FOGRA 39, GRACoL 2006, GRACoL 2013 Questions

[gordo]

I did a plot of the TVI curves embedded in the official characterization data sets for FOGRA 39, GRACoL 2006, and GRACoL 2013:

FOGRA 39:

GRACoL 2006:

GRAcoL 2013:

We're told that there is very little difference between GRACoL 2006 and 2013 and that they look almost identical. So why are the TVI curves so different?
The 2013 curves looks like the CMY TVIs have been adjusted to make them identical despite the difference in SIDs of the CMY inks.
Any ideas?

 

[Stephen Marsh]

Gordo, what TVI do you get with the new Fogra 51 data?

http://www.fogra.org/en/fogra-stand...-download-of-fogra-characterization-data.html

Fogra 51 is the update to Fogra 39, just as GRACoL 2013 is the update to 2006. Just curious, I am not expecting them to be similar.


Stephen Marsh

 

[gordo]

Gordo, what TVI do you get with the new Fogra 51 data?

http://www.fogra.org/en/fogra-standa...tion-data.html

Fogra 51 is the update to Fogra 39, just as GRACoL 2013 is the update to 2006. Just curious, I am not expecting them to be similar.


Stephen Marsh

OK,

FOGRA 51 has two data sets.

Here's the first (FOGRA51.txt) plotted:

The second is spectral data (FOGRA51_Spectral.txt):

Can you make any sense about what's going on?

 

[Deleted member 16349]

Can you make any sense about what's going on?

If the industry has been living in a fantasy world in thinking tonal curves predict colour, then what does it matter what curves are used.

 

[gordo]

If the industry has been living in a fantasy world in thinking tonal curves predict colour, then what does it matter what curves are used.

Mustn't jump to delusions. Tonal curves have never been postulated to predict color.

 

[Green Printer]

Hi Gordo

How does Fogra and Gracal determine the data in the data set? Do they run just one press and say here you are?

 

[Deleted member 16349]

Mustn't jump to delusions. Tonal curves have never been postulated to predict color.

You know this and I know this but I suspect most people in the industry think it does align their press to print predictable colours. It is what people believe that makes them think a specific method is the solution to their problems.

 

[Deleted member 16349]

Do you believe that the G7 standard is the most accurate way to consistently reproduce matching color, industry wide?

G7 is not a standard and it in no way deals with providing consistent colour matching. Don Hutcheson, the inventor of G7, as far as I know, has never said that G7 ensures matching colour. As I remember, he specifically stated that if the G7 method could not get the colour matching required, then you had to go to ICC profiles. I think he also meant custom profiles.

G7 might be a slightly better method than just adjusting dot gain curves since it is determining the gray line (or L line in the gamut). Of course with gray component removal, one is not usually printing any of the G7 gray line screen combinations.

I my view, Idealliance has done a disservice to the industry by pushing G7 so hard instead of looking into other methods that can actually predict and provide consistent and reasonably accurate colour in a simple way. But what else is new in this industry.

 

[gordo]

Do you believe that the G7 standard is the most accurate way to consistently reproduce matching color, industry wide?

No because G7 is a method for grey balancing output devices which, by definition, is not color matching.

 

[kansasquaker]

I my view, Idealliance has done a disservice to the industry by pushing G7 so hard instead of looking into other methods that can actually predict and provide consistent and reasonably accurate colour in a simple way. But what else is new in this industry.

I'm a complete neophyte when it comes to color. So I have a hard time reasoning through things. I do understand that grey balancing a device doesn't necessarily control color in any other part of the spectrum, but we've had reasonably good luck with Gracol/G7. What would be the 'simple' alternatives you mention? How do they function?

 

[gordo]

This thread is veering off topic. If you want to discuss issues other than these TVI graphs please start a new thread.
Any ideas/comments regarding understanding/interpreting these graphs would be much appreciated.

 

[Deleted member 16349]

What would be the 'simple' alternatives you mention? How do they function?

They would have to be developed. Custom ICC profiling is close to what I imagine but it would need to be more predictable and much more easy to apply and this is possible but it needs to be developed by people in the industry, who so far are not too interested in thinking about this problem, which is in the press and the prepress areas. This means you are stuck with what you have which may work OK for you most of the time.

 

[Morgan Manning]

Great catch Gordo! As I have to straddle G7 and Fogra methodologies, I'm always interested in anomalies like these. I see the FRED
data is taking bringing cyan, magenta and black to the same TVI area. Are we reducing black contrast by substituting cyan and magenta?

This reminds me of the Pantone books back in 2006 (?, sorry I don't have any in front of me). The TVI was all over the place under their specifications.
I always thought it odd that we all strive to calibrate presses using TVI curves to have a target with an entirely different TVI. Of course,
this was before Near Neutral/G7.

 

[Puch]

To the original question: Gracol basically normalized the TVI of the inks in a 3+1 manner. FOGRA went one step beyond, they realized that TVI can be the same across all colors, since we can compensate against any defiencies during color conversion. If we use the same standard in the prepress and at the press (eg. Fogra51), the process must be flawless.

