Perceptually Driven Interactive Sound Propagation for Virtual Environments
Public Deposited- Last Modified
- July 24, 2019
- Creator
-
Rungta, Atul
- Affiliation: College of Arts and Sciences, Department of Computer Science
- Abstract
- Sound simulation and rendering can significantly augment a user‘s sense of presence in virtual environments. Many techniques for sound propagation have been proposed that predict the behavior of sound as it interacts with the environment and is received by the user. At a broad level, the propagation algorithms can be classified into reverberation filters, geometric methods, and wave-based methods. In practice, heuristic methods based on reverberation filters are simple to implement and have a low computational overhead, while wave-based algorithms are limited to static scenes and involve extensive precomputation. However, relatively little work has been done on the psychoacoustic characterization of different propagation algorithms, and evaluating the relationship between scientific accuracy and perceptual benefits.In this dissertation, we present perceptual evaluations of sound propagation methods and their ability to model complex acoustic effects for virtual environments. Our results indicate that scientifically accurate methods for reverberation and diffraction do result in increased perceptual differentiation. Based on these evaluations, we present two novel hybrid sound propagation methods that combine the accuracy of wave-based methods with the speed of geometric methods for interactive sound propagation in dynamic scenes.Our first algorithm couples modal sound synthesis with geometric sound propagation using wave-based sound radiation to perform mode-aware sound propagation. We introduce diffraction kernels of rigid objects,which encapsulate the sound diffraction behaviors of individual objects in the free space and are then used to simulate plausible diffraction effects using an interactive path tracing algorithm. Finally, we present a novel perceptual driven metric that can be used to accelerate the computation of late reverberation to enable plausible simulation of reverberation with a low runtime overhead. We highlight the benefits of our novel propagation algorithms in different scenarios.
- Date of publication
- 2019
- Keyword
- DOI
- Resource type
- Advisor
- Lin, Ming
- Klatzky, Roberta
- Bishop, Gary
- Niethammer, Marc
- Manocha, Dinesh
- Degree
- Doctor of Philosophy
- Degree granting institution
- University of North Carolina at Chapel Hill Graduate School
- Graduation year
- 2019
Relations
- Parents:
This work has no parents.
Items
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
Rungta_unc_0153D_18617.pdf | 2019-07-24 | Public | Download |