Real-time audio rendering system for virtual reality

Abstract

Systems of virtual acoustics are more and more coming into the focus of research and application. Creating a virtual sound scene with spatially distributed sources requires a technique for adding spatial cues to audio signals and an appropriate reproduction. A real-time audio rendering system is introduced that combines room acoustics, dynamic crosstalk cancellation, and multitrack binaural synthesis for virtual acoustical imaging. The room acoustics component takes into account the full up-to-date algorithmic approach of specular reflections and scattering, but with real-time processing. The real-time performance of the software was reached by introduction of a flexible framework for the interactive auralisation of virtual environments. The concept of scene graphs for the efficient and flexible linkage of autonomous-operating subscenes by means of so-called portals has been incorporated into the existing framework, combined with an underlying BSP-tree structure for processing geometry issues very fast. Using this framework enables a significant reduction of computation time for both applied algorithms (deterministic image sources and a stochastic ray tracer). This enables the simulation of indoor or outdoor spatial distributed sources and, in addition to that, near-to-head sources for a freely moving listener in room-mounted virtual environments without the use of headphones.