David Burgess' Home Page

(Photo edited from NV video sequence.)

Hello! Welcome to my home page! Up until 1 July 1994, I was a reseach scientist with the Multimedia Computing Group at Georgia Tech's Center for Graphics, Visualiation and Usability (GVU) in Atlanta. I left Georgia Tech because I was not able to find funding for the work I wanted to do. I now work at Interval Research in Palo Alto, California.

While at Georgia Tech, my primary interest was at the intersection of:

I'm also interested in virtual reality, parallel computing, audio/video compression, and just about anything that has to run in real-time and demands a lot of resources.

Specific Funded Research

While working on my Master's degree, I studied spatial audio, also called 3-D sound, as part of the Mercator project. Most of this work involved determining the minimum computational requirements for a true 3-D audio system. I have developed systems that can generate effective 3-D sound in real time for as little as 20MFLOPS, which is within the performance range of low-cost commercial DSP devices. I am presently seeking funding for the development of a 3-D audio chip for embedded consumer applications such as video games and teleconferencing tools. (Teleconferencing isn't a comsumer app yet, but just wait for the NII to bring the Net into people's homes.) I am also working on protocols and techniques for managing auditory virtual environments.

My last project at Tech was an audio server, which we called NetAudio 3 (NA3). When complete, NA3 will provide three basic services:

audio processing, including 3-D audio, stereo effects, and environmental modeling (reverb, echoes).
network support, allowing audio resources to be used across a network, much as X allows graphical applications to be used across a network.
parallelization of audio tasks across multiple processors or even multiple workstations.

Such a server will allow programmers an easy means to include stereo, surround, and 3-D sound processing in complex, network-based applications such as virtual reality, teleconferencing systems, and user interfaces for the visually impared.

Signal Processing Code

As a byproduct of my work with digital audio, I've written a good deal of code for common signal processing operations and the manipulation of spatial audio. Most of this code is designed to run very quickly on Sun SPARCstations and SGI Idigo machines. Click here for more information.

Contact Information:

burgess@interval.com