Orthogonal loudspeaker matching for stable and accurate multizone sound field reproduction

Abstract

Multizone sound field reproduction employs loudspeakers to create personalized sound zones within a shared space. The selection of optimal loudspeaker placements is crucial for enhancing system performance, which has led to the development of various algorithms for loudspeaker placement optimization. Typically, these methods focus on reproducing a single desired sound field, often neglecting the exchange between bright and dark zones. This paper evaluates these methods in reproducing multiple desired sound fields. Additionally, existing approaches generally emphasize either reproduction accuracy or system stability, but rarely both. To address this limitation, this study introduces an iterative method that extracts the orthogonalization component between the candidate loudspeaker and the selected loudspeakers, referred to as the “orthogonal loudspeaker”. The sound fields reproduced by this component serve as the criterion for determining optimal loudspeaker placement during the iteration process. Experiments conducted in both narrowband and broadband settings demonstrate that the proposed method can accurately reconstruct the desired sound fields while maintaining a relatively low condition number for the selected loudspeaker configuration. Moreover, the proposed method exhibits high stability below 600 Hz and ensures accurate reproduction at higher frequencies, even in the presence of system perturbations.