Motion Capture

1. Inertial and magnetic motion capture systems capture 'joint angle' data
over the frame/time. Optical systems tend to capture" joint position"
information. What are the differences and advantages to each capture
approach?

2. Motion capture data files require some 'Human skeletal model' for a
character.What do you need to know to describe the skeletal model? How would
represent the root, and how does the choice of the root affect the overall
capture?


PIC/FLIP

1. In PIC/FLIP method, particles represent mass, momentum and/or energy.
Particles transport these quantities from cell to cell in grid at each time
step. Each particle carries two information: position and velocity. In PIC,
at each time step, velocity is calculated at the position of particle from
the grid by interpolation. To reduce the numerical dissipation, FLIP has a
small difference from PIC in terms of velocity information. Please explain
this.

2. Marching cube (3D) or marching square (2D) are methods to generate iso-
surface or iso-contour from scalar distance or density fields. Also there
are several methods to generate scalar distance field from unorganized point
clouds. Since PIC/FLIP uses particles, it would be a good idea to generate
free surface (water surface) out of these particles. In the paper 'Animating
sand as a fluid', surface reconstruction method is explained. Please
summarize it briefly.


Fluid simulation:

1. How do level set surfaces differ from mesh surfaces? List one advantage
and disadvantage of using a mesh to represent a surface.  List two
advantages and disadvantages to using a Level Set Surface.

2. For the fire simulation approach given in class, what is the flame
front?How can we implement advection and projection such that mass and
momentum is conserved across the flame front?  What do you expect to happen
if we did not maintain mass and momentum?

3. Smoke, fire, and water can all be simulated using fluids.  Briefly
explain how
they are similar and how they are different.

4. Why might one want to implement a water surface using a combined approach
(level set + spherical particles).  What would be the effect of only using
particles?  Of only using the level set?