Question: CYMK & ICC Profile Issues

[aprilk31527202]

Hello,

I am designing a book cover and I have my original design file in RGB (necessary for some of the graphic's layer modes) but I saved it out as a TIFF and converted that image to CYMK. The printer claims that they need an "uncoated" ICC profile, but if I try to change it to one of those my colors go from being blue / purple multi-toned to one value and they are extremely, extremely washed out. I cannot print it like that at all. I expect a bit of color shifting / not as vibrant of colors, okay. But this is truely terrible... I don't really understand how you could print anything half way bright or true to color in this gamut range... (To clarify, I can only find acceptable results for the image with "coated" profiles. But the printer warns that this will make the ink bleed on the paper. Still, there has to be a way to do this properly because most illustrators and book cover deisgners must work in RGB.)

Anyways, "uncoated" asside (though I am still wondering about that since some of the profiles do truely terrible things to my image) - although I have my image in CYMK I seem to have some kind of gamut warning. - But I thought switching from RGB to CYMK should eliminate this issue altogether? I am in the CYMK printing gamut, or am I not? I am obviously lacking a piece of the puzzle here.

Adding to my confusion, I have printed my files before using CYMK while leaving the standard ICC - the "Web Coated" one. And those images seemed to turn out very true to color... I am just concerned that with a new printer something may go awry, such as too much ink coating the paper.

(Do you know if the PDF itself comes with print color settings? I always understood that it inherited the profile of the original file.)

Thank you in advance! I look forward to hearing from you.

[CShubert]

Hi,

Have you asked the printer to supply you with a profile for their printer/paper combo? I would start there so that when you convert from RGB to CMYK it is based on how they have their machines setup. If they don't then ask them which of the uncoated profiles do they use? 

Here is a great resource on Soft Proofing in Photoshop and how to setup your prints.  

https://helpx.adobe.com/photoshop/kb/preparing-print-photoshop.html

Hope this helps.

[aprilk31527202]

I did ask, but they don't have one. They said "anything uncoated" but as I explained, that is not working with my image. Thanks for the link.

[aprilk31527202]

To clarify, it doesn't help me if I know that they need X and Z if my image looks like crap when I use those settings... Therefore, I am wondering how most people resolve these kinds of issues, becuase I would expect them to crop up rather often.

[D Fosse · Accepted solution]

There are no shortcuts to this. You need a specific profile that corresponds to the actual process. Don't talk to the customer relations people, they are surprisingly often rather ignorant of the technical aspects. Get the people who actually operate the press or the prepress operators. They are your friends and they always know.

Yes, printing on uncoated paper can be a challenge. The gamut is very narrow and some colors are inevitably lost. They just can't be reproduced and there's nothing you can do about it. So the first step is to soft proof to see what you're up against. Now you have a choice. You can leave the clipping and live with it, or you can try gamut remapping to compensate somewhat (but don't expect miracles). You won't get your colors back, but you may avoid hard clipping that wipes out texture and detail.

You can also try perceptual rendering intent instead of relative colorimetric. This already does some gamut remapping, by pulling in the saturated colors just before the clipping point.

The good news is that it usually works much better than you think. Remember, it won't be seen next to the original, and when you flip through something printed on uncoated paper, it sort of creates its own reality that the eye tends to just accept as-is.

[aprilk31527202]

Hello, I find this helpful and I get your point. I just never realized that uncoated papers were so unforgiving. All I know is that the image isn't retaining it's properties much at all in the softproof... Although I agree with you when it comes to the eye accepting what it sees when for example changing from RGB to CYMK, I cannot agree with you here. The image looks truely terrible with the uncoated ICC profiles. - I am talking about retaining a certain graphical feel as well as the colors. Things like 3D looking text created with text styles are flattening like crazy. Something that looks like a watercolor-like ink stain is left completely lifeless. And the image seriously went from three colors (blue, purple, and silver) to one color (a desaturated faded yet dark purple). And I was already testing it using "perceptive". It isn't even kind of okay... Does everyone who does nice covers always use printeres who use coated paper? Because I wasn't under the impression that this was so. Becuase honestly at this point I think bled covers would look better than the change  I see with these profiles, which frusterates me.

[aprilk31527202]

It seems I was checking a certain profile. I don't like the changes that uncoated makes, but I found one that will probably be doable. I just hope that they add a gloss finish on top of it to bring the colors back out... 

[D Fosse · Accepted solution]

No, you're still missing the point. You can't just "find" a profile. One one profile is the correct one, and that's the one that characterizes the actual printing process, the press/ink/paper system that produces the finished result.

The profile doesn't do anything. It's just a map, and it has to describe the actual terrain. If it doesn't, the map is useless. It's not about what you like. You need to deal with the reality of the print process.

If they add a gloss finish it's a different process, and then you need a different profile.

So to be clear: stop looking for a profile. Get the right one.

