Improving Depth Perception in 3D Interfaces with Sound

This thesis investigates the use of sound to improve user depth perception in 3D computer interfaces. Two experiments were performed to determine a) the effect that sound has on depth perception and b) how this effect varies over an extended period of use. When compared to a “no sound” environment with limited visual depth cues, sound feedback helped to reduce task errors while at the same time increased user appreciation. Task completion under the use of sound did, however, take longer than the no sound environment. “Sound only” trials, which have no visual feedback, were also tested and in some cases reached the same performance level of the no sound case suggesting a possible use in applications not requiring a visual display. Three types of audible sound environments were studied each using different sound feedback: a simple tone, a musical piece, and a more involved musical piece using a fixed orchestra arrangement. Overall, the tonal environment proved to have the lowest error levels, yet was more annoying than the other two. An extended period study was performed, that required subjects to repeat the experiment on a Monday, Wednesday, and Friday, to measure the effect that experience had on performance. The results showed little improvement in task accuracy but a significant improvement in task time. Extra trials were also performed to determine if sounds at random locations could be learned and later located by memory. Although performance for these memory trials did decrease compared to the regular trials, there was still a notable improvement over the no sound condition. In conclusion, sound was found to be an effective mechanism to improve depth perception.