This is post three in a series about my experiences in publishing a book. The series starts here.
In the 90s, when I was working at IBM as a color scientist with a special interest in color management, I worked with engineers from Kodak, Adobe, Xerox PARC, Apple, and hp on what we called in those days device-independent color. We all seemed to share a common vision of the future. You’d do your editing in some colorimetric color space, and then send the file, tagged with a profile to indicate its color space off to a display, or printer, or printing press. Before it got rendered, software would look at a color profile for the output device and at the file’s color profile, then it would map the colors in the input file to the colorants (electron beam currents, inkjet dot counts, whatever) of the output device, mapping colors outside the gamut of the output device according to rendering intents specified in the input file.
That’s pretty much how it works today for inkjet printing. Hold that thought.
In the 90s, there were several ink systems introduced that supplemented the traditional four offset printing process colors, cyan, magenta, yellow, and black (CMYK) with other colored inks. One such system was developed by Pantone and called HexaChrome; the two added colors were orange and green. The press that Hemlock is talking about using for my book is a six-color press.
My Epson 4900 uses orange and green inks to extend its color gamut.
Based on all that, I thought that Hemlock would supply me with a profile for their press so that soft proofing in Photoshop and Lightroom would work right, that they’d use six process colors, and that I’d send them files in my preferred editing color space.
Wrong on all counts.
First off, Hemlock uses only four process colors. The extra two inks that could be used in the press are for spot colors. Second, they want the files in CMYK. However, it’s not the CMYK native color space of the press, it’s a standard CMYK space that comes with Photoshop called “Coated GRACoL 2006 (ISO 12647-2:2004)”. It’s unlikely that the native space for their press and whatever paper is in it is exactly that standard space, so they must be tweaking the press so that it matches the standard. This turns color management on its head; rather than converting the colors in the image to colorants using a press profile, Hemlock is making the colorants of their press match those of a standard profile.
Poking around the web, I find that what Hemlock is doing is common. Not universal, certainly, but common. This scheme gives up some press gamut in the interests of standardization, but has the (small, I would think) advantage that they don’t have to distribute color profiles. It also finesses the issue of how to ensure that the gamut mapping software of the customer is executing the same algorithms as the gamut-mapping software of the printer, since the printer isn’t doing any gamut mapping.
In the case of Hemlock, they want the images in Coated GRACoL 2006 regardless of which paper and paper gloss I specify, whether we’re using 10 micron or 20 micron plates, and whether the images are varnished or not. All those things will influence the color gamut of the printer/paper combination, and will affect the mixture of colorants necessary to get any in-gamut color in the output.
This is not a color-managed workflow on the sense that we color engineers envisioned it 20-some-odd years ago! Hemlock must be counting on tweaking the press to get close enough.
Well, I’ve seen their work, and it looks very good indeed. I’ll proceed ahead expecting that mine will, too. However, this whole scheme gives the engineer in me more than a little nervousness.
I thought I’d export a bunch of files from Lightroom in the Coated GRACoL 2006 color space. Can’t do it. Lr doesn’t support any CMYK spaces for export. I wonder why they have this limitation, since the color space conversion can be done with the same Adobe color engine that handles color space conversions in Photoshop. The cynic in me says that they did it on purpose, since CMYK = pro, and pro = able to pay more, and, when you put the two together, it means that Adobe thinks that anyone converting to CMYK ought to pony up and buy Photoshop.
About 4 years ago I asked a leading Bay Area color management expert if there were any presses in this area that had color managed their press (not just the proofer). He said he’d have to get back to me, and a few days later called and said, “no”. I was surprised. Bill Atkinson got unbelievable results in his book “Within the Stone” by profiling the press at Vanfu in Tokyo. But the “old world craftsmanship” ethos among press people persists.
Charlie
“However, this whole scheme gives the engineer in me more than a little nervousness.”
I wonder how they feel with a client with your credentials coming at them?
I also think the comment by Mr. Cramer above is very interesting, and very telling.
And count me as one who might also be interested in getting a copy of your book, depending on cost(and I still want to find out how much a print of this one image of yours might be—not one in your list of available ones). As for me, my book dreams have to do with helping to market my own work to galleries, & etc. But I am watching these posts of yours because of my involvement in another project, an artist’s work for which I’ve been chief curator for the last 5-6 years—we’d like to get our own book out.
I think a lot of it must has to do with the fact that even with perfect profiling throughout the systems it’s still very, very difficult to get wysiwyg third party contract prints – that is, what you see on your screen is what you get in their print.
Too many variables, starting from something as simple as illuminance vs luminance, color temperature and viewing environment. There must be too many cases of folks who thought they had the perfect soft proof off their monitor from a fully end-to-end profiled system in their nicely lit studio who are disappointed when they look at the final print under natural light.
After too many ‘I don’t understand why it looks so awful, my system is perfectly profiled’, printers decide to disconnect, use their crafty side and go perceptual: ‘doesn’t it look good?’ – even though it is not tied colorimetrically to your original. We look at it and say, ‘Yeah, it does look good, you are a good printer. Not like on my perfectly profiled monitor, but still good. I’ll take it’.
Jack