Abstract: Loudspeakers of plural channels are used as microphones to convert a test sound generated at a listening point into corresponding electric signals respectively. Moments of arrival of the test sound at the loudspeakers are detected in response to the electric signals respectively. Desired delay times for audio signals of the plural channels are calculated from the detected moments of arrival of the test sound at the loudspeakers. Volume levels of the electric signals are detected. Desired volume levels for the audio signals are calculated from the detected volume levels. The audio signals are delayed by the desired delay times respectively and volume levels of the audio signals are controlled at the desired volume levels respectively to get processing-resultant audio signals of the plural channels. The processing-resultant audio signals are fed to the loudspeakers respectively.