Scattering Delay Network: an Interactive Reverberator for Computer Games

Enzo De Sena, Hüseyin Hacıhabiboğlu and Zoran Cvetkovic
Department of Informatics, King's College London Strand,
WC2R 2LS, London, United Kingdom
{e.desena|huseyin.hacihabiboglu|zoran.cvetkovic}@kcl.ac.uk

Many 3D computer games incorporate audio renderers simulating room acoustics to provide the user with a high level of immersiveness and realism. Full-scale interactive room auralisation systems are impractical for use in computer games due to their high computational cost. As a low-cost alternative, artificial reverberators can be used. This paper is concerned with the design of a scalable interactive reverberator inspired by digital waveguide mesh (DWM) models and feedback delay networks (FDN). This reverberator is by construction tightly linked to the acoustics of the enclosure that it simulates. Simulation of unequal and frequency- dependent wall absorption, as well as directional sources and microphones can also be incorporated. It is shown that the response of the proposed reverberator accurately renders the early reflections and room modes, as well as providing RT60 values consistent with Sabine and Eyring equations.

Audio samples

The simulated space has the same dimensions as our audio booth: 6.52x4.56x2.10 m. The source and the microphone array are positioned as shown in the figure below, at a height of 1.05 m. The microphone array is a stereophonic subset of our patented array for base-angle 72deg and radius 15.5cm. It is assumed that the stereophonic loudspeaker setup you are using to play back the samples has roughly 72deg base-angle. The source is moving back and forth on a line as deicted in red in figure. The original audio sample is Archimedes's anechoic percussion.

As a first sample, the walls have realistic frequency-dependent absorption. More specifically, the floor is cotton carpet, the ceiling is fissured tiles while the walls are walls, hard surfaces average, as reported in M. Vorlander's book. Here you can find the audio sample:
https://desena.org/sdn/realistic_walls.wav

As a second example, the walls have frequency-independant absorption, which changes over time. For the first 5 seconds the room is perfectly anechoic. Between 5 and 10 seconds, the walls have a reflectiveness of 0.75, which then increases to 0.85 between 10 and 15 seconds. Finally, at 15 seconds the reflectiveness is further increased to 0.9. Here you can find the audio sample:
https://desena.org/sdn/changing_reflectiveness.wav