G7 Usage

[Skinflint]

I'm wondering if the implementation and continued use of G7 is growing or waning.

We all know many companies have closed doors and all, but are shops maintaining the certifications and are new shops joining the standard?

 

[kansasquaker]

My personal, unscientific, impression is that more, perhaps most, commercial companies have adopted the methodology (especially as software has made implementation seem easier) but fewer maintain Idealliance membership and G7 certification.

 

[Correct Color]

The problem with G7 is that it was designed to address a specific purpose.

That purpose was that prior to computer-to-plate, while it was possible in litho printing to determine individual primary color values; individual primary color densities; media white point and total-of-all-primaries ink density, what it was not possible to determine with accuracy was dot gain -- now of course referred to by the more scientific-sounding Tone Value Increase.

CTP made it possible to do that, and that's very basically what the G7 methodology does. So the idea was then that if you could determine tone curves as well as the above-mentioned variables, you could have all the elements of a machine-state in one handy portable, universal, stock profile; so you could then use that stock profile and get matching results whether you were printing in New York or Shanghai.

And to that extent it does work...

I'd say the downside is that all these printers who struggled hard and put all this effort into making their product a commodity shouldn't be surprised when their clients start wanting commodity pricing.

But anyway...

The thing is that G7 has branched out far afield of its original idea. But the problem is that beyond its original idea, it does not have any benefit.

The original idea again is that if you have all machine-state variables defined and articulated as standards except dot gain (TVI), then you can adjust dot gain (TVI)on a press to match a standard, use that standard as the expected tone curves in a stock profile, and print to that profile universally and get consistent results between presses.

That does not work at all in -- for instance -- large format inkjet printing. There are no industry-defined primary color values, no standard white points; no standard anything really.

And, since digital is, well, digital, the original issue in litho with tone curves in analog plates simply never existed in any sort of digital printing. Defining tone curves -- whether called linearization or calibration -- has been part of the process since day one. And it's important to understand that the only way G7 works as intended is when there are no changing machine-state variables but tone curve. Any other changing variables require a new profile be made, and once that profile is made, it defines the colors a device prints, not whatever intermediate tone-curve adjustment is used.

But a lot of effort, time and money has been expended trying to sell G7 to digital printers and their clients, and a lot bought in.

But a lot are now buying out.

So overall, my opinion is that while G7 may still have benefit in certain applications in lithography, other than that its influence is most definitely on the wane.



Mike Adams
Correct Color

 

[gordo]

The problem with G7 is that it was designed to address a specific purpose.

That purpose was that prior to computer-to-plate, while it was possible in litho printing to determine individual primary color values; individual primary color densities; media white point and total-of-all-primaries ink density, what it was not possible to determine with accuracy was dot gain -- now of course referred to by the more scientific-sounding Tone Value Increase.

In my experience dot gain/TVI has been a measurable metric for many years prior to G7. It should not have been a target for printing (tone value should have been) - but that's how the industry malfunctions.

CTP made it possible to do that, and that's very basically what the G7 methodology does.

CtP and digital files/PDFs made it possible to have whatever TVI the shop wanted by eliminating linear film as the file interchange format.

So the idea was then that if you could determine tone curves as well as the above-mentioned variables, you could have all the elements of a machine-state in one handy portable, universal, stock profile; so you could then use that stock profile and get matching results whether you were printing in New York or Shanghai.

And to that extent it does work...

G7 was originally promoted as a method to apply tone curves on printing plates to achieve grey balance in the final presswork. The assumption was that ISO 12647 didn't already deliver that and that the resulting grey balance and tone reproduction delivers a common "appearance" (whatever that means) across different printed pieces.

I'd say the downside is that all these printers who struggled hard and put all this effort into making their product a commodity shouldn't be surprised when their clients start wanting commodity pricing.

Print became a commodity with the advent of linear film as the file interchange format and specifications such as SWOP long before G7. Idealliance is merely casting the commodity aspect in solid rock. Sad.

The original idea again is that if you have all machine-state variables defined and articulated as standards except dot gain (TVI), then you can adjust dot gain (TVI)on a press to match a standard, use that standard as the expected tone curves in a stock profile, and print to that profile universally and get consistent results between presses.

I think the notion was that you should adjust solid ink density on press to match CMYK solid ink hue targets. I believe that resulted from a lack of understanding by the members of Idealliance as to why SIDs are measured in the first place (not for color). I was a member of the original GRACoL/G7 committee and attended meany of the press test runs. I was the only one on the committee who could actually run a press so, of course, my concerns were ignored.

Defining tone curves -- whether called linearization or calibration -- has been part of the process since day one.

