Dan Margulis Applied Color Theory

CMYK Confusion

CMYK Confusion: What is Info telling me?
Posted by: "williamtheis"
Sun Mar 25, 2007 10:26 pm (PST)

CMYK is not my native colorspace so I am struggling somewhat to do what PP5 is telling me. I'm trying to make some Irish grass to be green (pp74) by looking in CMYK at the dominating ink, the unwanted color (lowest + K), and the middle ink. My Info palette does not read 30C80M90Y or anything like this. Instead it reads more like: C:72% M:68% y: 94% K:34%

Obviously this doesn't add to 100%... so 72% of 256 is 184Y, 68% of 256 is 174M, etc. Is this the correct way to get the comparison of middle to dominating/unwanted color or is there some setting in photoshop that allows me to skip this % thing and go right to CMYK in values not percentages?

Bill Theis
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: "J Walton"
Mon Mar 26, 2007 3:55 am (PST)

On 3/25/07, williamtheis wrote:

My Info palette does not read 30C80M90Y or anything like this.

Good! That sounds like an odd color of red - not good for grass.

Instead it reads more like: C:72% M:68% y:94% K:34%

This is an odd color of brown - also not good for grass! You need to take a lot of magenta out of that dirt before anyone will believe that's grass.

A grass callout could be 70-80C, 15-35M, 85-100Y, 0-20K, depending on lighting and content. The Cyan is typically less than Yellow and the Magenta is *way* less than either. The black and magenta will vary based on the black generation used when the image is converted to CMYK.

Obviously this doesn't add to 100%... so 72% of 256 is 184Y, 68% of
256 is 174M, etc.

Nothing about CMYK is supposed to add up to 100%, and while 72% of 256 might be 184 that has nothing to do with CMYK either.

CMYK numbers are based on percentage of ink. You can't have 256 Cyan; the most you can have of a certain ink is 100%. It's like hearing someone say they give 110% effort - it's not possible!

Solid Cyan is 100/0/0/0, not 256/0/0/0. 100% Cyan basically says that the printing press will use the most Cyan ink it possibly can in that area.

Is this the correct way to get the comparison of
middle to dominating/unwanted color

Yes, you can set your info palette to either HSB or LAB but that won't really help you because you'd have to learn how to interpret those numbers - just like you will need to do in CMYK. You can always find sample images with colors you like and use those as "good numbers." Whether you use LAB, HSB, RGB or CMYK numbers is up to you, but if you are eventually going to CMYK the logical choice is clear.

or is there some setting in
photoshop that allows me to skip this % thing and go right to CMYK in
values not percentages?

No, for the reasons stated above. Skipping the % thing in CMYK is like skipping the sugar in Fruit Loops - it is firstly impossible and more importantly undesirable.
--
J Walton
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: "Lee Clawson"
Mon Mar 26, 2007 6:30 am (PST)

First think about the 4 colors, CMYK, each being able to measure, at full strength (saturation), 100% or very light pastel and tints at the other end at 2 to 5%.

C&Y makes the green and M and/or K making the green (grass colors) look darker and/or browner. Think of CMYK like mixing paint colors instead of light (RGB). The M and K will go up in the shadow areas of the grass and even present throughout to give the color a bit of dimension.

Lee Clawson
2/\V/\7 Studio
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: "williamtheis"
Mon Mar 26, 2007 11:22 am (PST)

Thanks Lee & J Walton,

I have no real reason for CMYK in my workflow being one of those who are "helped" by a CMYK Epson printer that insists it is an RGB. Another topic. I would consider brief excursions into CMYK (especially wide gamut because I need to return to RGB after the excursion) to take advantage of the many advantages of K that Dan points out in PP5, such as getting the K to multiply into an RGB image, etc.

I appreciate all the comments about the % being a % of the ink used by the printer... logical. What is not logical to me is Dan talking throughout the entire PP5 book in terms of explicit values of C,M,Y, and K which he calls "points". Doing a quick look at his value for shadow (80C70M70Y70K which sums to 290) confirms that my RGB-centric logic (where 256 is saturation of a color and 0 is none) is flawed when applied to CMYK.

