Facial Skin Layered
Mapping
Part 3.6a – Mapping The Face:
Creating the Base from an Image
Before You Begin
Load the reference images (front and side) and also
SkinFrontSectionWork000.psd into Photoshop.
Introduction
The front/side images will be used in two ways: to
create a base layer for the skin color map; to create a base layer for the skin
wrinkle bump map. This method was used
for the layered maps used for the model shown in figure 3.6a.a.

Figure 3.6a.a
The reference images used are show in Figure
3.6a.b.

Figure 3.6a.b
Note 1: The following instructions are for Windows.
- Make
copies of the reference images to work with (it’s a good idea to write
protect the original reference images so that you never goof up and modify
them).
- Place
the reference images (the copies) next to each other and make them the
same size. Then, using the
Hue/Saturation (<ctl>u), get the images looking as much alike in color
and lightness as possible.
- Now
duplicate the Background layer and then delete the Background layer.
- Using
the tool of your choice remove all but the skin areas of both images. Save.
- If
needed, for each one select Image->Rotate Canvas->Arbitrary to make
the images level. Save and preserve
these.
- Create
a new image the same height but 3 times the width as the largest of the
reference images.
- Select
all on the front image (<ctl>a) then past this into the new image
and center this using the move tool (v).
- Do
the same for the side image and line it up as best as you can with the
front image so the nose tips meet.
- On
the side reference image do Image->Rotate Canvis->Flip Horizontal
and repeat step 8 for the other side of the face. You should wind up with something like
Figure 3.6a.1.

Figure 3.6a.1
- Now
duplicate the bottom layer (the front view) and move this to the top and
adjust the Opacity of this layer to about 38%.
- The
next step is to put a base layer below all this that contains just the
approximate color of the images. To
create this base layer do Image->Duplicate (<alt>i, d, enter) to
duplicate this image and then select the top most layer and then either:
a) ‘link’ the layers below it then do <ctl>e to Flatten Linked
Layers or b) do <ctl>e over and over until all the layers are
flattened.
- Next
do Filter->Blur->Gaussian Blur.
Set the blur radius to about 30 and press okay. You should wind up with something like
Figure 3.6a.2.

Figure 3.6a.2
- Now
duplicate this layer 6 or 7 times and flatten all these layers (be sure to
maintain the transparency – do not do ‘Flatten Image’ but rather flatten
each layer to the next one using <ctl>e).
- Repeat
steps 12 and 13. You should wind up
with something like Figure 3.6a.3.

Figure 3.6a.3
- Select
all (<ctl>a) then paste this into the image with the faces then move
the new layer to the bottom and name it ‘Base’. You can delete the work image you used
to create this layer. You should
now have something like Figure 3.6a.4.

Figure 3.6a.4
- Next
duplicate this image and set the duplicate aside. This duplicate is so you can go back and
get any of the individual layers you might want to use in the next
step. You might save this so that
you archive layers in case you need them downstream. Go back to the original image and link
and flatten the layers (again, be sure to maintain the transparency).
- Now
work to breakdown and smooth out the lines between the front and side
images. Every image is different so
exact steps are impossible to document but in general you do this by
adjusting the opacity of the layers and using the Clone Stamp tool with an opacity of
about 75 and the Smudge tool with an opacity of about 15 varying the brush
size as needed. Maintain as much of
the detail as possible and strive for a flattened face. The process may take a little time. While your at it, if needed, close the
mouth by duplicating the upper and lower lips over the opening and do the
same with the eyes. You should wind
up with something that looks like Figure 3.6a.5.

Figure 3.6a.5
- Now
its time to bring this into the.
Flatten the layers again leaving the transparency, select all and
copy. Move to the
SkinFrontSectionWork000.psd image, select the Color layer set (or position
just below the Color layer for version 5 people) and paste. Name the new layer Color Base.
- You
will need to fit the image to the mesh guide. To do this first make the Guide 2, Mask,
Color Base and Backing the only visible layers. Turn down the opacity of the Guide 2 and
the Mask layers so you can see the Color Base layer clearly. Now <ctl>click the Color Base
layer to select it and also select all the colored pixels on it. Then, go into free translate mode (<ctl>t)
and adjust the image as best as you can to the guide. You must line up the eyes, nose, mouth,
chin and ears. You may not be able
to do all this just using the translate tool and it is likely you will
need to cut some parts out in order to position and rotate them into
place. You should eventually wind
up with something like Figure 3.6a.6.

Figure 3.6a.6
- Once
you are happy with the fit… save!
Duplicate the layer and move the copy to the Bump set (or area) and
name it Wrinkles Base. Hide the
Bump set (or layer) and reselect the Color Base layer..
- The
color image should only contain color information and not texture,
wrinkles or shadows. But… texture
and wrinkles actually do have some color differences – the skin is
compressed and so the color is darker.
We’ll perhaps get a tad of this coloring from the Color Base layer
but most of it will come from an overlay we’ll be doing later. For now well just try to get the over
all colors in the Color Base.
- First
use a Gaussian Blur with a Radius of about 3.3.
- Next
using the smudge and clone stamp tools as before smooth out the ears and
get rid of shadows so that only color remains.
- Be
sure to review all of the image and get rid of any dark shadows using the
clone and smudge tools.
- Finally,
use the Gaussian Blur again with about the same radius as before. You should wind up with something like
Figure 3.6a.7.

Figure 3.6a.7
- Save.
- Hide
the Color set and unhide the Bump set and select the Wrinkle Base. In this image you want to remove the
shadows but keep the wrinkles and texture as much as possible.
- After
you’ve removed all the shadows, look at the texture and wrinkles. If any have softened or blurred, go back
the original image layers you archived or even the original front/side
images and copy the areas and paste them in to restore the textures and
wrinkles (be sure to smooth out the edges of the paste using the clone and
smudge tools).
- When
you are happy with this image, de-saturate the image
(<ctl><shift>u) making it a gray scale. You should have something like Figure
3.6a.8.

Figure 3.6a.8
- Not
that we have this base established, save then save to a new version
(SkinFrontSectionWork002.psd).
Go to Part 4.1 – The Bump Maps:
Finishing the Wrinkle Map