GRACoL 2006 Hard Proofing

[Jim Kasson]

I'm working on a book which the publisher wants in CMYK, specifically Coated GRACoL 2006 (ISO 12647-2:2004).

Yesterday, my designer asked me for the final copies of each of the images for the book. I had previously identified seven images for which the standard Adobe gamut mapping with relative colorimetric intent were not acceptable. I turned on Lightroom’s soft proofing feature and set the output medium to Coated GRACoL 2006 (ISO 12647-2:2004). I created virtual copies of the seven images. I made corrections with the adjustment brush, and exported the images to .psd files in ProPhoto RGB.

I opened the files in Photoshop, converted them to Coated GRACoL 2006 (ISO 12647-2:2004), saved them as TIFFs, and proofed them on my Epson 4900 on Exhibition Fiber paper. I had previously determined that the entire Coated GRACoL 2006 (ISO 12647-2:2004) gamut lies within the 4900/Exhibition Fiber gamut. I made sure that the Black Point Compensation box in the Ps print dialog was unchecked.

The prints looked good. Too good, in fact. I got suspicious. I measured some dark areas on the prints with a reflection densitometer, and got readings in the 1.8s. That seemed too high for offset lithography. I got a recent issue of LensWork, and measured the darkest spots I could find in the 4-color images: 1.66. Then I measured the darkest spots I could find in the duotone images: 1.76.

I created a PPRGB image in Ps, and filled it with 0/0/0. Then I converted it to Coated GRACoL 2006 (ISO 12647-2:2004), and printed it with the Black Point Compensation checked and unchecked. With the box checked, I got readings of 2.10 with the densitometer. With the box not checked, I got 1.88.

So, it seems that the collective wisdom of the profile and the Adobe color management software is that the printer can print as dark as a density of 1.88. If that’s true, that’s wonderful.

But is it true? Does anyone know the Dmax of Coated GRACoL 2006 (ISO 12647-2:2004)?

As a side issue, things would be a lot easier if Lightroom supported CMYK files. I’m not sure why they don’t. The program knows enough about CMYK files in general, and Coated GRACoL 2006 (ISO 12647-2:2004) in particular, to make soft proofing work. That means it can get from its internal color space to CMYK. Why can’t it get from CMYK to its internal color space?

Thanks,

Jim

[digitaldog]

I'd ping Don Hutcheson (let me know if you need his email) about the dMax. He's on the committee working on these spec's. He's your man!
LR does now allow CMYK out the Print Module as a JPEG. Probably not what you want. Otherwise, since it's an RGB processing engine, doubt it will every 'support' CMYK files other than what we have now.

[Jim Kasson]

I'd ping Don Hutcheson (let me know if you need his email) about the dMax. He's on the committee working on these spec's. He's your man!
LR does now allow CMYK out the Print Module as a JPEG. Probably not what you want. Otherwise, since it's an RGB processing engine, doubt it will every 'support' CMYK files other than what we have now.

Andrew,

Thanks for the tip. I'll PM you.

WRT Lr and CMYK, I'm not asking for CMYK editing; that would be a big deal to do. I'm just asking that it be able to deal withfiles that start out as CMYK. In my case, I don't really want to edit the images, just use Lr's Print module to deal with them in batches, which Ps isn't so good at.

I have used the Lr CMKY output before, and I am grateful for it as a bulk converter, but, because of the JPEGness, I almost always do bulk conversion to CMYK with a Ps action.

Jim

[LenR]

Hi James.
CMYK is a "color space". 
Gracol is a "profile" for one of many press/printing conditions.
You might want to consider a little consulting time with Andrew to get a deeper sense of what you're doing and how to successfully manage your project.
By the way, Photoshop is the primary application used by Desktop pros in the printing industry, not Lightroom.
Cheers
Len

[Jim Kasson]

Hi James.
CMYK is a "color space". 
Gracol is a "profile" for one of many press/printing conditions.
You might want to consider a little consulting time with Andrew to get a deeper sense of what you're doing and how to successfully manage your project.
By the way, Photoshop is the primary application used by Desktop pros in the printing industry, not Lightroom.
Cheers
Len

I take that as condescension. I think something that I said made you think I am not sophisticated in the field of color management. You might read a few of the following and see if you still think that's true.

