Skip to Main content Skip to Navigation
Reports

Perceptual Audio Rendering of Complex Virtual Environments

Abstract : Including spatialized audio in virtual environments significantly enhances the sense of realism and immersion. With the constant increase in complexity of virtual scenes, rendering spatialized audio is currently often impossible, due to the prohibitive signal processing costs of rendering a large number of sound sources. We present a novel solution to this problem permittin- g audiovisual rendering of environments containing hundreds of moving sound sources with complex frequency-dependent effects and reverberation processing on an consumer grade system. Our approach introduces three algorithms, based on perceptual culling, a clustering strategy allowing for rendering a large number of sound sources with only a limited number of available hardware 3D audio channels, and the use of graphics hardware to perform required pre-mixing operations on the audio signals. With little overhead to the graphics resources, our method leverages the use of existing audio rendering hardware while introducing additional controls and processing capabilities. We have implemented our approach, and we present results for a complex, highly dynamic virtual environment containing hundreds of sound sources, animated characters and moving objects. The results show that our approach permits high-quality interactive audio-visual rendering of such scenes.
Document type :
Reports
Complete list of metadata

Cited literature [33 references]  Display  Hide  Download

https://hal.inria.fr/inria-00071853
Contributor : Rapport de Recherche Inria <>
Submitted on : Tuesday, May 23, 2006 - 6:58:02 PM
Last modification on : Saturday, January 27, 2018 - 1:31:32 AM
Long-term archiving on: : Tuesday, February 22, 2011 - 12:01:32 PM

Identifiers

  • HAL Id : inria-00071853, version 1

Collections

Citation

Nicolas Tsingos, Emmanuel Gallo, George Drettakis. Perceptual Audio Rendering of Complex Virtual Environments. RR-4734, INRIA. 2003. ⟨inria-00071853⟩

Share

Metrics

Record views

250

Files downloads

407