Optimal spatial sampling for spherical loudspeaker arrays

The design of spherical loudspeaker arrays involves the discrete sampling of the sphere. In practical applications the desired number of sample points is very low (possibly 16-32 transducers) and there are typically one or more areas in the room that cannot accommodate loudspeakers. This paper presents an optimization method for choosing loudspeaker locations and driving functions in such a constrained environment. An additional weighting is added to allow for selective optimization of both low spatially-varying components and azimuthal components.