ARTS COLLEGE 694Z - Motion Capture
GREEN GROUP
Ying Chen, Mark Keck & JulieAnna Facelli
INTRODUCTION:
We used motion capture to tell a story through movement. Since we were telling a story about two specific dance styles, motion capture was useful in giving us data to show the difference between the two types of motion.
Why Motion Capture?
If we had used 3D or 2D
animation the weighted feeling of the ballroom dip and the frantic steps of
the DDR dancing would have been less effective.
-In the DDR dance sequence we did not use the original dancing data for the
center character, Charlie, because there were too many marker problems in the
data. Instead, we adjusted data from another section and put him on the machine.
If you compare Charlie’s dancing to the other two characters’ dancing,
the active DDR movement is much more effective.
Maya Trax Editor vs Motionbuilder
We used a combination of Maya Trax Editor and Motionbuilder to blend our motions. For the most part, we used Motionbuilder to change movement to fit the sequence. In the Trax Editor it was easy to cut out sections of movement that weren’t working. In the Trax Editor we had to line the character’s position, but Motion builder could blend their position into a new one.
What went right?
The preparation and storyboarding for the project was our group’s strong point. We planned the capture well, and got data that worked well for blending. At the final stages of our process it was easy for us to put the sequence together.
What went wrong?
The most challenging aspect of this process was our unfamiliarity with Maya. We decided on humanoid characters, so we downloaded models instead of creating them ourselves. These avatars presented many problems, including their size, their position and their many extra poly-surfaces.
What we'd do differently...
If we were to do this project
again we would simplify a lot. Although it was an exciting challenge to have
multiple characters, it was very difficult to match their motions, especially
in the ballroom dancing where the two actors were touching.
The step, that upon looking back, we wish we had spent more time on, was setting
up control rigs to anticipate how the motion would translate onto the model
with the skin on it. Since we only imported their skeletons into Motionbuilder,
we didn’t know that we would have so many problems on the next step in
Maya.
What we learned!
Throughout the course of
this class we came to an understanding of the caveats of Motion Capture. While
watching the RDT capture, the motions that would present cleaning problems were
obvious to us. We also learned a lot about the three pieces of software, and
how they interface. Going from program to program required a great deal of foresight.