[aprilk31527202]

Thanks. To be clear, I specifically asked that printer for their profile, but they were unable to send me one. They claimed any "uncoated" would do... which did not at all seem professional to me either.

In general, the uncoated CYMK profiles were changing my image A LOT, and so my questions were more focused on that issue rather than on any particular ICC-profile. - It seems like an issue that a lot of artists could run into. - And since the profiles are a preview of how the image will print (under certain conditions; in contrast to the real conditions, which remained unknown to me) - I figured that a preview that COULD look alright was better than one that was obviously far off the mark. (But yes, I was "shooting in the dark" anyway, I get that.)

In the end, I suspect that they actually used the coated variant instead, as the colors look darker and more accurate than in the uncoated variant... (assuming the screen previews are accurate in contrast to the actual printing method [paper, ink etc.], the specifics of which still remain unknown to me.) But if they did, this would be unadvisable for ink-bleed reasons. From my standpoint, (in this case) I don't have a lot of control over the final print, which is why I was hoping someone more seasoned might have good advice for these kinds of situations...

We agree that it is best to work directly with the printer and to get their specific ICC Profile, but regardless of this, my initial question remains - What if the artwork looks horrible with that specified ICC profile? What do you do? And, no, it cannot always be taken into account (designed for) from the very beginning for multiple reasons...     

[D Fosse]

If it looks horrible with the profile they provide, there are three possible reasons:

• they're not providing the right profile for the actual print process
• there's a lot of gamut clipping which you weren't prepared for (soft proofing is the answer to this one)
• your monitor isn't properly calibrated to the actual output conditions

The last point is crucial and needs a little comment. The most efficient way to deal with this is to calibrate your monitor so that monitor white is a visual match to paper white, and monitor black is a visual match to max ink. Obviously, having a previous print example made with the same process helps a lot.

This isn't about numbers, just using your eyes and adjust calibration parameters accordingly. The point is that if you have a visual match, what you see is what you get. No more guesswork. Then you can soft proof with confidence, and know what comes out the other end.

The thing about offset print is that it usually has a very high black point (dark gray) - much higher than your monitor has natively. The profile won't account for this (or only up to a very limited point). If your calibrator lets you set the black point, you may need to go as high as 1.0 - 1.5 cd/m². This is way higher than a monitor out of the box, which is usually around 0.2 - 0.3 cd/m².

Black level is hugely important. It has a very direct impact on the perceived "punch" of an image. It is the main reason for disappointment when people see the final printed result. The notion that "screen and print can never match" is in fact a myth - but you need to make the monitor match the final print. You can't do it the other way around.

So, to sum up, this is about setting the framework for color management and icc profiles to work in. To the profiles, white is just white, and black is just black. It all maps to those endpoints, and that's why it's important to visually define them.

[NB, colourmanagement]

Uncoated Print is basically incapable of the same saturation as Coated printing media. Every image printed on uncoated has to endure this lack of saturation and punch, so, in an uncoated print environment, where all images are suffering in the same way, you'll be fine. 

As an artist you can appreciate that if you were painting with watercolours you'd have far less saturation that oils or Gouache - however you would still be able to make a pleasing rendition of the scene.

Uncoated print is like that.

Using a coated icc profile for softproof (or even for conversion) isn't going to help at all if the image is reproduced uncoated. The "coated" soft proof will be far "nicer" than the "Uncoated" print and if the RGB is made into CMYK (separated) using a coated profile, there will be far too much ink on the press - if it gets that far, prepress may spot it.

In any case its not going to give you more colour than the ink on paper can physically provide.

I hope this helps
neil barstow, colourmanagement net :: adobe forum volunteer
google me "neil barstow colourmanagement" for lots of free articles on colour management
[please only use the blue reply button at the top of the page, this maintains the original thread title and chronological order of posts]

[CShubert]

Without a custom profile to convert to, there is no way to know how something may print at this lab.

MOST labs are color managed and have profiles for all their machines and papers, or at least the most comonly ordered. 

The soft proofing in Photoshop attempts to simulate what will transfer from a backlit device, your screen, to a front lit device, the paper. To do this, it must take the profile you are viewing and with known values, ie the custom profile, attempt to show you how it will print from that device on that paper.

It is true that colors will shift from screen to print based on the type of paper, and with the View/Gamut Warnings you can see what colors fall outside this printing profile/paper combination.

Are you going to use this printer for many projects? Maybe they would work with you create a custom profile for CMYK conversions? 

Have you seen this page on using soft proofing? Might be able to work with this with the printer?

https://helpx.adobe.com/photoshop/using/proofing-colors.html

[Stephen Marsh]

Can you attach a flattened, lower resolution RGB copy of the layered file, with the RGB ICC profile embedded? Sometimes you just have to accept physics and make "artistic" adjustments, knowing that end users don't know what the original RGB looked like so they don't have the same expectation as you do.