ISO 12647 Paper Types

[JDT]

Hi Everyone,

I am new here so would like to briefly introduce my self. I am a photographer. In my small little world with my Epson wide format printer, and creating my own custom ICC profiles, I am getting very accurate color reproductions. So I know what an end-to-end color managed work-flow is capable of, and I know that it works. But whenever I have to send artwork to a print shop, I want to pull my hair out. Which is why I am here. From reading this forum I can tell that there are many people here with lots of experience, so hopefully I can learn how to properly communicate color with a print shop.

What I understand so far about printing presses (which is not a lot), is that you do not create custom profiles for every press, ink and paper combination. Instead every press is adjusted (or calibrated) to behave as per some or other standard profile.

Currently I am dealing with a print shop that claims to be printing according to ISO 12647-2. So this is my question: How do I find the appropriate ICC profile to use? I noticed that there are several different profiles, depending on paper type. So how do you decide which one to use?

I have a sample of the paper that will be used for my print job. I measured the white as Lab=94,2,-6. On areas of the paper that have the optional spot UV coating, the white is Lab=94,1,-4. So there is already a deltaE=2 difference just between parts of the paper with and without the spot UV coating. Does this mean that I should use two different ICC profiles for the areas with and without the spot UV? Or is this compensated for automatically by the press calibration?

Thanks
Johan

 

[gordo]

I have a sample of the paper that will be used for my print job. I measured the white as Lab=94,2,-6. On areas of the paper that have the optional spot UV coating, the white is Lab=94,1,-4. So there is already a deltaE=2 difference just between parts of the paper with and without the spot UV coating. Does this mean that I should use two different ICC profiles for the areas with and without the spot UV? Or is this compensated for automatically by the press calibration?

A couple of things to consider:

1- If you calculate using CIE76 the delta E is 2.24
If you calculate using CIE94 the delta E is 1.75
So it's a good idea to specify how you are calculating deltaE whenever giving numbers.

2 - The deltaE you are seeing may just be the result of how your particular instrument "sees" the samples. A different instrument may "see" and report very different values. (According to a recent study done by Greg Radencic, Research Scientist, at PIA/GATF). So the deltaE value may not reflect what you actually see.

I've never heard of anyone using different ICC profile for areas with and without spot UV. You have to be careful about the meaningfulness of deltaE values. They are for comparing two samples, however, presswork is usually viewed in isolation. Also, instruments do not agree with one another (even the same brand/model), they don't respond the same way to different types of samples, and there is no standard for how the difference should be calculated.
So while it appears very scientific and precise to say that such and such have a deltaE of 2.24 it doesn't mean that is so, nor does it mean that there is a meaningful difference. It may simply be in the "noise" of the instrument.

best, gordo

 

[JDT]

A couple of things to consider:

1- If you calculate using CIE76 the delta E is 2.24
If you calculate using CIE94 the delta E is 1.75
So it's a good idea to specify how you are calculating deltaE whenever giving numbers.

Yes, good point about the different deltaE calculations. I did the measurement with i1Share - I don't even know which deltaE calculation the software uses.

2 - The deltaE you are seeing may just be the result of how your particular instrument "sees" the samples. A different instrument may "see" and report very different values. (According to a recent study done by Greg Radencic, Research Scientist, at PIA/GATF). So the deltaE value may not reflect what you actually see.

I don't know what the absolute accuracy of the i1 is, but I suspect there is probably a fair amount of variation between instruments. But I know from experience that it is fairly consistent from one sample to the next, so the difference between two measurements taken directly after one another is probably not too far off.

But the point I was trying to make is that there is a clear difference between the paper color with and without the spot UV. I can easily see it with my eyes. Surely this must affect everything printed as well?

However, I only mentioned the spot UV issue as an example. The real problem is the paper white point. There can easily be deltaE 10 difference between different papers, so I still don't understand:
1) how do you pick the correct standard profile for your paper, and
2) how can these "standard profiles" even work considering the wide range of paper white points available.

Thanks
Johan

 

[rich apollo]

...hopefully I can learn how to properly communicate color with a print shop.

