lower cyan without increasing magenta?

[anderson234]

Hi,
I've learned several tutorials of Lee Varis. Regarding his method of adjusting/controlling skin color with CMYK values, I have a question: say with selective color tool, I first lower magenta in Reds, then lower cyan if nessisary, but that means increasing red, and as a matter of fact magenta will increase,too, thus spoiling my magenta control.
What should I do?

[Tony W]

Re: lower cyan without increasing magenta?

If this is purely for editing images within PS for colour correction I would suggest ignoring CMYK completely until such time as converting and presenting for press.

View this article by our own Andrew Rodney for a sensible approach
https://www.youtube.com/watch?v=dWaFDKrNrwc

[andrewrodney]

Re: lower cyan without increasing magenta?

Ignore CMYK; it's an awful, output color space for this work! Stick with Lab (or RGB):
Here's a video on correcting skin tones without having to resort to CMYK:

Low Rez (YouTube)
http://www.youtube.com/watch?v=dWaFDKrNrwc

High Rez
http://digitaldog.net/files/SkinToneVideo.mov

[Sabrina81]

Re: lower cyan without increasing magenta?

Ignore CMYK; it's an awful, output color space for this work!

With this technique, CMYK isn't used for output. Just to clarify.

[Sabrina81]

Re: lower cyan without increasing magenta?

Hi,
I've learned several tutorials of Lee Varis. Regarding his method of adjusting/controlling skin color with CMYK values, I have a question: say with selective color tool, I first lower magenta in Reds, then lower cyan if nessisary, but that means increasing red, and as a matter of fact magenta will increase,too, thus spoiling my magenta control.
What should I do?

When I lower magenta enough to reduce the red in skin tones, I then need to lower yellow to counteract the green that this introduces -- and maybe increase black a bit. Like this:

[Tony W]

Re: lower cyan without increasing magenta?

In these case the use of HueSat adjustments maybe a better choice than Selective Colour as you will reduce the red content (or indeed any other colour) without usually adversely affecting the others.

If you have a crossed curve type situation or wish to create one then Selective Colour may be a reasonable choice.

Example shown just altering the Red from HSB layer.
EDIT: Forgot this - you mentioned removing Cyan well this is -100 Cyan HSB layer. Not measured other values but Cyan gone without any observable effect on rest of image

[Shoku]

Re: lower cyan without increasing magenta?

A magenta adjustment and then a red adjustment - I used your posted jpeg.

[anderson234]

Re: lower cyan without increasing magenta?

Thanks guys, very helpful.

When I lower magenta enough to reduce the red in skin tones, I then need to lower yellow to counteract the green that this introduces -- and maybe increase black a bit. Like this:

-magenta = +green
-yellow = +blue
+green + blue = +cyan = -red

I've seen this method before and thought about it, but can not figure out: why not simply reducing red ?

[Guest]

Re: lower cyan without increasing magenta?

Ignore CMYK; it's an awful, output color space for this work!

Aside from this work do you think CMYK can still have a point when retouching for pre-press because of the density concept and the "Maximum Dot" and "Minimum Dot" in print specification?

I believe it's useful to take total ink readings in the palette just to see how the image will print. It's just one value (Σ) to assess the image.

I've seen this method before and thought about it, but can not figure out: why not simply reducing red ?

I am not familiar with the selective color adustment. Yet, why not just operating on the midtones slider?

[Shoku]

Re: lower cyan without increasing magenta?

Aside from this work do you think CMYK can still have a point when retouching for pre-press because of the density concept and the "Maximum Dot" and "Minimum Dot" in print specification?

Yes. We've been retouching in CMYK for more than two decades. We also use RGB and LAB as needed. CMYK is very intuitive and has some benefits the other color spaces don't. Our output is for press, so we would naturally convert to our output CMYK anyway. Using it for editing has never slowed us down or made the process more difficult, in some cases it makes things easier.

[andrewrodney]

Re: lower cyan without increasing magenta?

With this technique, CMYK isn't used for output. Just to clarify.

The scale of the numbers is always based on some specific output and that's the issue with using a silly color model based on many differing output devices and thus scale and ratio. Absolutely not the case with Lab.
These two sets of RGB values have four different CMYK equivalents due to four different CMYK profiles; which is correct?

http://digitaldog.net/files/CMYKs1.jpg

[Tony W]

Re: lower cyan without increasing magenta?

Aside from this work do you think CMYK can still have a point when retouching for pre-press because of the density concept and the "Maximum Dot" and "Minimum Dot" in print specification?

I believe it's useful to take total ink readings in the palette just to see how the image will print. It's just one value (Σ) to assess the image.

CMYK changes are best reserved IMHO to deal directly with press variables and solving their specific problems.

RGB deals directly with our photographic variables and are usually the best for such problems.

I believe that anything that can be done in CMYK has at least an equal or even better alternative approach in RGB. Not everything that can be done in RGB can be repeated in CMYK

RGB is a wiser choice than editing based on the needs of one very specific CMYK output device - you are pretty much stuck as sending the data to any other (including CMYK devices) is likely to lead to a world of pain. Whereas work in RGB (preferably the wider spaces) and convert to virtually any output device know to man - providing you have the necessary profiles.

Of course just my opinion and I would be open to seeing objective testing to prove or disprove

EDIT

I knew I had seen something in the past - I think this fairly succinct

http://rgbcmyk.com.ar/en/rules-for-c...mages-to-cmyk/

• Rule 1: DO NOT convert to CMYK.
• Rule 2: If it is not possible to apply Rule 1, then DO NOT convert to CMYK yet!

[Guest]

Re: lower cyan without increasing magenta?

I'd love to understand if being aware of Ink limit (Total Area Coverage) is part of a retoucher "best practise" or if this is just a subject that the prepress people must face.

What if the retoucher is being informed by the publication or the photographer that the Ink limit should be under 260?

Should the retoucher convert to a custom CMYK profile (GCR, Medium, 260%) and then assign a known working profile (swop)?

[Shoku]

Re: lower cyan without increasing magenta?

Not everything that can be done in RGB can be repeated in CMYK

True. And it is also true that every color space has it's limitations. That's why we use CMYK, RGB, and LAB. We use what works, and that can vary from image to image.

[anderson234]

Re: lower cyan without increasing magenta?

I could learn a lot from your discussion. I appreciate it.
But at this point, I care most about the numbers/pattern that Lee Varis's CMYK method showed us, which directly helpes me quickly adjust skin tone to normal. Maybe RGB or LAB is better, but I need that pattern, too. I'll try to find it within RGB and LAB.