Kasson, J.M., Nin, S.I., Plouffe, W.E., and Hafner, J.L., “Performing Color Space Conversions with Three-Dimensional Linear Interpolation, Journal of Electronic Imaging, vol. 4, July 1995, pp. 226-249.
Kasson, J.M., “Efficient, Chromaticity-Preserving Sharpening for RGB Images,” Device-Independent Color Imaging, Walowit, E., Editor, SPIE vol. 2414, pp. 134-145 (1995).
Kasson, J.M., “Efficient, Chromaticity-Preserving Midtone Correction for RGB Images,” Second Color Imaging Conference, Scottsdale, AZ, November 15-18, 1994.
Kasson, J. M., “Tetrahedral Interpolation Algorithm Accuracy,” Device-Independent Color Imaging, Walowit, E., Editor, SPIE vol. 2170 (1994).
Kasson, J. M., Plouffe, W.E., and Nin, S. I., “A tetrahedral interpolation technique for color space conversion,” Device-Independent Color Imaging and Imaging Systems Integration, Motta, R. J., and Berberian, H. A., Editors, SPIE vol. 1909, pp 127-138 (1993).
Nin, S. I., Kasson, J. M., and Plouffe, W., “Printing CIELAB images on a CMYK printer using trilinear interpolation,” Color Hard Copy and Graphic Arts, Bares, J., Editor, SPIE vol. 1670 (1992).
Plouffe, W. Kasson, J.M., Easy-to-Compute Non-Linearities for Efficient Encoding of Color, Society for Information Display International Symposium, Digest of Technical Papers, Volume XXII, May 1991, pp 814-816.
Kasson, J.M. and Plouffe, W., “Subsampled Device-Independent Interchange Color Spaces,” Image Handling and Reproduction Systems Integration, SPIE vol 1460, pp 11-19, 1991.
Kasson, J.M., Color Science for Device-Independent Color Reproduction, Society for Information Display Conference, Las Vegas, NV, May 1990.
Kasson, J.M. and Plouffe, W., Requirements for Computer Interchange Color Spaces, SPSE/SPIE Electronic Imaging Conference, Santa Clara, CA, February 1990.
Kasson, J.M., “Computationally efficient low-artifact system for spatially filtering digital color images,” USA 5,793,885, issued August 11, 1998.
Kasson, J.M., “Method and Apparatus for Tone Correction of a Digital Color Image with Preservation of the Chromaticity of the Image,” USA. 5,774,412, issued June 30, 1998.
Kasson, J.M., and Plouffe, W.E., Pryor, D., Nin, S.I., “Function Approximation Using a Centered Cubic Packing with Tetragonal Disphenoid Extraction,” USA 5,751,926, issued May 12, 1998.
Edgar, A., and Kasson, J.M., “Automatic Cross Color Elimination,” USA. No. 5,509,086, issued Apr 16, 1996.
Kasson, J.M., “Method and Apparatus for Interactively Indicating Image Boundaries in Digital Image Cropping,” USA. 5,473,740, issued December 5, 1995.
Kasson, J.M., “Color Image Gamut-Mapping System with Chroma Enhancement at Human-Insensitive Spatial Frequencies,” USA. 5,450,216, issued September 12, 1995.
Kasson, J.M., and Plouffe, W.E., “Tetrahedron/Octahedron Packing and Tetrahedron Extraction for Function Approximation,” USA. 5,390,035, issued February 14, 1995.
Edgar, A., and Kasson, J.M., “Display Calibration,” USA. No. 5,298,993, issued Mar 29, 1994.

Jim

[digitaldog]

I have used the Lr CMKY output before, and I am grateful for it as a bulk converter, but, because of the JPEGness, I almost always do bulk conversion to CMYK with a Ps action.

The entire JPEG output from print is ridiculous! How hard could it be to give us the option for a TIFF? I mean seriously, that's not big engineering. I've asked, the LR people at Adobe don't listen to me any more. Maybe Schewe can twist their arm.

[Jim Kasson]

The entire JPEG output from print is ridiculous! How hard could it be to give us the option for a TIFF? I mean seriously, that's not big engineering. I've asked, the LR people at Adobe don't listen to me any more. Maybe Schewe can twist their arm.

Andrew, you're preaching to the choir on this one.

Jim

[Jim Kasson]

Hi James.
CMYK is a "color space". 
Gracol is a "profile" for one of many press/printing conditions.
You might want to consider a little consulting time with Andrew to get a deeper sense of what you're doing and how to successfully manage your project.
By the way, Photoshop is the primary application used by Desktop pros in the printing industry, not Lightroom.
Cheers
Len

CMYK is not a color space, any more than RGB is a color space.

RGB is an umbrella name for a class of color spaces that share one thing in common: they correspond to the additive mixing of three primary colorants. Within the umbrella, RGB spaces may be generated by specifying the chromaticities of each of the primaries, the white point, and the nonlinearities (if any) associated with each primary. The primaries do not have to be physically realizable.

Similarly, CMYK is an umbrella name for a class of color spaces that share one thing in common: they correspond to the mixing of four primary colorants. The usual name for the mixing is subtractive, but that is an oversimplification; in general, the way the inks mix to form colors is more complicated than that.  Within the umbrella, CMYK spaces may be generated either by models, as with RGB spaces, or empirically, by tables. There have been many attempts to form model-based CMYK spaces, and some have been reasonably successful, but CMYK spaces are today almost always defined by lookup tables.

