---------------------------------------------
    "Imperfect Voxelized Shadow Volumes"
        by Chris Wyman and Zeng Dai
Proceedings of High Performance Graphics 2013
        (Supplementary Materials)
---------------------------------------------


We include full-resolution images of all figures in our paper, and include some 
additional figures we did not have space for:

  1) FigureXX_comparison*_NaiveVsImperfectVSV.png:  A pair of figures comparing
the quality of imperfect voxelized shadow volumes with the naive approach described
and compared against in the paper (one using a fairly low-frequency model and a the
other a fairly high-frequency model)

  2) FigureXX_dynamicLightSponza.png is similar to figure 10 (it shows multiple frames
from a video sequence with a moving light).

  3) FigureXX_hairballPointVsAreaLight.png is similar to Figure 2.

Additionally, we include a video in DivX format showing our algorithm in action.  
Things we explicitly demonstrate are mobile geometry (and corresponding shadows),
mobile lights, mobile viewpoint, animated light sources (e.g., a video texture), 
comparison of shadows with and without interpolation, comparison of scattering
with and without shadows, quality comparison with varying resolution VSVs, light
moving behind an occluder, and quality and performance in complex scenes.

Please note: Figure 12 contains an explicit comparison to a ground truth ray casting
method for accumulating single-scattering light with (ray traced) shadows.