A sound field in a vehicle can be further readily and optimally corrected without forcing a user to do troublesome work. As an embodiment of the present invention, a measurement sound emitted through a left speaker or a right speaker at a further position in a view from a driver's listening point that was assumed on the head touching surface of the headrest of the driver's seat or the passenger seat is picked up, with a microphone provided as buried on a front panel of a head unit to be attached to the almost center position between the driver's seat and the passenger seat in the vehicle. And the frequency characteristic of the above sound is corrected, based on a tendency that the frequency characteristic of an audio sound reaching from the left speaker or the right speaker to the microphone is almost approximate to the virtual frequency characteristic of an audio sound that will reach from the left speaker or the right speaker to the driver's listening point assumed on the head touching surface.