up anymore?

Scanners do not mechanically operate like presses - if they did, using just one example - 50c, 50m and 50y would produce neutral grey. Since they don't you end up with a G7 methodology.

That's not the point, Gordo. The point is that gray-balance is a foundational concept in scanning and color correction. Don Hutcheson figured out a way to use that in lithography. I have the tools to make gray-balance happen - instead of hoping that it will.

PS, I'd still like to see examples of those pre-G7 press curves and the post-G7 press curves. Also what is the range of SIDs that is represented by a delta E of 5.

I can't give you a satisfactory answer on SIDs. I could root around and find some G7 curve data.

 

[Alois Senefelder]

Colour Specification UK

Colour Specification UK

Gentlemen,

Throughout my working life, I maintained close collaboratoin with my colleagues
in the Photo Litho Reproduction department (Pre Press) - All Final 4 colour halftone screened Positive Printer Separations
had the required tone reproduction curves and dot gain corrections built in for each colour.

I enclose a 1985 PIRA Guide PDF


Regards, Alois

 

[Alois Senefelder]

********

********

One more PIRA Guide PDF - which completes this post



Regards, Alois

 

[gordo]

Thanks for your patience Rich.

I think the difference is the non-linearity in your example was not intentional, or controlled. Plates weren't exposed to achieve a specific TVI; blankets and plates are packed relative to bearer height, not to achieve a particular TVI; impression cylinder pressures are set, generally, to achieve .004" "squeeze", not to hit specific tone reproduction. We have greater control, now; or perhaps it's more accurate to say that we have greater control in PrePress now.

Actually, as the Pira PDFs that Alois posted show, the whole prepress process (film, plate, inks, press) was controlled to deliver, among other things, a specified TVI.

I will disagree with you on this point, Gordo. Change paper and you change your gray-balance. Print on Balboa Velvet and you get pink, print on Carolina cover and you get yellow, print on Sommerset and you get blue. I have to be ready to run on both newsprint and #5 gloss text. Running plates properly exposed from linear films, should I expect gray-balance on each?

I'm speculating here, but, if you're talking about what an instrument measures then you'll likely not have grey balance on papers that have different color casts. If, instead, you're talking about what the human sees, then you will have grey balance. Unfortunately I have not seen samples of what you are describing - so I don't know.

There are too many ink packages available to be able to say that gray-balance is the result of a properly set up press. ISO color conformance is only ONE condition for which an ink might be chosen. What about UV, proximity to food, rub resistance, drying time, chalking, compatibility with coatings, cracking on folds, ability to be run through a laser printer, adhesion to plastics, et cetera.

Well you chose the base ink hues according to ISO 12647 - X. In the steam powered days of yore, that meant that if you hit the correct TVIs/tone reproduction curves using those inks run at their optimal SIDs (not part of the specification but included as guidance) then, yes, you achieved grey balance.

What about screening? A properly set up press condition established around what screening? When you change the screening, is the press not properly set up anymore?

If you are printing to an industry specification/standard then the screening is already defined. If you target your own standard (e.g. FM), then, yes you may need to adjust your curves to achieve grey balance. Although in my experience that's not been an issue.

The point is that gray-balance is a foundational concept in scanning and color correction. Don Hutcheson figured out a way to use that in lithography. I have the tools to make gray-balance happen - instead of hoping that it will.

I just think that it is dangerous to apply a methodology from one technology to another. It may work, or it may not. After all, a scanner is designed to digitize color - however, a press is not a color device, it is a device for laying down a film of ink (and water). That's why a scanner has color controls but a press does not.
I, and thousands of other lithographers, have set up many presses to grey balance before G7. It's always been foundational to setting up a press. I know that Heidelberg, MAN, Creo/Kodak, and likely other vendors have software tools to grey balance a press (or proof).

I guess I'm still stuck with my bad experiences at the GRACoL 7 press runs conducted by Don and the idealliance committee that I attended. My apologies. The issue that I struggle with is that there ever occurred a simple test where the press was set up to ISO 12647 and that result then compared to a press set up using G7. Put another way, if you're going to change how printers operate, I think that it is important to show/remind them of how they are currently operating and the issues they face doing it that way. Then show them how the new way is better, more effective etc. The only positive experiences from G7 adopters that I have heard (some are in this forum) are from printers who didn't know what they were doing in the first place, had no QC processes, no process control, etc. So anything would be better than the way they were operating.
So, from my point of view, complexity has been added but I can't see the benefit vs the traditional methods.

I can't give you a satisfactory answer on SIDs. I could root around and find some G7 curve data.

I'm surprised that no one seems able to provide such simple pieces of information.

