An audio architecture integrating sound and live voice for virtual environments

Abstract

The purpose behind this thesis was to design and implement audio system architecture, both in hardware and in software, for use in virtual environments. The hardware and software design requirements were to provide the ability to add sounds, environmental effects such as reverberation and occlusion, and live streaming voice to any virtual environment employing this architecture. Several free or open-source sound APIs were evaluated, and DirectSound3D was selected as the core component of the audio architecture. Creative Labs Environmental Audio Extensions (EAX) was integrated into the architecture to provide environmental effects such as reverberation, occlusion, obstruction, and exclusion. Voice over IP (VoIP) technology was evaluated to provide live, streaming voice to any virtual environment. DirectVoice was selected as the voice component of the architecture due to its integration with DirectSound3D . However, extremely high latency considerations with DirectVoice, and any other VoIP application or software, required further research into alternative live voice architectures for inclusion in virtual environments. Ausim3D's GoldServe Audio Localizing Audio Server System was evaluated and integrated into the hardware component of the audio architecture to provide an extremely low-latency, live, streaming voice capability.