What I understand so far about printing presses (which is not a lot), is that you do not create custom profiles for every press, ink and paper combination. Instead every press is adjusted (or calibrated) to behave as per some or other standard profile.

This is only partially true. Some folks calibrate their presses, some don't, some folks create custom press profiles, some don't.

Currently I am dealing with a print shop that claims to be printing according to ISO 12647-2. So this is my question: How do I find the appropriate ICC profile to use? I noticed that there are several different profiles, depending on paper type. So how do you decide which one to use?

You can determine whether, or not, they are printing to 12647-2 or not, by measuring some colorbars. As to the appropriate profile, find out what the print shop is using for proofing. THAT is what they are trying to match.

I have a sample of the paper that will be used for my print job. I measured the white as Lab=94,2,-6. On areas of the paper that have the optional spot UV coating, the white is Lab=94,1,-4. So there is already a deltaE=2 difference just between parts of the paper with and without the spot UV coating. Does this mean that I should use two different ICC profiles for the areas with and without the spot UV? Or is this compensated for automatically by the press calibration?

First, the paper is pretty blue. It's actually outside of spec. The spec reads 93 0 -3 on black backing or 95 0 -2 on white backing with tolerances of ± 3 ± 2 ± 2. So, they may be printing to hit one of the TVI curves defined by the spec, but that does not ensure a colorimetric match.

As far as I know there is no spec that deals with coating. Sounds like this is creating a fairly significant color shift.

What is the complaint you have with the product. What I mean is, tell us what you see; describe what you find objectionable and let's go from there.

 

[JDT]

First, the paper is pretty blue. It's actually outside of spec. The spec reads 93 0 -3 on black backing or 95 0 -2 on white backing with tolerances of ± 3 ± 2 ± 2. So, they may be printing to hit one of the TVI curves defined by the spec, but that does not ensure a colorimetric match.

Do I understand you correctly: The ISO standard only applies to paper within those tolerances (as specified for the different paper types), and other paper cannot be used under the ISO standard?

What is the complaint you have with the product. What I mean is, tell us what you see; describe what you find objectionable and let's go from there.

Very often when I see my work published the color accuracy is what I would consider poor. So I am trying to find out why that is and what can be done about it. Right now my guess is that people are simply not using the correct ICC profiles.

For example, I have seen my work published in two different magazines - in the one it looked good, but the other not (washed out look). I managed to get a copy of the final PDF that the graphic designers sent out and saw that the images were encoded in US Web coated SWOP. (The graphic designers insist that they always use this profile and they are not interested in hearing about anything else.) So I phoned up the printers of the one magazine and asked them what profile they use and what do they do when they get a profile mismatch. Apparently they ignore the embedded profile and print the CMYK values as is. This is the equivalent of an "assign profile" in Photoshop. Could it be that the other magazine, the one that got it right, actually did a profile conversion to the correct profile before printing?

So what I find strange is why there is no clear message coming from the print shops that says: You absolutely MUST use this or that profile. Furthermore, would it not be a good idea for print shops to reject artwork that is not encoded in the correct profile?

 

[gordo]

For example, I have seen my work published in two different magazines - in the one it looked good, but the other not (washed out look). I managed to get a copy of the final PDF that the graphic designers sent out and saw that the images were encoded in US Web coated SWOP. (The graphic designers insist that they always use this profile and they are not interested in hearing about anything else.) So I phoned up the printers of the one magazine and asked them what profile they use and what do they do when they get a profile mismatch. Apparently they ignore the embedded profile and print the CMYK values as is. This is the equivalent of an "assign profile" in Photoshop. Could it be that the other magazine, the one that got it right, actually did a profile conversion to the correct profile before printing?

So what I find strange is why there is no clear message coming from the print shops that says: You absolutely MUST use this or that profile. Furthermore, would it not be a good idea for print shops to reject artwork that is not encoded in the correct profile?

From the ad specifications of Southern Living Magazine - a US publication: "Do not embed ICC profiles or any other type of color management." (The full specification is here: http://img4.southernliving.com/static/pdf/2011_market_specs.pdf)

This is very typical. The profile is not relevant for their purposes. In NA the magazine ad specification will say that it prints to SWOP. If your ad was created to SWOP specification and your proofer was set up to SWOP then you should get what you expect color-wise. Publication ad color and specifications are based on the proof conforming to the specification. The prepress and pressroom's job is to bring the presswork into alignment with the specification as embodied in the proof.

