February 4, 2007

Busting the Myth: achieving consistent color across Photoshop, Illustrator, and InDesign

Plenty has been written about the holy grail of color consistency -- getting your screen and your print output to match across Photoshop, Illustrator, and InDesign (and Acrobat too). Adobe themselves will even talk about the newer setting in Bridge or in Adobe Creative Suite 2 in general which allows you to synchronize your color settings across the suite. And the nice thing is that it works -- for the most part.

But even with these settings, we still find those out there who complain of colors not matching. Mainly, these are files that include spot colors. I've seen some people talk about these issues, and I've seen a few suggestions for solving the issue as well. In fact, it always seems that spot colors throws a wrench into the workflows we depend on every day. A while back I discussed how spot colors and transparency can present some unexpected results (and I presented a solution to the issue).

I've touched on the real solution to this problem before, but only in smaller specific cases. Here, I hope to present a clear understanding of these settings that accomplish three very important things:

1. Your spot color plates are preserved.

2. The integrity of your spot color CMYK values are preserved should you choose to convert your spots to process late in the workflow.

3. Your onscreen proofs AND your composite proof prints that contain spot colors will match accurately across Illustrator, InDesign, Photoshop, and Acrobat.


I'll start by saying that some have offered solutions of switching the spot color definitions in Illustrator or InDesign to use LAB values (which match Photoshop) -- but this can be VERY dangerous and results in the loss of number 2 mentioned above. Sure, we all know that a designer should use spot and process colors as needed, but we also know that many designers will spec spot colors (either because it's easier to do so by picking them from a book, or because they don't know any better). In addition, printers may convert spots to process late in workflows as job specs change under deadline. Item 2 listed above can often mean the difference between a job that prints correctly or one that must be reprinted (and at who's cost?).

UNDERSTANDING BOOK COLOR

So first, let's understand something that was introduced to Adobe Creative Suite 2: Book Color. You may notice that at times, a swatch will show up in Illustrator CS2 as something called a "book color". This occurs in InDesign CS2 as well, and here in the United States, you'll find it most often when you specify Pantone colors. So what is this book color setting? Well, here's the story: one of the reasons why we use spot colors is simply because there are colors that simply aren't reproducable in CMYK. Bright blues, purples, greens and oranges are perfect examples. But you also know that there are CMYK conversions of these Pantone spot colors. Pantone delivers an entire Excel spreadsheet to Adobe that lists each Pantone color, along with their LAB color equivalents. Obviously, the LAB color values match closer to the real color. Pantone also delivers a Solid to Process library which specifies the CMYK values of Pantone colors. Obviously, the latter is far less accurate and results in color shifts (and if you're paying attention, you know that by default, Photoshop uses the LAB values and AI, ID, and Acrobat use the CMYK values, which means that placing PS content into the other apps can result in color shifts).

Traditionally, a single swatch can only contain a single color definition. Since Illustrator or InDesign always used the CMYK values, proofing spot colors from these apps -- either on screen or on a composite proof -- was less than desirable. Especially considering the statement made in the previous paragraph, color shifts are also common due to this.

A book color swatch is a swatch that contains BOTH the LAB and the CMYK values of a spot color. In this way, Adobe can be more intelligent and use the values that are best for the situation. But there's a catch. You need to specifically tell Illustrator (or InDesign or Acrobat) that you want to use one value or another for the task at hand. But the important thing here is that you don't want to choose one OVER the other -- you always want to have both at your disposal. By this I mean, when you are actually printing and separating as a spot color, you want the best possible proof. But when you decide late in your workflow to convert your spot to process, you don't care about the LAB value -- you care about the CMYK breakdown and the consistency of that breakdown with other artwork, apps, etc. If you instruct your app to use the LAB color space for your swatch, then if you ever convert your spot to process, you get a LAB to Process conversion -- NOT the CMYK breakdown that the spot color manufacturer has specified in their book.

