page 1  (1 pages)

(a) ff = :10 (b) ff = :15 (c) ff = :20 (d) ff = :30

Figure 10: DVR images with corresponding visualizations for the number of samples to reach the given opacity threshold. Regular ray sampling is used to render this Hipip data set. (a) 0.10, (b) 0.15, (c) 0.20, and (d) 0.30. Black regions indicate that rays entered and exited the volume without reaching the opacity threshold.

(a) images (b) (c) (d) (e)

Figure 11: Volume rendered images and visualizations of their distance metrics and difference metrics. The top row is rendered with regular ray sampling, while the bottom row is with cell face intersection. Column (a) are the rendered images and columns (b) - (d) are visualizations of metrics. The metrics are measured until each ray reaches the opacity threshold of 0.11. Column (b) shows eye distance metric, (c) shows volume distance. In column (d), the volume distances from (c) are mapped to height and rendered as a surface. To generate the top image of column (e), we take the difference between the number of ray samples used by the two DVR algorithms and map that to color. For the lower image of column (e), we take the difference between the volume distance used by the two DVR algorithms and map that to color.