Sparse Framework for Personal Sound Reproduction with Differential Phase Constraint

Abstract

Sparse multi-zone sound reproduction provides multiple users a personalized sound experience. The primary objective of these methods is to minimize the reproduction error in the zones of interest along with the simultaneous reduction in the number of active of loudspeakers. The accuracy of these methods depends on how well the reproduced sound field resembles the desired sound field. In this work, a sparse framework for personal sound reproduction using a differential phase constraint is developed. The additional constraint enhances the spatial quality of the reproduced sound field in the bright zones by preserving the phase component from the direction of interest. The formulated optimization problems is tested in simulated environment with different scenarios. The performance of the proposed framework is evaluated objectively on the basis of obtained acoustic contrast, reproduction error and the directivity of the reproduced sound fields.