So what do you do? You use a feature that you would normally never imagine would have anything at all to do with this.

OVERPRINT PREVIEW TO THE RESCUE

OverPrint Preview is a wonderful feature that exists in Illustrator, InDesign, and Acrobat. But it's the worst named feature ever to grace those programs. That's because you might think that OPP only simulates or displays any overprint commands on your screen (which is very helpful). But it also goes a step further and uses a different algorithm for displaying spot colors. When you have a book color specified, we discussed that the swatch contains both LAB and CMYK values. In normal Preview mode, AI, ID, and ACRO all use the CMYK values for displaying the colors on your screen. But when OPP is turned on, the apps display those colors on the screen using the LAB values in the swatch.

So the first thing we've learned is that if you want to see consistent spot colors on your screen across all of your applications, don't touch the swatches -- rather change your preview mode by turning Overprint Preview on. This way, the integrity of the swatch remains, but you're basically instructing your application to display the color on your screen using the more accurate and consistent LAB values.

But the question is, what about composite proofing? I understand that I can use OPP to SEE my colors in a consistent way on screen, but how can I also get color proofs without color shifts? The answer is...

SIMULATE OVERPRINT

I always wondered why if I create a file without any overprints in it, I can go to the Print dialog and see that the Simulate Overprint command is available. If there are no overprints present, the item should be grayed out, no? The answer is that when you turn on simulate overprint in the Print dialog, the same thing is happening as we discussed above with Overprint Preview. The printer is using the LAB values for the color, not the CMYK values.

So the second thing we've learned is, if you want to see consistent spot colors on your composite proof prints, turn Simulate Overprint on in the Print dialog box. Again, this feature is available in AI, ID, and ACRO.

If you use OPP and Simulate OP, then you never lose the CMYK definition of the color that was specified by the book manufacturer, so it's always safe to convert to process at any time in the workflow if necessary.

In closing, you can ask why AI or ID even allow users to switch their spot colors to use only the LAB equivalents at all, if based on what you read here, it's such a bad thing. The answer is that in a color managed workflow, printers or RIPs may be able to better match spot colors to specific press or paper stock conditions when going straight from LAB. So I would strongly suggest that if you do want to use the methods described by others (such as Steve Werner's suggestion which I've linked to above), you first speak with your printer -- or you make sure to never convert your spots to process.

16 comments:

Anonymous said...

Oh my! I never realized Over Print Preview was able to do this. incredible. I'm all for colormanagement (Certified color proofing) but this in brand new information for me. Thanks!
Woz
www.macmojo.nl

Anonymous said...

Mordy,

Thank you for this wonderful post. I had actually heard recently that LAB was THE way to go with spot colors since Photoshop did this by default and so I had gone ahead and modified my ID defaults to always display spot colors using LAB values. I also had done my best to get this to be the default in ILL CS2 (although it doesn't seem possible to set this as a default setting for new docs).

I actually had noticed that Overprint by itself seemed to show spot colors much more accurately (without even having LAB turned on) and now it all makes sense that it's actually showing the Pantones with LAB values instead of CMYK.

Thanks to your post, I'm now going to disable my default of rendering Pantones with LAB and go back to the CMYK method. Overprint ON seems like the way to go now so you can get the best of both worlds --- accurate on-screen viewing and the ability to keep the CMYK values for spot-to-process conversion.

Thanks!

-Josh Gordon

Mordy Golding said...

Thanks for the comments Woz and Josh.

While it's certainly possible to work in OPP, screen redraw is slower.

I should point out that Adobe engineers feel that Overprint Preview should be the DEFAULT preview mode -- and that it should be on all of the time. But you can blame the Adobe product managers for refusing to do this. Why? Because OPP is too slow. People don't enjoy waiting for screens to redraw, or for things to move slowly when dragging them across the screen. So the product manager's reply to the engineers is "make overprint preview as fast as regular preview, and we'll make OPP the default".

And so it goes...

Nikos said...

