Sculpting for Normal Maps - Game Character Creation with Blender and Unity - page 106

The Purpose of Sculpting
Sculpting is a seductive method for beginning 3D artists, especially those trained in
traditional art. Setting up a new sculpting file is a technical process, but sculpting itself
allows you to take painter-like strokes to modify the shape of a 3D model. It is possible
to begin with a cube and sculpt a complex monster's head, for example (Figure 5.1), as
though you are working with clay.
Because of its ease and fun factors, many new designers wonder why all 3D models
cannot be created with sculpting programs like ZBrush.
The reason for this is model topology. Creating 3D assets with traditional methods like
box or poly-by-poly modeling requires artists to interact directly with a model's struc-
ture. Sculpting, on the other hand, often
displays models' surface features without
wireframes. This can cause users to create
a model with little thought to how they
will animate that model. If topology is
ignored, sculpted models may have topol-
ogy inappropriate for animation in areas
such as the eyes and mouth.
The other topology pitfall of sculpt-
ing is polygon count. Sculpting modifies
geometry by adjusting the positioning of
the model's vertices. To incorporate fine
details, models must have polygon counts
in the millions. As beautiful as sculpted
models are, they are often dense to the
point of being unusable.
These pitfalls should not deter you
from sculpting in 3D art. In fact, many
new features are being added to proprietary sculpting programs that help users translate
their sculptures into usable game meshes. However, knowing the realities of sculpting can
help you understand sculpting's place in the game art process. Sculpting is an easy and
fun way to add detail to low-polygon models. This is accomplished with a normal map
(see Figure 5.2), a special type of texture that tells the 3D program or game engine to dis-
play details on a polygon surface as though they were geometry.
The normal map is made up of a mix of red, green, and blue (RGB) color information.
Game engines and 3D programs interpret this information as directions for changing
how a face's surface normals are displayed, thus giving the impression that details are
built into the modeling topology. This can be useful for both rendering and using charac-
ters in games—you don't have to compromise art quality for technical limitations.

Figure 5.1
Sculpted ogre head
created from the
default cube