Home Use for CMYK Separations

[Jeff F]

I'm currently working my way through the CMYK separations section of Richard's book for the second time. Are there any significant advantages to printing on an inkjet printer from a CMYK split image? More efficient use of ink possibly, or more accurate colors during print? Just wondering. I'm enjoying the section of the book regardless because it's furthering my understanding of colors and how they mix in photographs. Any thoughts would be appreciated.

Thanks,
Jeff

[Richard_Lynch]

Deeper color, control of your output...These are two that come to mind. If you haven't tried doing a duotone with the separation technique, give it a go.

In order to use the CMYK separations, you'll have to do some checking and will have to print from another program that handles CMYK. Depending on your printer or other utilities, you may be able to control your inkjet and stop it from outputting CMYK>RGB>CMYK...there is a section on printing which discusses that.

OK?

[Jeff F]

Richard,

Thanks for the quick reply. Could you expound a little on the "control of your output..." part of your answer. Would it not look exactly like what's on the monitor (assuming all is calibrated) and isn't that what you'd want? Or perhaps there is a lot of variation in inks used in home printers so that consistent calibration of the printer is impossible?

I know of no programs that I have on hand that handle CMYK printing, but I haven't reached your chapter devoted to printing yet. So I'll wait until then to worry about that. I will, however, try the duotone printing as soon as I get a chance.

Thanks again,
Jeff

[Stephen M]

Jeff there are two answers to your question, depending on how one reads it.

There is the simplistic technical viewpoint...unless you use a PostScript RIP, _any_ colour mode which is not RGB is going to be converted on the fly from an unknown RGB source to the RGB that the inkjets printer driver uses, before the colour is passed off to the internal separation 'black box' before the print heads. A common press CMYK gamut is often smaller than an inkjet, so bright rich blues/reds/greens are going to suffer compared to keeping the file in RGB. Non PostScript printer drivers like QuickDraw on the classic Mac OS and GDI on windows are sort of like a translation of the monitor to paper - so CMYK is a no go.


Then there is the practical view where anything goes to make the image better, and working in modes which are not RGB may be the thing that is needed sometimes, so a quick duplicate of the original image, separate it to CMYK - work the CMYK seps (often the K which does not exist in RGB), then convert back to RGB and blend as a layer in luminosity blend mode over the original RGB image (so that hue/saturation are not affected, only the luminance of the CMYK converted image is added).

RGB and CMY channels are often similar, it is the K or black channel/plate of CMYK that most folk are interested in CMYK for inkjet use (it can often be better than creating a selection of the black areas of a RGB file).

Hope this helps,

Stephen Marsh.

[Richard_Lynch]

Jeff,

I suggest you use the CMYK pdf from the book and do some testing to see what result you are getting with your printer/driver combo. The goal of the CMYK separation is to control the black, as Stephen said, however, I disagree that the printer has to be postscript to affect the result. I have outlined how to do this in the book.


Stephen,

My book skins the color horse in a few different ways. I understand and respect your expertise, but there are a few unconventional solutions that I use to get results--even when a stubborn driver won't accept CMYK seps. These can be things that most Photoshop users would never consider, yet they are based on solid theory from pre-digital color press work. You might want to take a look at the book ( http://hiddenelements.com ) -- or you can wait for the Photoshop version, which will be out in the fall. With that book you'll find a few additional surprises.

Instead of working with CMYK to manipulate blacks in RGB, I would, instead use my RGBL solution, which will prove more effective if sticking with RGB output (to an LED or other photographic light-process printer). In it you work with RGB channels as layers with a luminosity layer. This effectively lets you separate color control from luminosity. In the Elements environment where you "can't" use CMYK channels or LAB, this solution allows you to affect almost identical adjustments with less color trade-off and conversion damage.

Hope that helps!

[Stephen M]

Richard, I can believe that you can do some 'unconventional' things - and I would be interested if you can work around the basic printing pipeline offered by QuickDraw/GDI and perhaps Quartz (PDF based OS X). Infact, I would think that Chris Murphy and Andrew Rodney and a few other vocal names in the colour game would be very interested. <g>

There is a very simple test to see if your printer works like a PostScript printer in the way input colour modes and ink on paper is performed. I am not talking about the interpretation of vector data which is the major function of a PS RIP.