Amazing. Every time I visit this blog I feel like every piece I've sent off to the printer turned out ok by sheer luck alone.
There's no end to the detail you must see to to get something done correctly in Adobe's apps.

Thanks so much for enlighting our ignorant minds.

BTW how come Adobe decided that PS should behave differently? Isn't that a bit counterintuitive?

Mordy Golding said...

Great question nikos...

While InDesign, Illustrator, and Acrobat all share a single unified imaging model (based primarily on the PDF spec), Photoshop uses a different imaging model entirely. Displaying and printing pixels is different than drawing and printing objects.

Craig S said...

Mordy,

I tested Sim OP using a 50% tint PMS 3025 (a dark blue) from ID. When I printed to a Xerox 7750 color laser with Sim OP OFF, the color looked good - a lighter blue. When printed with Sim OP ON, the color looked horrible - a greenish gray.

To see what was going on I repeated the test printing to PS to run thru Distiller to a PDF. I then used Pitstop's tools to inspect the colors.

The Sim OP OFF PDF indicated a color definition of 50% 3025 with an alternate color value of:
50 C
8.4 M
0 Y
25.5 K

The Sim OP ON PDF color was defined simply as a CMYK value of:
47.1 C
31 M
25.9 Y
0 K

That's a big difference and explains the greenish gray. Am I missing something?

Thanks for all your time and efforts educating those of us in the trenches.

Mordy Golding said...

Hi Craig!

It appears that for some colors, there is a VERY wide delta between the eventual CMYK breakdowns of colors. Your example can even be seen on screen in Illustrator simply by turning on OPP.

I don't know enough about color (I'm slowly learning), but tint values seem to be VERY different when you compare how they are calculated in the CMYK color space or the LAB color space. The color you chose is a really good example of that.

In duplicating your example, I noticed that the difference between OPP and non-OPP with 100% of the color was slightly different, but the difference between OPP and non-OPP with a 50% tint of that color produced a VERY large gap. Here are my numbers:

Pantone 3025: non-OPP (CMYK -> CMYK)
100% color: C=100 M=17 Y=0 K=51
50% color: C=50 M=8.5 Y=0 K=25.5

Pantone 3025: OPP (LAB -> CMYK)
100% color: C=100 M=65 Y=38 K=24
50% color: C=50 M=33 Y=19 K=12

The second set of numbers (with OPP) is obviously what matches that found in Photoshop (since Photoshop also does the LAB-CMYK conversion).

As you can see by the numbers, the tint values for BOTH sets of numbers are really half of their full-color equivalents.

I guess the question is -- what more accurately describes what you'll see on press?

Anonymous said...

If I may add my 2 cents, these CMYK-tables of Pantone... what kind of CMYK were they talking about? For years Pantone offered the same CMYK-values for coated, uncoated and all other CMYK-values. This proofed to be extra faulty for printers outside the USA, because the CMYK-values they provided Adobe were based on standard Americain ink. It was not until recently Pantone decided to provide Europe with ISO12647-2 compatible colors based on ISO2846-1 ink. The NEW 'Solid to Process' flyer is now available as the 'Pantone color bridge coated EURO'. A 2nd Color Bridge UNcoated was released last year. THESE CMYK-values do take in account the paper you're printing on. ALSO they are based upon Computer To Plate process most printers use now a days. Pantone itself states 98% OF ALL PMS->CMYK colors have new values and people should go out and buy the new color bridge and NOT use the older ones. More information: http://www.gammag.com/WhitePapers/doc/PANTONE_bridge.doc

PS: Almost finished the Dutch translation, Mordy ;-)
PS PS: Hope this information was helpful for the not-usa printers. I've got a more detailed article but it's in Dutch (sorry).

Woz

patrick said...

I recall reading a post on AskPantone.com where Pantone stated that the only CMYK values they supply to Adobe were the ones available in the Adobe apps' "Solid to Process" swatch library.