Linearization and calibration do not mean the same thing. But people (vendors and printers) often get these terms confused or misuse them.

 

[Ralph Corneliussen]

We use a Measure Color system to read or sheets here on our Man Roland 700 presses. We use conventional inks and print on double sided poly SBS board. We use alcohol to help dry as well in our fountain solution. The G7 conversation has been going on for a few years here and we do not use it yet. We do use ICC profiling for our presses. Can the two work together? Is there pros and cons to doing this? Would using both give my operators a bigger window of control around color control and variance?

We do keep our presses in good running shape. TVI is consistent, not perfect numbers but consistent.

Thanks for your comments in advance.

 

[gordo]

We do use ICC profiling for our presses. Can the two work together?

Yes. G7 is a calibration method to put a process into grey balance.

Is there pros and cons to doing this?

No.

Would using both give my operators a bigger window of control around color control and variance?

No to the window. No if your variance is within the press run. Possibly if the variance is between presses.

We do keep our presses in good running shape. TVI is consistent, not perfect numbers but consistent.

TVI is not color. If your printing aligns with your proofs and your customers are happy and you are profitable - then there would have to be a compelling economic reason to make a change. I don't see G7 providing that in your case.

 

[Ralph Corneliussen]

Thank you Gordo.

 

[Correct Color]

Just a couple added thoughts...

Linearization and calibration do not mean the same thing. But people (vendors and printers) often get these terms confused or misuse them.

Well, yes and no. In actuality, it works out along the lines that linearization is always calibration, but calibration is not always linearization. So for that reason after years of correcting people I've become a little more sanguine about not picking too much over the difference, as calling linearization calibration is never technically wrong.

We do use ICC profiling for our presses. Can the two work together?

They can... But it's pointless to do. The whole, entire, original point of G7 was to make a tone curve for standardized printing conditions that worked with standardized profiles, thereby precluding the need for individual profiles.

If you're profiling, you can run the G7 routine as your linearization/calibration routine, but once you make an ICC profile, it's irrelevant, and your end result is no longer originally-described G7. Any other properly run routine will accomplish the same result. It's the ICC profile that tells the RIP what dots to create. Whatever calibration routine you use is simply part of the machine-state described to it.




Mike Adams
Correct Color

 

[SteveSuffRIT]

I would call no curve at all "raw".
"raw" may not produce a linear result, 50% input NOT equal to 50% output.
In this case, you would need a "linear" curve to get 50=50. CtP tolerance limit +/-1%.
Steve Suffoletto

 

[Deleted member 16349]

Well, yes and no.


They can... But it's pointless to do. The whole, entire, original point of G7 was to make a tone curve for standardized printing conditions that worked with standardized profiles, thereby precluding the need for individual profiles.

If you're profiling, you can run the G7 routine as your linearization/calibration routine, but once you make an ICC profile, it's irrelevant, and your end result is no longer originally-described G7. Any other properly run routine will accomplish the same result. It's the ICC profile that tells the RIP what dots to create. Whatever calibration routine you use is simply part of the machine-state described to it.



Correct Color

If I understand you correctly, I totally agree. There is a need to get on the path to the future and stop following legacy thinking which G7 follows. IMO G7 is a totally wasteful method and the strong marketing of this method has not been good for the industry. It holds the industry back by keeping it in the past.

As many have stated, custom ICC profiles work the best. IMO the real issue is how to make a custom ICC profile in the easiest way possible.

About 15 years ago, I realized that the path to make that happen would be a simple, accurate and fast way to capture thousands of colour patch data. Scanners can't do that because they are RGB based devices. Scanners can capture data from a large number of patches but that data is not colour data. What was needed is a device that captures true colour data, which in my opinion should start with the XYZ values. I have been waiting to see this kind of technology become available for 15 years.

Probably if people realized that automating accurate ICC profiles with thousands of data points was the goal, this technology might have been developed much earlier. I do see a glimmer of hope. Epson has just announced that they have developed a spectral camera that they think should be used for process control. I think it should and could be used for ICC profiling.

Here is a link.

https://global.epson.com/newsroom/20..._20191209.html

Potentially capture of thousands of small colour patches with one exposure. Done. Easy. Accurate.

As I have stated in the past, I think the printing industry does things backwards. True process control does not come from measuring and then adjusting but from making the processes inherently consistent, predictable and repeatable. The logic of colour management via ICC can only work well if the processes are predictable. Fix the processes and logically simplify colour management.

 

[Correct Color]

Correct me if I’m wrong but I always thought that linearization is simply no curve at all, and calibration is a curve that’s applied to achieve a given target?

I.e., A linearized set of plates is used to determine the amount of calibration needed.

Well...