You need to confirm via the publication's media representative that the publication is run to the appropriate specification or not.

There are many reasons why one magazine got it right and the other didn't that have nothing to do with your embedded profile.

best, gordo

 

[JDT]

From the ad specifications of Southern Living Magazine - a US publication: "Do not embed ICC profiles or any other type of color management." (The full specification is here: http://img4.southernliving.com/static/pdf/2011_market_specs.pdf)

Why would they do that? Why do they not ask the opposite, that the client must embed the profile so that they can check it? And when there is a mismatch they can warn the client that they are probably not going to get what they are expecting?

This is very typical. The profile is not relevant for their purposes. In NA the magazine ad specification will say that it prints to SWOP. If your ad was created to SWOP specification and your proofer was set up to SWOP then you should get what you expect color-wise.

OK, but the problem is that the NA spec does not actually say "SWOP". All it says is "CMYK", which is not very helpful.

For example, what would happen if some European company send an ad to this magazine which is in ISO Coated v2 (ECI)? The magazine can now either:
1) see that there is a profile mismatch and refer it back to the client, or
2) do a profile conversion from ISO Coated v2 (ECI) to US Web Coated (SWOP) v2, which would probably still produce good results, or
3) ignore the embedded profile and just print it as if it was SWOP. This is the equivalent of assigning the profile, which of course will produce inaccurate colors.

However, since the magazine asked that "No ICC profile should be embedded", option 3 is what will happen by default. I don't understand, why does everyone work like this?

I started this thread trying to figure out what profile to use, considering that there are many different profiles for different paper types. So far I have not seen a single media kit or print shop that clearly states what specific ICC profile to use. And knowing that I am not allowed to embed a profile, or if I do, that the profile will be ignored, only makes the situation worse.

 

[rich apollo]

I don't understand, why does everyone work like this?

"Everyone" doesn't work this way.

The reason that so many do work this way, I believe, comes down to liability. The less a printer does to a file, the less liability they incur. In your situation, you're displeased with the printed result, you call the printer to audit the workflow, they say they performed such-and-such actions. If one of those actions is a color transform then you (the client) insist that they damaged the file and request a refund.

Prepress folks have long been taught NOT to alter the file. Some of us ignore this "conventional wisdom," but at our peril.

As to the paper, the values I gave you were for gloss-coated, wood-free - in other words for a #1 or #2 sheet on a sheet-fed press. For magazine work it's more likely that the paper specs would be:

Gloss-coated, web 87, -1, 3 on black backing, or 92, 0, 5 on white backing. So, again, the paper falls out of spec. If the paper is out of spec, then the print job is out of spec.

For example, what would happen if some European company send an ad to this magazine which is in ISO Coated v2 (ECI)? The magazine can now either:
1) see that there is a profile mismatch and refer it back to the client, or
2) do a profile conversion from ISO Coated v2 (ECI) to US Web Coated (SWOP) v2, which would probably still produce good results, or
3) ignore the embedded profile and just print it as if it was SWOP. This is the equivalent of assigning the profile, which of course will produce inaccurate colors.

1) Most of the time this just starts a fight, and the printer NEVER wins.

2) This is what I do; take WHATEVER you give me and make it into what I want/need.This has cured a whole HOST of evils for my shop.

3) would lead to a rather full print, but colorimetrically shouldn't be too bad.

Again, I would say you should find out what the printer is proofing to. THAT is what they're trying to match. Give me the screening and paper specs and I can steer you to some possible profiles. 12647-2 includes information for 5 paper varieties. There are separate profiles for non-periodic (FM) screening, too, so there are some very different possibilities. Contact me off-list at richard(dot)apollo(at)okoffset(dot)com - that will ensure that I receive your message. I don't check these threads every day.

It would also be worth figuring out how close you feel the print needs to come to your expectation. The window of operation for printing presses can be disturbingly large. Make sure that your expectations are within the capabilities of the presses.