Week 3 Skin Tones [manda]

This board is for the March 2009 Curvemeister 101 class.
mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Mon Apr 13, 2009 3:04 am

As skin tones are a subject dear to my heart, I've spent a long time playing with the exercises.  The pdf is brilliant - thanks for the resource.  I've made an action out of the guidelines which I'd need to tweak for public use if any one is interested?

I couldn't seem to download ganna's baby image, but I've used one that I found on the internet quite a while ago.  I use this image a lot to practise my skills and never end up at the same place. 

Some random thoughts:
~ for the first time I saw the difference skipping the image from RGB (most familiar space but more random results) /LAB (better tan tones) /HS (better coverage).

~ the gradient gives great 'even skin tones', but would I want this on all images - I'd need to be careful of creative choices as people don't have perfectly even skin e.g. one of my practise shots has a classic Mongolian child with tanned forehead / burnt cheeks / baby skin on the chin.

Question: I use the 'colour by numbers' skin tone approach watching the hue clocks carefully.  I've chosen to show the dials: display>other colour>working CYMK, as oppose to Mike's wgCYMK.  It works for me, but it took me ages to decide which 'cymk' display to use.  Any thoughts on this choice?

Workflow:
use the instructions in the pdf
experiment using the jpg 'skin tones' and creating my own from existing tones on the photo
final curve adjustment using the numbers (aim for C -24, M - 52, Y -60, K - 1) and adjust to suit the image.
apply a sepia filter for creative effect to warm up the overall image

The attachments are work in progress, it got to the point where I'd lost all sense of judgment and was creating the most unflattering tones imaginable! 

thanks Mandy

mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Mon Apr 13, 2009 3:08 am

Final image ... for the moment!


ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Mon Apr 13, 2009 11:27 am


As skin tones are a subject dear to my heart, I've spent a long time playing with the exercises.  The pdf is brilliant - thanks for the resource.  I've made an action out of the guidelines which I'd need to tweak for public use if any one is interested?


Most certainly I'm interested.


I couldn't seem to download ganna's baby image, but I've used one that I found on the internet quite a while ago.  I use this image a lot to practise my skills and never end up at the same place. 


I'll look into it.


Some random thoughts:
~ for the first time I saw the difference skipping the image from RGB (most familiar space but more random results) /LAB (better tan tones) /HS (better coverage).


RGB can seem random with the placement of Pins but it has more to do with the pixels you are assigning values to, if you hit odd pixels you get wild color shifts.


~ the gradient gives great 'even skin tones', but would I want this on all images - I'd need to be careful of creative choices as people don't have perfectly even skin e.g. one of my practise shots has a classic Mongolian child with tanned forehead / burnt cheeks / baby skin on the chin.

All the gradient gives you is a wider range to select the skin tone from.  The pins are quite small and hard to judge the actual color from.  The gradient gives you more choice and better matches overall IMHO.


Question: I use the 'colour by numbers' skin tone approach watching the hue clocks carefully.  I've chosen to show the dials: display>other colour>working CYMK, as oppose to Mike's wgCYMK.  It works for me, but it took me ages to decide which 'cymk' display to use.  Any thoughts on this choice?


I would say you should use any color space that makes sense to you that can also be translated to your finished color space with minimal or no loss of fidelity.  As for not using WgCMYK the same should apply.  I use the LAB space for the hue clocks since it makes the most sense to me.  I can use the others when necessary but I prefer LAB.

Workflow:
use the instructions in the pdf
experiment using the jpg 'skin tones' and creating my own from existing tones on the photo
final curve adjustment using the numbers (aim for C -24, M - 52, Y -60, K - 1) and adjust to suit the image.
apply a sepia filter for creative effect to warm up the overall image


The attachments are work in progress, it got to the point where I'd lost all sense of judgment and was creating the most unflattering tones imaginable! 


It is a wonderful image and thanks for sharing it with us. You appear to have hit the "time out" part of the work flow.  Your image needs to stand the test of time...over night should usually do for images that matter.  If.. after you come back to it... you feel everything you do makes the image look worse then you have probably done as much as you would want to.

Looking at the image there are 2 areas you might want to revisit.  Right now she appears to have been lit by a warm light source.  Her fingers are slightly cooler than her hand... might be normal... I don't know...and there appears to be a warm spot in the area where the upper hue clock is shown in my screen shot. Otherwise the image is just delightful..
Greg


mikemeister_admin
Posts: 4927
Joined: Fri Sep 20, 2013 8:29 pm

Postby mikemeister_admin » Mon Apr 13, 2009 1:39 pm

Thanks for such a detailed response, it is a wonderful image, but it isn't mine - I just nabbed it off the internet.

I've just realised that you are posting replies in the early hours of the morning (unless the times are out of sync), lets hope you are up partying or at least enjoying a bit of quiet time in the early hours of dawn.


ggroess
Posts: 5342
Joined: Wed May 24, 2006 2:15 am
Contact:

Postby ggroess » Mon Apr 13, 2009 2:51 pm

Early dawn quiet time...
I usually get up at 5:30 Am

Greg


Return to “Curvemeister 101 March 2009”

Who is online

Users browsing this forum: No registered users and 33 guests