• What SIDs are your presses running at to deliver the target CIE L*a*b* values? (on what paper)
• What SID tolerances result in keeping the target CIE L*a*b* values within a deltaE of 5 (the specification).
After all, when a press operator is at the press console adjusting ink keys - that's what is being adjusted - ink film thickness.
• What were your press tone reproduction curves before G7? In my experience you cannot build any plate/press curves until you first see the current press condition. The tone reproduction condition of the press may mean that you should not attempt to build curves G7 or otherwise. So every printshop/consultant should have this info at hand.
• What were your press tone reproduction curves after G7? Again, every printshop/consultant should have this info at hand so that they can show the customer where they were and where they're now at.

best, gordo

PS, If there are any printers in the Seattle, Portland, Vancouver, Vancouver Island areas that have done the G7 thing and would let me visit their shop and see what they are doing and share their experience I would sure like the opportunity. Contact me at pritchardgordon @ gmai (dot) com

 

[rich apollo]

I'm surprised that no one seems able to provide such simple pieces of information.

• What SIDs are your presses running at to deliver the target CIE L*a*b* values? (on what paper)
• What SID tolerances result in keeping the target CIE L*a*b* values within a deltaE of 5 (the specification).
After all, when a press operator is at the press console adjusting ink keys - that's what is being adjusted - ink film thickness.
• What were your press tone reproduction curves before G7? In my experience you cannot build any plate/press curves until you first see the current press condition. The tone reproduction condition of the press may mean that you should not attempt to build curves G7 or otherwise. So every printshop/consultant should have this info at hand.
• What were your press tone reproduction curves after G7? Again, every printshop/consultant should have this info at hand so that they can show the customer where they were and where they're now at.

All very good points. I think I see what you're driving at.

 

[gordo]

All very good points. I think I see what you're driving at.

When I worked with engineers at Creo, one of the things that they used to say whenever they heard claims about new products or or the benefits of different ways of doing things was a quote by W. Edwards Deming the father of the science of quality in manufacturing: “In God we trust; all others must bring data.”

If there's no data to support it, then it's faith-based. I've been asking these same simple basic questions for years, including at the idealliance committee, and no one seems to have it. Or, if they do, they won't share it. I'm expected to just believe it because some authority has said so, or because an unknown believer has said it is so. I find this very strange, especially when everyone seems to want to promote printing as a science rather than a craft.

best, gordo

 

[Deleted member 16349]

One of the engineers (an American) that I worked with at Creo used to say: "In God we trust. Everyone else needs data."

If there's no data to support it, then it's faith-based. I've been asking these same simple basic questions for years, including at the idealliance committee, and no one seems to have it. Or, if they do, they won't share it. I'm expected to just believe it because some authority has said so, or because an unknown believer has said it is so. I find this very strange, especially when everyone seems to want to promote printing as a science rather than a craft.

best, gordo

Amen.

Science and technology are two very different things.

In the printing industry, technology is more related to craft than science. One uses a technology to do something but doesn't exactly know why or how things actually work. That is craft based usage. Nothing wrong with doing that but it is not science based.

At the heart of science is theory and data either supports a theory or does not. If the data does not support a theory than the theory is invalid. It only takes one confirmed invalidation to invalidate a theory and then you do not have a theory and one should not keep calling it a theory.

Observations are not theory. Theory tends to be mathematical.

There is room and need for both science and technology/craft in the printing industry but it is good to know the difference so one is not mislead by marketing hype and the promotion of helpful but wasteful concepts that are presented as being solutions.

 

[David Milisock]

Isn't this all academic BS anyway? What percentage of real print shops can for their print presses run anyhting but general coated and uncoated profiles? Let's face it they run 20 different papers at least with miniscule run lengths and even the white point of the house gloss changes from lot to lot. wE CAN banter on and on about the specifics of color management processes but for 99% of the shops out there beyond a couple general profiles and a set of application procedures it means little more then a real possible we

 

[gordo]

Isn't this all academic BS anyway? What percentage of real print shops can for their print presses run anyhting but general coated and uncoated profiles? Let's face it they run 20 different papers at least with miniscule run lengths and even the white point of the house gloss changes from lot to lot. wE CAN banter on and on about the specifics of color management processes but for 99% of the shops out there beyond a couple general profiles and a set of application procedures it means little more then a real possible we

To a certain extent I agree with you.

But there are a few different issues involved in this thread.

One is about how specifications, standards, methodologies, "information" are propagated in this industry. This was eloquently captured in Erik's post. There is a great deal of BS in this industry, and as a result printers lose precious time, money, and waste resources.

Another is the impact on printers when print buyers/specifiers, who very often have little understanding of the process, start demanding that their print suppliers adhere to poorly developed and articulated specifications, standards, and methodologies.

Another is how difficult it is for a printer who wants to improve their processes to get basic, useful, actionable, accurate information.

Those are not academic BS issues.

best, gordo

 

[David Milisock]

>Another is how difficult it is for a printer who wants to improve their processes to get basic, useful, actionable, accurate information.

Those are not academic BS issues.

