Abstract: The data put to the digital filter operation and the data not put to the digital filter operation are synthesized together at a predetermined distribution ratio to form a stereophonic sound image. A digital filter operation is executed for forming a sound image in the vertical direction that reaches one ear being diffracted by the head, and a digital filter operation is executed for forming a sound image in the vertical direction in the other ear, thereby to form a stereophonic sound image in the vertical direction. Of the sound generated from one of the different sound systems, the sound reaching one ear is canceled and the only sound reaching the other ear is validated by generating a counterbalance sound component. Excess sound component is erased, the excess sound component being generated from the other sound system and being not desired for forming the stereophonic sound image.