Create a CMYK mode file. Fill the document with the default rich black which is going to be something like 80c 70m 70y 100k. Inside this large area of four colour black, create a selection of a smaller circle or square - but fill this with 0c 0m 0y 100K.

Onscreen without softproofing, both look the same visual density. But one only has 100% ink cover, while the other has 300% or more. There are going to be density differences between the 100K area and the rich black background - even though the monitor is showing no difference.

When output to a PostScript printer - there should be a noticeable difference between both blacks.

A non PS device (any printer without a PS RIP) - will render the two different source objects as the same black.

There is quite some difference between a L channel and a skeleton black plate from a UCR separation - but one can edit the L channel to be similar. The interaction of Luminance blending is not as good as the working the plate directly in CMYK mode, but it can come close sometimes.

Luminance and Colour blend modes (and LAB, HSB etc) which separate colour from tone really change how one approaches editing - I would be lost without my layer blends.

Stephen Marsh.

[Jeff F]

Thanks Richard and Stephen. I appreciate all of the information I can get from experts on the subject. Richard, I'll give that pdf file a go as soon as I get a chance.

Jeff

P.S. Stephen, thanks for the additional information. The more I gather about CMYK, the more I realize that it's probably well outside my perview right now. I'm really just a beginner in the world of photo editing, and I probably should be concentrating on color corrections in the RGB colorspace (which I am). But I'm working my way sequentially through Richard's book and am currently on the CMYK section. I never pass up an opportunity to learn, so any advice you can give in any aspect of this subject will always be welcome and appreciated.

Thanks again,
Jeff

[Richard_Lynch]

Stephen,

Working within the limitations of Elements can be pretty enlightening, and I think there would be some drastic simplification of process and thinking should you give it a spin.

I am not arguing that one thing is different from another (K and L are not the same, no...but it also doesn't have to be different if you want it not to be)...I am arguing that CMYK separation is probably not going to be used in elements for color correction as there are not the same mechanisms. The CMYK separation in Elements is made manually -- not using channels in the conventional sense you would in Photoshop. Plates are generated as separate files. Perhaps you can surmise where that goes from there. The elements user is liberated from the standard separation forced by Photoshop settings...a custom black separation is the only solution, really, and I show how to do that...and the reader can customize separations however they would like -- straight from luminosity, with luminosity and saturation considerations, and with any ramping of any sort -- using curves or levels to control -- or both.

Conventional wisdom is sometimes its own worst enemy. I know of Mr. Murphy and Mr. Rodney and yourself, and none of us would concur on everything, except technical behaviors which pose limitations. Limitations can be a brick wall if you look at them and stamp your hooves. I think one thing that poses a limitation that few people argue with is Photoshop itself and its helping hand. It keeps users from understanding the process. In the good old days of single color presses, one made separations. They put down inks one at a time. With the accuracy of conventional paper gripping, it is possible to mimic 4 color by running a sheet multiple times through the printer. Any printer. It is also possible to test quite accurately for separation using a CMYK file and printing as a PDF to test the drivers -- and it is possible to get different output from different machines, programs and configurations. I don't know if you can say for certain how every combination handles separation -- that would be a claim more grand and generalized than I would feel comfortable with. I certainly wouldn't claim that I know every possibility -- but I know what I have tested and what works. I was actually surprised (pleasantly) when I was able to get my inkjet printer to pay attention to a CMYK separation I'd created in Elements (and ran from Quark) after hearing both that inkjets wouldn't do it and that Elements wouldn't support CMYK. The trick was twice as sweet. The printout retained my rich black tests, and printed them accurately. The printer is not a postscript printer.

I am aware of densities, rich blacks and all sort of other color wash having been a pre-press person, editor and designer for a photography book publisher, where I scanned, picked photos, set up the color work, and collected the files for output. Our shop was above a printshop where I was able to gather all sort of knowledge about the process from both ends.

My suggestion in the previous post was only that you have a look at the techniques in the book the forum is named for, and perhaps poke around a bit in Elements. I know the reluctance of PS users to look at Elements as I was one once. The whole question of CMYK and color handling becomes more elemental in Elements...and for me it was enlightening. Just as Blend If remains a mystery to many, CMYK separation is 'magic', and it need not be.