I should point out that colors in CMYK spaces are overdefined, in that there are many ways to specify a (perforce three dimensional) color in four-space.

Jim

[Jim Kasson]

RGB is an umbrella name for a class of color spaces that share one thing in common: they correspond to the additive mixing of three primary colorants. Within the umbrella, RGB spaces may be generated by specifying the chromaticities of each of the primaries, the white point, and the nonlinearities (if any) associated with each primary. The primaries do not have to be physically realizable.

Similarly, CMYK is an umbrella name for a class of color spaces that share one thing in common: they correspond to the mixing of four primary colorants. The usual name for the mixing is subtractive, but that is an oversimplification; in general, the way the inks mix to form colors is more complicated than that.  Within the umbrella, CMYK spaces may be generated either by models, as with RGB spaces, or empirically, by tables. There have been many attempts to form model-based CMYK spaces, and some have been reasonably successful, but CMYK spaces are today almost always defined by lookup tables.

In considering the relationship between RGB color spaces and CMYK color spaces, it occurs to me that there is an odd asymmetry.

You can have an RGB color space where the primaries are not red, blue, or green. In ProPhoto RGB, the R primary is a red color, and the "green" and "blue" primaries are not colors at all, just mathematical constructs. In CIE 1931 XYZ, none of the primaries is any color.

I know of no CMYK color space where the colorants are not at least rough approximations to the block dye spectra of cyan, magenta, yellow, and black. I think that's because there is no CMYK space (that I know of, someone please correct me if I'm wrong) that is intended as a pure editing space, uncoupled from the output device. I have not experience with current practice, but when I was in the business, people using a CMYK variant as an editing color space intended to send the file to a printer with those colorants. In other worlds, they were doing an end-around device independence.

You could imagine creating a model-based editing color space using a CMYK model. Such a space could have a respectable gamut (far exceeding the gamut of actual CMYK printers) if the model were based on pure block dyes and idealized mixing. But other than familiarity to the diminishing number of people who are more comfortable editing in CMYK than RGB spaces, I can think of no advantage (and several disadvantages) to this approach.

Jim 

[digitaldog]

You can have an RGB color space where the primaries are not red, blue, or green. In ProPhoto RGB, the R primary is a red color, and the "green" and "blue" primaries are not colors at all, just mathematical constructs.

Yup. I prefer to call these device values. Clearly not colors (we can't see them).

I know of no CMYK color space where the colorants are not at least rough approximations to the block dye spectra of cyan, magenta, yellow, and black.

Agreed. CMYK is an output specific color space so yes, it's based on something real.
As to RGB or CMYK as written, I like to call those color models. Adobe RGB (1998) or  SWOP TR001 are color spaces (based receptively on the models)

[Jim Kasson]

Yup. I prefer to call these device values. Clearly not colors (we can't see them). Agreed. CMYK is an output specific color space so yes, it's based on something real.
As to RGB or CMYK as written, I like to call those color models. Adobe RGB (1998) or  SWOP TR001 are color spaces (based receptively on the models)

I like the "color models" terminology. Much better than my "umbrella". I think it's OK to elide the fact that most CMYK spaces are table-based, not model-based.  In fact, the only way to specify a CMYK space in the ICC framework is with a table, right?

Jim

[LenR]

Hi Jim,

Sorry if I came across that way.  I meant no disrespect but there were a few things I read in your OP that raised some red flags. 

In the graphics industry RGB, CMYK, LAB color etc. are generally referred to as "Color Spaces" and the profiles we're using in those color spaces are referred to as a "Working Space".  Certainly everyone is free to make up their own names for these.
 
Regarding "Does anyone know the Dmax of Coated GRACoL 2006 (ISO 12647-2:2004)". 
There's a boatload of information available on the GRACol specification and judging from your academic experience you're probably aware of it.
You've chosen a high quality printer who may be using as much a 320% ink coverage for your book.  I would reach out to them with technical questions since they, more than anyone,  are most suited to give you meaningful answers that you can use. These people are going to bend over backwards to do a great job and I'm sure they would love to make your experience one that would generate repeat business.

Best of luck to you in your project.  Your photos are really cool and I'm sure everything will work out to your satisfaction.

Regards

Len

[Jim Kasson]

Hi Jim,

Sorry if I came across that way.  I meant no disrespect but there were a few things I read in your OP that raised some red flags.

Thanks. I've calmed down now, but it did seem to me that you were being patronizing. I will put that behind me.