I had always wondered why I sometimes got different sets of CMYK values when the same spot colour was converted to CMYK in Photoshop/Illustrator. Very frustrating when two objects that are meant to be the same colour are superimposed (especially as the difference usually becomes more apparent when printed).

I'm sure Pantone also stated that when swatches from the other available libraries (in Adobe apps) are converted to CMYK, the apps use an algorithm to calculate the values. The suggestion being, I suppose, that the Pantone-supplied CMYK values have all been visually proofed by living breathing human beings, and the Adobe "calculated" values have not.

Pantone's advice was that, when approximating Pantone spot colours in a job destined for CMYK output, you should only use swatches from the Solid to Process library.

(I note that this advice has now been removed from the Pantone site, I guess because some of the detail became irrelevant with the advent of Pantone Color Bridge. I believe the current equivalent to the "Solid to Process" library, for those of us that have our swatch libraries up to date, is the "Pantone Color Bridge CMYK" library.)

Whoever said achieving colour consistency was like negotiating a minefield? If nobody did, please, allow me. Designers need to know this stuff! And they don't!

Which brings me to your post, Mordy. This is the first time I've EVER seen anybody mention an Excel spreadsheet of LAB values. Valuable stuff, if I can get it straight enough in my head. Does this mean that Photoshop etc are translating the LAB values to RGB for display on screen when we use Overprint Preview? I know that LAB has the broadest gamut of colours and is thus the most accurate virtual model available for describing a printed ink – just not sure how this relates to what we are shown by our monitors, which are light sources.

Mordy, you're a wizard and a true star.

By the way, I'm in the UK, so although it may jar when some of you see "colour" spelled with a "u", it looks just dandy to me.

Mordy Golding said...

Yes Patrick, Photoshop is using a LAB value and using that to display colours (spelled special for you). Which is also what both InDesign and Illustrator do when Overprint Preview mode is turned on.

Anonymous said...

i've been working in prepress and design for 8 years here in Australia. And every time I think I've got the whole colour management cmyk/spot/screen/print thing sussed out, I read another paragraph of information and blow my mind all over again!
SCARY STUFF!

joel@goodiesgraphicdesign.com.au

Martin said...

Good Job! :)

dstudent86 said...

I'm a design student and really appreciate your post. I'm doing a folder project for a company and have used a lot of transparency effects, such as opacity masks. When I export or view overprint and view proof colors, the whole work gets a grey cast on it. I'm not sure what I'm doing wrong. Maybe I'm still using spot colors? Should I go to every element and change the color to one from the pantone solid to process library? What's the difference between process coated and uncoated? Also, I have three "pantone color bridge CMYK" options 1) EC 2) PC 3) UP and have no idea what the difference is. This is my first job and I really need to get this right. I'll be printing with uprinting.com and wondered what color book would be safe?
Thanks.

EugeneWhite said...

Been working in design for 5 years here in Ukraine. every time I think I've got the whole colour management cmyk/spot/screen/print thing sussed out, I read another paragraph of information and blow my mind all over again)

Luke said...

Hi,
I work for an events company and have a client who has re-branded, but the colours are throwing up a few issues. They are now looking to use Pantone 335C and 376C, however the RGB and CMYK values they have quoted are different to those quoted by a number of our printers (who incidental, have all provided different figures). Are you able to shed some light on whether this is normal (a range of RGB and CMYK values for a single pantone - and if so why? or whether we are going wrong somewhere. It'd be great to share findings with the rest of the company!
Thanks in advance.
Luke

Luke Power said...

Hi,
I work for an events company and have a client who has re-branded, but the colours are throwing up a few issues. They are now looking to use Pantone 335C and 376C, however the RGB and CMYK values they have quoted are different to those quoted by a number of our printers (who incidental, have all provided different figures). Are you able to shed some light on whether this is normal (a range of RGB and CMYK values for a single pantone - and if so why? or whether we are going wrong somewhere. It'd be great to share findings with the rest of the company!
Thanks in advance.
Luke