Remember, those TVI curves were invented in the pre-color management times, when color conversion happened at the computer of the drum scanner, and the result was imaged to film immediately. No need to stick with that idea today, when we can manipulate color at will, without any quality degradation.

That move should make the pressroom's task much more easier. Less room for error when checking and re-setting curves for CTP.

 

[gordo]

Great catch Gordo! As I have to straddle G7 and Fogra methodologies, I'm always interested in anomalies like these. I see the FRED
data is taking bringing cyan, magenta and black to the same TVI area. Are we reducing black contrast by substituting cyan and magenta?

This reminds me of the Pantone books back in 2006 (?, sorry I don't have any in front of me). The TVI was all over the place under their specifications.
I always thought it odd that we all strive to calibrate presses using TVI curves to have a target with an entirely different TVI. Of course,
this was before Near Neutral/G7.

These shouldn't be anomalies since they are in the official datasets for the industry's standards/specifications. I'm just not sure what they mean or what their significance is. Pantone is different in that it is a private business. Pantone can do whatever they want and you can take it or leave it. ISO/GRACoL on the other hand are industry-based and all relevant industry standards and specifications derive from and point to those organizations.

If you're doing G7 - could you post a similar graph of the TVIs of a press that's been so calibrated?

 

[Morgan Manning]

No Title

Here's something (to be taken with a really large grain of salt), I was helping out another department with their cut sheet digital:

 

[Correct Color]

Gordo,

May take would just be that they're different. For whatever reason, you'd have to ask whoever made them.

Thing is though, that from my understanding, all these embedded tone curves are is an equivalent of linearizations on an inkjet printer. Or rather that this or that tone curve is embedded into this or that standard CMYK space, so therefore that space is only valid as a profile to a particular device when it's printing -- among other parameters -- with that exact tone curve.

But as has been alluded to here, that doesn't mean that one or the other wouldn't print identically.

They would. Same as I can make any number of inkjets print basically identically with far different linearizations, two standard spaces could print exactly the same if they had the same outer dimensions: White point; chroma values; SID's; and TID.

Their internals -- tone curves -- could be different, as you see here. But what that would mean is that in order to be printing absolutely correctly, and to match, each machine would have to be actually printing the tone curve in the color space/profile it was using.

(Edited to add:

In my view, Idealliance has done a disservice to the industry by pushing G7 so hard...

Amen, Brother. Amen.)



Mike Adams
Correct Color

 

[gordo]

Gordo,

May take would just be that they're different. For whatever reason, you'd have to ask whoever made them.

The characterization data is provided by ISO.

Thing is though, that from my understanding, all these embedded tone curves are is an equivalent of linearizations on an inkjet printer. Or rather that this or that tone curve is embedded into this or that standard CMYK space, so therefore that space is only valid as a profile to a particular device when it's printing -- among other parameters -- with that exact tone curve.

The characterization data represents a press configured to the ISO 12647-2 specification. I.e. the press characteristic that printshops are supposed to align to in order to conform to standards such as GRACoL and ISO. These TVI curves don't make any sense.

But as has been alluded to here, that doesn't mean that one or the other wouldn't print identically.

They would. Same as I can make any number of inkjets print basically identically with far different linearizations, two standard spaces could print exactly the same if they had the same outer dimensions: White point; chroma values; SID's; and TID.

Their internals -- tone curves -- could be different, as you see here. But what that would mean is that in order to be printing absolutely correctly, and to match, each machine would have to be actually printing the tone curve in the color space/profile it was using.

GRACoL 2006 and 2013 are based on essentially the same specification - the TVIs should be virtually the same. Just the fact that the SIDs are so different - the TVIs of GRACoL 2013 should not overlap almost identically the way they do.

(Edited to add: Amen, Brother. Amen.)

I second that demotion.

 

[Green Printer]

http://www.sovsib.ru/color/iso12647_en.pdf

Read Annex A page 14 and 15.
How do you measure ink film thickness on a blanket?
How do you measure ink film thickness on printed sheet?
Who did the test and what were their procedures to arrive at the ink film thickness data?

This ISO standard is based on 75% of a committee agreeing to the numbers presented at the ISO meeting

 

[gordo]

http://www.sovsib.ru/color/iso12647_en.pdf

Read Annex A page 14 and 15.
How do you measure ink film thickness on a blanket?
How do you measure ink film thickness on printed sheet?
Who did the test and what were their procedures to arrive at the ink film thickness data?

This ISO standard is based on 75% of a committee agreeing to the numbers presented at the ISO meeting

Well, you could use an Ellipsometer to measure the ink film thickness on the blanket.
Densitometers give you an indirect measurement of ink film thickness (which is why press operators use them). The different solid ink density values for the different process inks are based on a target ink film thickness on the press sheet of about 1 micron.