To obtain sound characteristics by an optimum virtual loudspeaker, according to the position of a user.

A user's position detection part 110 detects distance between a human body sensor 30 and a person (user), according to the level of the detection signal from the human body sensor 30 and outputs the distance data to an angle calculation part 120. The angle calculation part 120 calculates the angle of spread between two loudspeakers 50A and 50B with respect to the position of the person, according to the distance data from the user's position detection part 110 and the interval between the two loudspeakers 50A and 50B which is stored previously. An arithmetic part 130 applies arithmetic processing to an input audio signal according to the angle calculated by the angle calculation part 120 to generate an audio signal representing a virtual loudspeaker having a full sense of spread and output it to the two loudspeakers 50A and 50B. Consequently, when the position of the user to a display device is changed, an optimum virtual loudspeaker can be set corresponding to the changed position.