When you work on skin tones in CM one of the biggest hurdles is knowing what "Skin Pin" to set. I have created a few resources for you to use and build upon.
The first is a JPG image that I have posted that has an extended range of skin tones for you to choose.
You will need the ID and Password to get to this image.
www.curvemeister.com/downloads/Class/greg/pinning_skin_tones/skin_tones.jpg
The next is a set of pins that I have posted in the Pins Area. Pins.txt needs to be copied to your pins folder.
The Jpg is available there too but in a smaller version.
http://www.curvemeister.com/forum/index.php?topic=1667.0
Lastly is the work flow...
www.curvemeister.com/downloads/Class/greg/pinning_skin_tones/pinning_skin_tones.pdf
A special thanks to Ganna for the wonderful example of needing skin tones corrected.
Greg
Pinning Skin Tones: Advanced Workflow
-
- Posts: 462
- Joined: Sun Jul 11, 2010 4:23 am
- Contact:
BTW, CM (in my case, I might be strange) puts the swatches directory in the curvemeister directory. That directory is hidden. That is a mystery because cm's files tend to be plain text files.
Hidden directory??? I hope I did not screw things up too much, I un hid them.
I assume that pallor would be good for babies and young children. Or at least a good starting point.
Hidden directory??? I hope I did not screw things up too much, I un hid them.
I assume that pallor would be good for babies and young children. Or at least a good starting point.
-
- Posts: 462
- Joined: Sun Jul 11, 2010 4:23 am
- Contact:
I tried the skin pinning technique.
1. I took the Caucasian pallor swatches and made a photoshop image with gradient.
2. I copied that image and pasted to a new layer in the baby jpeg. After I am done, I can delete the layer or make it invisible.
Now I try to select two points of the swatch to match two points of the baby image. Do I do that by comparing the L points in LAB? Theory, is that with the same luminosity that a and b should match.
Or do I do math and add the R,G and B numbers and try to match?
Here I tried something different.
After pasting the swatch into a blank layer, I ran CM on the composite image. . I marked a spot on the baby and the swatch where the L numbers matched. I pinned the swatch and set the pin mode to match image original color. I move the pin to a new location on the baby with the same L value. Perhaps (????) I made a mistake. I did not set the mode to hue/saturation. Since the brightness was the same, that does not make a difference?? It probably was set to H/S/B.
BTW thanks for the class it has been great. I have been in Amsterdam at the ons. Can I ask for assistance if I do not understand the moves?end of last week. I tried to finish, but I was too busy. I will do my work and compare to the solution.
Can I retry to thai image? Can you publish the solution?
1. I took the Caucasian pallor swatches and made a photoshop image with gradient.
2. I copied that image and pasted to a new layer in the baby jpeg. After I am done, I can delete the layer or make it invisible.
Now I try to select two points of the swatch to match two points of the baby image. Do I do that by comparing the L points in LAB? Theory, is that with the same luminosity that a and b should match.
Or do I do math and add the R,G and B numbers and try to match?
Here I tried something different.
After pasting the swatch into a blank layer, I ran CM on the composite image. . I marked a spot on the baby and the swatch where the L numbers matched. I pinned the swatch and set the pin mode to match image original color. I move the pin to a new location on the baby with the same L value. Perhaps (????) I made a mistake. I did not set the mode to hue/saturation. Since the brightness was the same, that does not make a difference?? It probably was set to H/S/B.
BTW thanks for the class it has been great. I have been in Amsterdam at the ons. Can I ask for assistance if I do not understand the moves?end of last week. I tried to finish, but I was too busy. I will do my work and compare to the solution.
Can I retry to thai image? Can you publish the solution?
- Attachments
-
- baby01-ss-jpg (248.8 KiB) Viewed 10796 times
-
- baby01-jpg (201.83 KiB) Viewed 10796 times
I tried the skin pinning technique.
1. I took the Caucasian pallor swatches and made a photoshop image with gradient.
2. I copied that image and pasted to a new layer in the baby jpeg. After I am done, I can delete the layer or make it invisible.
Now I try to select two points of the swatch to match two points of the baby image. Do I do that by comparing the L points in LAB? Theory, is that with the same luminosity that a and b should match.
Or do I do math and add the R,G and B numbers and try to match?
LAB will have problems with more than one pin you might want to try this in RGB mode
Here I tried something different.
After pasting the swatch into a blank layer, I ran CM on the composite image. . I marked a spot on the baby and the swatch where the L numbers matched. I pinned the swatch and set the pin mode to match image original color. I move the pin to a new location on the baby with the same L value. Perhaps (????) I made a mistake. I did not set the mode to hue/saturation. Since the brightness was the same, that does not make a difference?? It probably was set to H/S/B.
I'll have to try thi sbut if you do this in RGB you will get a better match.
BTW thanks for the class it has been great. I have been in Amsterdam at the ons. Can I ask for assistance if I do not understand the moves?end of last week. I tried to finish, but I was too busy. I will do my work and compare to the solution.
Can I retry to thai image? Can you publish the solution?
You can always give the Thai image another go around. I'll look and see if I can find a good solution to the Thai image...
Keep posting and I'll keep answering within reason...let's say another week for the class forum and anytime on the open forum...
Greg
-
- Posts: 462
- Joined: Sun Jul 11, 2010 4:23 am
- Contact:
I think you are right. RGB is a better color space for pins with multiple points. LAB is better (in my mind) for selecting the points.
I think this technique would work, it I used LAB to selected the points. (I could put hue clocks). Change to RGB mode staying within CM. Mark points on the swatch and pin the image with these points.
This seems to be to be much easier than setting foreground and background points in PS.
I think this swatch technique is good and better than pins for skins. It is less important or unnecessary for other corrections. I will continue to use this to fix critical skin tones. It is fairly easy since my grandchildren have Caucasian pallor skins.
I enjoyed the class and do not want to overstay my welcome.
I will try Thai one more time. I did a few other un-posted tries.
I think this technique would work, it I used LAB to selected the points. (I could put hue clocks). Change to RGB mode staying within CM. Mark points on the swatch and pin the image with these points.
This seems to be to be much easier than setting foreground and background points in PS.
I think this swatch technique is good and better than pins for skins. It is less important or unnecessary for other corrections. I will continue to use this to fix critical skin tones. It is fairly easy since my grandchildren have Caucasian pallor skins.
I enjoyed the class and do not want to overstay my welcome.
I will try Thai one more time. I did a few other un-posted tries.
Return to “Curvemeister 101 September 2010”
Who is online
Users browsing this forum: No registered users and 2 guests