Iterative inverse filter relaxation algorithm for adaptation to acoustic fluctuation in sound reproduction system

摘要

In this paper, we propose an adaptive relaxation algorithm for inverse filter design of the multichannel sound reproduction system. In the conventional multichannel sound reproduction system, a design method of inverse filter using Moore-Penrose generalized inverse matrix has been proposed. However, the inverse filter of the conventional method often amplifies background noise and line noise when the room transfer function changes. In order to resolve the problem, the adaptive relaxation algorithm based on Truncated Singular Value Decomposition (TSVD) method is introduced, in which the only one microphone to observe the noise condition is placed apart from listener's ears. The experimental result of the computer simulation reveals that the proposal method can prevent the noise amplifications. Also, the subjective evaluation in the real acoustic environment indicates the trade-off relation between the sound quality and the localization.