Let me digest this a little and reread the book now that I realize he's talking about %'s, like you folks have just described.

Bill Theis
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: "Henry"
Mon Mar 26, 2007 3:05 pm (PST)

It is just easier to conceive or think about color as a set of values in CMYK. Colors described in Lab and HSB would even be easier to visualize than would RGB values. Even with a well-calibrated monitor, being familiar with numerical values for color will only make things easier for you. For one thing, an adjustment made "by the numbers" may not be readily obvious on a monitor, yet show up in print. Likewise, a substantial adjustment made visually might translate to what you might think of as only a small change by the numbers. Excepting for web imagery, how an image prints is the ultimate concern, not how it looks on a monitor. There is very little of RGB that is intuitive beyond values for black, white and neutrals. Given the values of a CMYK color, one can visualized the color, maybe not exactly, but the values do give you a pretty good idea. It is not the same with RGB values.

Don't begin by memorizing a lot of colors - there are a lot of them. Try to get an understanding of the relationships in the values for basic colors, then see how these relationships change as these basic colors get lighter and darker. There isn't too much to actually memorize. It more like recognizing that the color values are telling you that it is a green, what kind of green is it, and if its light or dark, saturated or dirty. Much the same kind of thinking an artist uses when mixing paints.

Dan's writing on this subject is probably the best real world handling of this topic that one could hope for. Some experience along with some good guidance, and the light will come on, and it won't take you very long.

Henry Davis
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: Dan Margulis
Mon Apr 9, 2007 11:44 am (PST)

Before I left for Photoshop World, Bill Theis wrote,

I appreciate all the comments about the % being a % of the ink used by
the printer... logical. What is not logical to me is Dan talking
throughout the entire PP5 book in terms of explicit values of C,M,Y,
and K which he calls "points". Doing a quick look at his value for
shadow (80C70M70Y70K which sums to 290) confirms that my RGB-centric
logic (where 256 is saturation of a color and 0 is none) is flawed when
applied to CMYK.

Nothing (except convention) prevents you from using a 255-0 light-to-dark scale when in CMYK, or a 0-100% scale when in RGB. It's all cosmetics. In each space a darker channel equates to a darker composite. The red channel is similar to the cyan, the green to the magenta, the blue to the yellow.

In spite of the similarities, there is a tradition that RGB users call CMYK counterintuitive and CMYK-oriented folk consider RGB a crippled version of the real thing. The result is that people are afraid of the one they're not familiar with, when in fact it should be trivial to make the adjustment.

Those who are having this difficulty may face an uphill struggle in future years. As usual, I polled the audiences at Photoshop World on the question of what output is important to them: is CMYK output very much more important than RGB; is RGB output very much more important than CMYK; or can you frequently be asked to do important work in either?

This year, over several hundred people, it looked like a straight 40-40-20 split. That's about double the number of folks finding both outputs important as last year, when I would say it was more like 40 (CMYK)-50 (RGB)-10 (both). I would expect this growth to continue among professionals.

Dan Margulis
___________________________________________________________________________

Re: CMYK Confusion: What is Info telling me?
Posted by: "Alessandro Bernardi"
Tue Apr 10, 2007 10:56 am (PST)

Dan Margulis wrote:

In spite of the similarities, there is a tradition that RGB users
call CMYK counterintuitive and CMYK-oriented folk consider RGB a
crippled version of the real thing. The result is that people are
afraid of the one they're not familiar with, when in fact it should be
trivial to make the adjustment.

I agree completely, it's a question of numbers used normally.

In addition to this, I suggest, to those who normally use the RGB values to make color correction and have difficulty to use the CMYK ones, to do this:

in a CMYK image, if you make option+click on the gray bar in the Curves Command window, the IN and OUT values of each point of the curve will switch to the 0-255 values.

The gray bar will be inverted to the normal RGB position, with the black on the left and the white on the right.

It works also in the RGB images.

Maybe this could be more familiar to the RGB people...

Alessandro Bernardi