Computational Acoustics At Meyer Sound

Perrin Meyer

Senior DSP Engineer, Meyer Sound Laboratories Inc., Berkeley, CA.

Meyer Sound, a local company that manufactures professional audio equipment, collaborates and sponsors innovative research in computational acoustics. I will describe two projects. The first is in collaboration with the UC Berkeley interdisciplinary Center for New Music and Audio Technologies (CNMAT). We are investigating parallel algorithms for the real-time signal processing of hundreds of channels of live streaming audio. These signals create real-time three dimensional sound scapes utilizing special purpose loudspeakers. One such speaker is a 120 transducer compact spherical array. Another loudspeaker contains 96 transducers tight packed in a line, intended for horizontal wavefield synthesis. On the measurement side we have created a 144 element spherical microphone. The second project is in collaboration with the Courant Institute of Mathematical Sciences (CIMS). We are sponsoring the development of a Fast Multipole based accelerated Boundary Element code for the Helmholtz equation in complex three dimensional geometries. The code is designed to scale to meshes of millions of triangles, allowing the solution of acoustic radiation and scattering problems hundreds of wavelengths in diameter. I will also show some recent results of modeling Meyer Sound loudspeakers, and compare that to empirically measured data from our anechoic chamber. The eventual goal of the project is to release an open-source version to the research community, with hopes that the code can be parallelized on new multicore/GPU architectures.