Not in digital printing. Not being totally up-to-date on the process in litho, I can see the process you're describing, but... linear means linear. In digital printing, there are no plates, so you send linear data to the printer, then it prints whatever dot gain it prints, you read the resultant patches, and typically make a calibration routine that corrects for the dot gain and returns the result to linear -- thus the term "linearization." Although it's possible here to create some sort of non-linear curve, and some do, but it serves no practical purpose.

So yes, in litho you would have to have linear plates to accomplish the same thing, but again, the print is the point, not the plate. Obviously the press will have some resultant dot gain from printing with those linear plates. You read the patches printed with the plates, then create some kind of calibration from the results. That can be a linear calibration -- linearization -- or some curve that is not linear, designed to work with some pre-made profile.



Mike

 

[Correct Color]

As many have stated, custom ICC profiles work the best. IMO the real issue is how to make a custom ICC profile in the easiest way possible.

Okay, I'll get out my soapbox here and say I completely disagree.

The real issue is how to make an ICC profile in the best way possible.

And it is not so easily done, and from my perspective, all the people who are looking for a foolproof, easy "scientific" way to do it miss the point.

RIPs convert pixels into dots using information in profiles.

Before making an ICC profile, you must first describe the machine-state the ICC profile is characterizing. Nothing in the article you linked does that. And the variables in creating the machine state actually control a good bit of print quality that is not specifically related to color.

It takes experience, talent, and judgment to create a truly superior machine state, and then once you do, you have to print patches from a machine in that state, then read them with a spectrophotometer. Reading them with a camera, even if you could get a camera to differentiate individual patches and assign a L*a*b* value to each, would introduce ambient light into the final result -- which would not work.

Profiling is an art. It can't be learned in a webinar, or at a three day "boot camp."

What I wish, is that the industry would recognize this.



Mike

 

[Deleted member 16349]

Okay, I'll get out my soapbox here and say I completely disagree.



Profiling is an art. It can't be learned in a webinar, or at a three day "boot camp."

What I wish, is that the industry would recognize this.



Mike

Does not sound too convincing to me. It sounds a bit like when press operators say how important they are to the process. Some truth to it but science and technology has been proving that to be a flawed view as the skill and art have been designed out of the process. The same trend will happen with colour management. How fast? Probably not so fast but eventually. It is possible that you don't want that to happen but I suspect it will trend that way anyhow.

 

[gordo]

Correct me if I’m wrong but I always thought that linearization is simply no curve at all, and calibration is a curve that’s applied to achieve a given target?

I.e., A linearized set of plates is used to determine the amount of calibration needed.

On a CtP device, calibration (putting into a known condition) of the device is done by an engineer. It is done by setting exposure intensity, dwell time, plate processing etc. in a state where the imaged area is robust and the non image area is cleared. It has nothing to do with linearization. The result of the calibration is a plate with a (characteristic) typically non-linear tone response (e.g. (50% in file = 43% on plate)

Linearization is a modification of the inherent tone response of a calibrated plate by the use of tone curve applied in the RIP in order to achieve a linear plate (50% in file = 50% on plate).

 

[Alois Senefelder]

Hello Erik and Festive Greetings to all.


Where do these " Magicians " reside that have slain the 150+ variable inputs that need to be mastered and understood............ in the

Lithographic Printing Process?


Regards, Alois

 

[gordo]

Hello Erik and Festive Greetings to all.

Where do these " Magicians " reside that have slain the 150+ variable inputs that need to be mastered and understood............ in the

Lithographic Printing Process?

Regards, Alois

For the most part they reside in various vendors' offices. AFAIK none of those folks know how to run a press.

 

[OffsetStorefront]

All a pressman's skill is for naught if the customer buying the printing thinks the quality you get from the "button pressing" level of running the press is good enough.

 

[gordo]

All a pressman's skill is for naught if the customer buying the printing thinks the quality you get from the "button pressing" level of running the press is good enough.

A common misunderstanding. Actually, it’s not the primary job of the press operator to deliver “quality” in the press work. A related misconception is that offset presses are designed to print color.

 

[Deleted member 16349]

A related misconception is that offset presses are designed to print color.

Very true. Presses are designed or at least the attempt is to design them to print a repeatable ink film in the solid and a repeatable ink structure in the screen by managing the ink films on the form rollers. Colour is just a by product of that. People might argue how much an operator affects this and how much of it can be automated. I would say quite a lot can be designed out but others will argue differently. The historical trend is clear.

By not developing offset more, the print community has allowed the development of digital technologies which looks to replace a lot of the smaller offset technology in the future. They made it easy for the opposition printing technology to kill them. They deserve what they get.

 

[Alois Senefelder]

Gentlemen,