Adobe Photoshop Training with Dan Margulis

Hi, a question, ironically I've had two jobs sent recently, a light skinned person and another that has awkward baby skin tones, which have made me stop and think! The two wet proofs that I'm about to have output are now a major concern. So to cut to the chase, can the magenta be higher than yellow when colour correcting these types of skintones, especially in the palm of the hand (light skin tones), and in the face of (baby skin) blotches?

Many thanks,

Christian Macey

___________________________________________________________________________

Hi Christian:

If the white and gray areas do not have a color cast I would say that the skin color you see is most probably the correct color. Babies are notably "pink skinned" in the first few months. Blemishes and skin blotches are also reddish as the body is providing an extra amount of blood to heal the malady. Very light skinned people tend to go to more of a pink (magenta) skin tone. I would balance the overall color to eliminate any color cast and print it as is. Rubbing your cheek bones is cheap rouge :)

Michael C. Demyan

Fine Photography & Digital Graphic Design

www.mikedemyan.com

www.pbase.com/mdemyan

___________________________________________________________________________

Michael, cheers for the malady info. Just to clarify there was nothing in the original pictures that I would bet all of my two Christmas presents on would be 100% neutral. I also had to assign a false profile 1.4 or 1.0 Gamma to open one of the pictures up which concerned me even more, because both pictures had poor lighting. That's my excuse anyway.

<g>

Thanks again,

Christian

___________________________________________________________________________

Hi Christian, happy new year from Oz.

Yes, [assuming SWOP type conditions] it is fine to have higher magenta values in certain key areas, or even in 'average' areas - evaluating skintone is a tricky thing and depends on many factors, with the sample point being critical to evaluations or mis-

evaluations (no shadows, makeup, certain body areas etc). I find both the numbers and a good display help, as even relatively small moves of around 3% can make or break the edit (this is often where edits in final output space are often better than hoping that an colour transform will do the job).

It all depends on how dark the skintone needs to be, as well as the hue. Cyan can often range from Dans recommended 1/5 - 1/3, and even 1/2 the cyan value of the magenta value can be fine for a very deep caucasian tan. The yellow ratio compared to magenta is also very flexible, with lesser/equal/greater proportions being quite acceptable - depending on the subject and other variables. The basic rule of thumb of say a 10% higher ratio of yellow over magenta for a deep tan is a good place to start.

The points you mention are not ideal for evaluating an 'average' skin reading using a large sample reading. The palms of the hand will probably be more magenta than other areas, as will a blotchy baby complexion. These are not ideal sample reading points.

Very young caucasian children often have higher ratio of green/magenta to blue/yellow when compared to adults, with the red/cyan ratio being either too high or too low to the often desired 'ideal' (often more pronounced in fair haired/skin kids than say brunettes).

Channel mixer and or apply image are great way to blend the channel data, often better than curves for this type of work (replacing/mixing good channel content into bad, as opposed to curving existing bad channel content), often using colour blend mode and or a layer mask based off the red/cyan channel or using other methods to help isolate the edit to the skintones.

There will likely be many variations to the hue of the skin over the image, which is why it is a good idea to use a few sample points in average areas which are not misleading - and to have a good display preview of the edits being made. By all means use the numbers and ratio guidelines - they are a good guide, but that is all they are, one has to use judgement.

Stephen Marsh.

___________________________________________________________________________

Not knowing what the original image looked like puts one in the "dark" as to a proper response to a problem. Sorry!

I can only imagine your dilemma when you first saw theses images. Why is it that these "so called photographers" do not understand the essence of photography? Lighting - Lighting - Lighting! Photography is another word for "Light Capture". Modern cameras all have light meters built in, and the cheapest point and shoot digitals have a preview screen. If you are unsure of the capture - TAKE ANOTHER!

My Word.

I sympathize with you and all those that are presented with unusable images and told to "make it work" :(

Michael C. Demyan

Fine Photography & Digital Graphic Design

www.mikedemyan.com

www.pbase.com/mdemyan

___________________________________________________________________________

How true that is. But most images need some kind of correction. You can just send images from camera to whatever output (not correcting the image(s) any further). But you can improve on those images in an image-editing program. Whether it's for inkjet, wet printing or print. Image-editing programs, "I" like to use it as a tool, not as a band-aid.

Today, photographers have_control_over the conventional process, not only for print. The wet process also. Which they can enhance contrast, color in that image. Though the use of curves, channel blending or any of the other commands that program has to offer for editing. Retouching, btw, is not considered a band-aid (for me).

I remember when a brand and model of camera (professional), the images where too magenta (skin tones) on a lot of images. I guess its better to have more red than cyan. You had to know how to handle those either with an image-editing program (channel-blend) or messing with the cameras settings. The next generation of that brand and model had changed that problem.

I would not put too much faith in built-in light meters. Even though they' re pretty good. Built-in light meters can be fooled. My saying: Fool me once, shame on you. Fool me twice, shame on me. I use a hand-held light meter. I feel the same about the screen preview as well. I had noticed when taking shots (professional camera) with certain colors in them. 15c,94m,94y comes to mind. Those colors on the preview were more like, 15c,82m,94y. That 's guessing at the numbers, but you could see the difference (color it should have been in the preview). When the images opened on the monitor (calibrated of course) in a neutral-balanced (well-behaved) color space, you saw red not orange. What is should have looked like in the preview. But the preview is pretty good.....blinks, composition. Those previews are not like monitors. I'm not trying to be funny here. Really. But to me its like playing a video game on Game boy (tiny screen), and playing the same companys' game on a (well-behaved, well calibrated) HD-TV.

John Opitz