In
the beginning of the project, the texture mapping and
2D painter group already started to gather material and designed
textures according to the reference source. This information along
with the final model and lighting was passed on to the material
and texturing group, where everything would be integrated, then
transformed the animatics into 3D model with complete texture
settings. Now, the computer will start to render images scene
by scene (fig. 10) Every rendered image must meet the client's
expectation and approved by the supervisor before moving on to
the next step.
Animation
Among our 3 teams, the animation team probably had the most complicated
work of all. Unlike the other 2 teams, the animation team, lead
by Chris Chen, does not always divide into the same number of
working groups. For this BMG MTV project, the animation team was
divided into 5 group:
1. Camera movement, 2. Character Setup, 3. Motion Capture, 4.
Facial Animation, 5. Character Animation.
The camera movement group first took the animatic elements and
designed camera movements according to the story board. If there
was any interaction between characters or objects, the timing
needed to performed those actions must be reserved. The character
setup now bind skin the animatics with motion capture data, and
adjust the relationship between camera movement and subjects.
The character setup group was also responsible for creating each
character's skeleton setting; this entire process only took about
1-2 days to complete. By applying the motion capture data and
camera movements, the animatic is completed and ready for a rough
cut. At this stage, we already have a good idea about lighting,
camera movement, and character motions.
The character setup group's next job was to finalize all the character
settings (fig. 11). We were roughly at the mid-phase of the project,
when the modeling team completed a usable model for each character.
Modeler will spend a lot more time in detailing the facial and
muscle shapes, to reassures the texture quality is closest to
a very human being. However, this does not affect the character
setup group. Character setups can be made with the rough models,
we will just have to make sure the character's number of polygons
and the control points are the same as the final models. All the
character setups can be applied henceforth, including skeleton
binding, paint cv weight, etc.. ( Chris Chen has developed a program
that enabled us to copy a character's bind cv weight on to the
final model. ) The character setup group also had to set the interactions
between cloth, hair and the character; this would include the
dynamic setting of collision and gravity.
fig 11. hair setup
|
fig 12. hair and body setup
|
fig 13. hair and body interaction
|
fig 14. hair and body setup
|