Agreed, with that said I was just ask by a client about a G7 certification for her industry because a prospective client was demanding it. So a quick check showed 0 G7 certified suppliers for her industry in the State and for print proofing 7 out of 20 in the State had let their certification expire. I have over 40 printers in my county alone, several large and successful ones.

I think that one can deduct that the information on standards, the accuracy and usefullness of such information has much more limited value then one might first glean from the large amount of press and forum chatter about them.

 

[Deleted member 16349]

Isn't this all academic BS anyway?

David, I have to agree. Maybe not academic BS but sad BS.

As you say, there are problems in print shops. I tend to look at the process and I see problems at every level.

Presses don't print exactly the same image from impression to impression.
The plate is not inked consistently in all locations in the image. This can change and be dependent on the image itself. How can one have a accurate profile of a press under this kind of condition? Can't.
Averaging out a wide variation of printed data for a profile does not result in an accurate profile, but the idea of averaging is used often without concern for standard deviation.
Density varies through the run.
Presetting the press only gets close to target densities but not within a required tolerance.
Density tolerances are too wide in order to provide colour consistency.

Colour science is helpful but only represents an average of how people see under limited conditions.
People see colour differently. A range of about Delta 2 has been suggested in a technical paper. (which one I can not remember just now)
Fluorescence in papers and inks cause a problem in predicting colour.
One can't print every colour that is within the Gamut volume that is suggested by the gamut plots for a device. This means that just because one colour is in a gamut volume of one device, one might not be able to match it even if that same colour is in another gamut volume of another device.

It goes on and on. With so many potential problems in the process, I am amazed it is commercially workable. But of course it is but I also see lots of frustration in this industry and any help I am sure is appreciated, even if sometimes it is imaginary.

 

[Deleted member 16349]

There is a great deal of BS in this industry, and as a result printers lose precious time, money, and waste resources.

Gordon,

This is the sad BS that I refer to. Wasted time and resources. I said it years ago, that G7, even if it is marginally better, was still the wrong thing to do because it wasted time and effort that could have been directed to solutions. Not perfect solutions but much easier solutions and ones that would be more general in solving colour matching problems using the capability of new technology.

The problems with the press I have talked about for a very long time but there was and still is no interest to investigate that area. The biggest resistance comes from people who love the process the most. Unfortunately the result is more wasted time and resources.

When the culture changes to one which demands valid knowledge, then maybe things will change. It is not here yet.

 

[gordo]

Not to put too fine a point on it....but.

Perhaps the lack of data is related to the parachuting color expert John Frum.

Management decides that the shop needs some of that G7 because customers are asking sales about it.

So in parachutes John Frum the color expert - likely the same one that ten years earlier told the shop that the proof needs to match the press. He talks to the pressroom in terms of deltaE, deltaF, C*, and H (not to be confused with h). Of how density and dot gain are of the old ways and not the new and improved ways. John Frum sets up the shop to G7 using a plethora of exotic tools and spreadsheets the likes of which the pressroom has never seen before. After much plating, and printing and measuring one press sheet is selected by John Frum as the chosen one. It is blessed, stamped and sent off to be certified so that the shop can place the hallowed G7 sign on their door post so that all who enter will know that they adhere to the ways of John Frum.
After just a day, John Frum promptly leaves taking his shiny tools with him. The people pressroom are left stunned. "We don't understand what happened, but it must be good and right" they agree. Maybe John Frum will return one day and see for himself how well we have kept his ways.

Well, it is Friday after all.

best, gordo

 

[David Milisock]

>I am amazed it is commercially workable. But of course it is
I agree and in my experience on press I'm no better off then when all we had was postscript color management. In terms of digital be it ink jet or a Next Press we are certainly better off. If I do the leg work I can get REPEATABLE color from any digital vender, however to get information from a shop so that I can create files that work on press while minimizing color shift from my file, forget it.Can I get color from a shop that matches from their digital print engine and their press? No. Is it reasonably close considering the different technologies? Yes, not bad. The industry has the ability to print with different methods with reasonable color but the BS is that it matches. It does not and we need to tell the truth or admit we are blind.

 

[David Milisock]

I agree I think it starts with being truthful. We can be better than we were even 5 years ago. Can we match across processes and medias? No, but we are better and we have better repeatability and in the real world that's a real true point.

 

[jstotz]

One more question to add. Which delta E calculation method is used to evaluate the sheets. I found a document on the idealliance web site, called G7 targeted, inside the G7 & GRACoL Qualification Kit for Sheetfed Offset. On page 1, it says to use Delta E*ab, which I think is the original delta E76 calculation which doesn't take into account perceptual non-uniformities. On page 25, the column heading is delta E'ab, but there is a footnote that says something like: ISO 12647 currently uses Delta E'ab (1976) but that one is inaccurate ... so you should use one of the weighted ones instead like CMC, 1994, or 2000.

And the proof verifier spreadsheet says delta E76

So come on people, which is it?