The final challenge is to ensure that the physical and virtual ears align in the matched zone. If the user turns her head, then the sound should be adjusted accordingly. If the sound emanates from a fixed source, then it should be perceived as fixed while turning the head. This is another example of the perception of stationarity. Accordingly, tracking of the ear pose (position and orientation) is needed to determine the appropriate ``viewpoint''. This is equivalent to head tracking with simple position and orientation offsets for the right and left ears. As for vision, there are two choices. The head orientation alone may be tracked, with the full pose of each ear determined by a head model (recall Figure 9.8). Alternatively, the full head pose may be tracked, directly providing the pose of each ear through offset transforms. To optimize performance, user-specific parameters can provide a perfect match: The distance along the axis from the eyes to the ears and the distance between ears. The latter is analogous to the IPD, the distance between pupils for the case of vision.
Steven M LaValle 2016-12-31