In the graphics industry RGB, CMYK, LAB color etc. are generally referred to as "Color Spaces" and the profiles we're using in those color spaces are referred to as a "Working Space".  Certainly everyone is free to make up their own names for these.

First off, I'm not making up my own names. I'm using the terminology of Color Science. It is also the terminology used by the ICC.

Second, the three things that you lumped together are apples and oranges. CIEL*a*b* is an actual color space. If you give me a value in Lab, say 50,10,10, I know what color it is. The other two things, which, to use Andrew's terminology, we'll call "color models" don't work that way. If you tell me that an RGB color is 128,140,90, I don't know what color it is without knowing more. That additional information can be supplied either by reference, such as "sRGB" or "Adobe 1998 RGB" or by specifying primary and white point chromaticities and nonlinearities. Same with CMYK.

Third, although I haven't worked with the printing industry in a long time, I do remember attitudes and terminology that ridiculed and thwarted device-independent color. I can't tell you how many times people from Scitex (not Efi Arazi; he was gone by then, he was smart and far-thinking, and he got the message loud and lclear), Hell, and Dai-Nippon told me that device independent color was a pipe dream, that high-end scanners would always output files in CMYK, and that the stuff I was working on was sloppy, unserious, and just for toy color.

To this day, I don't think they've all gotten the message. I'm working with Hemlock, a well-respected printer. Rather than give me their printer profile and let me output the files in the printer's native space, they calibrate their printer so that it's accurate when fed files in GRACoL 2006, and throw away any colors their printer can print that don't happen to fall within that gamut.

Not only that, but hard proofing is pretty close to impossible, because I can change paper stock, paper finish, and varnish type, and the profile doesn't change. When I profile my own printers, every time I change paper or ink setting, I reprofile the printer. Hemlock doesn't.




 

Regarding "Does anyone know the Dmax of Coated GRACoL 2006 (ISO 12647-2:2004)". 
There's a boatload of information available on the GRACol specification and judging from your academic experience you're probably aware of it.

I actually have no academic experience with color science. All my research was performed in an industrial setting. I haven't found the GRACoL 2006 Dmax for anything but 100%K, but I'm still looking. I may have to look inside the profile itself.

You've chosen a high quality printer who may be using as much a 320% ink coverage for your book.  I would reach out to them with technical questions since they, more than anyone,  are most suited to give you meaningful answers that you can use. These people are going to bend over backwards to do a great job and I'm sure they would love to make your experience one that would generate repeat business.

Thanks, but my conversations with Hemlock have been frustrating. They know a lot about printing, but the people I'm talking to don't know much about color management. However, they may be able to tell me the actual press Dmax with the paper and varnish I'm using, and then I can get closer  with m hard proofs than I fear I am now.

Best of luck to you in your project.  Your photos are really cool and I'm sure everything will work out to your satisfaction.

Thanks. I appreciate it.

Jim

[digitaldog]

Rather than give me their printer profile and let me output the files in the printer's native space, they calibrate their printer so that it's accurate when fed files in GRACoL 2006, and throw away any colors their printer can print that don't happen to fall within that gamut.

Just speculation but they might be using a Device Link Profile workflow where the ask (demand) CMYK in some established form form everyone. Then convert CMYK to CMYK with the DLP. IF that's the case, I suspect getting the Dmax for Coated GRACoL 2006 might not be an accurate value. Ideally you would get some output from them with max density, then measure it yourself.
This DLP workflow is what enables them to get one flavor of CMYK from you and others, then end up with differing output based on paper options. Again, speculation but IF so, you need the actual printed output for the job you're aiming for. Then measure the actual Dmax and other attributes you need to uncover.

[Jim Kasson]

Photoshop is right after all.

I looked at the GRACoL 2006 gamut in Gamutvision.



The shaded gamut is GRACoL 2006; the wireframe gamut is PPRGB.

The GRACoL black point has an L* of 11.

I ran this little code snippet in Matlab:

>> xyz = lab2xyz([11,0,0]);
>> y = xyz(2);
>> density = log10(1/y);
>> density

density =

    1.8993

Thanks, Andrew.

Jim

[Jim Kasson]

By the way, from the GRACoL 2007 Print Characterization Chart :


 
"Some printers may suggest that their printing parameters differ from those listed here
 . For example a printer may suggest using a 200-line screen or vary their ink densities
 . While this may enhance print properties such as resolution, there can be a penalty if the same
allowances are not made in proofing and other areas such as file preparation
 . Print buyers need to understand that printers offering to print “better” than a specification
such as GRACoL 2007 or SWOP 2007 may well be able to do so, but files and proofs generated by that
printer may not be interchangeable with another printer, and vice-versa, without careful color management."

So in a sense, GRACoL is advertising itself as an alternative for those who don't want to fully understand and implement color management.

Is that too cruel an interpretation of the above?

Jim