Three-dimensional sound reproducing apparatus and a three-dimensional sound reproduction method

A three-dimensional sound reproduction method, comprising the steps of:

• providing two-channel first audio signals defining sound images which are to be respectively localized on left and right sides of a listener by one of reproducing from a medium and receiving from an outside, the two-channel first audio signals to which a center-channel audio signal defining a sound image to be localized at a center is commonly added, and the two-channel first audio signals to which nonlocalization audio signals separated in phase by 180 deg. from each other are respectively added;
• conducting filtering processes respectively corresponding to transfer functions of paths from a virtual point in a three-dimensional acoustic space to left and right ears of the listener on the two-channel first audio signals, to generate two-channel second audio signals defining a sound image to be localized at the virtual point;
• and
• conducting a crosstalk canceling process on the two-channel second audio signals to generate two-channel third audio signals, so that, when the two-channel third audio signals are respectively is generated by two loudspeakers positioned in front of the listener, the two-channel third audio signals reach the left and right ears of the listener without producing crosstalk.

A three-dimensional sound reproduction method, comprising the steps of:

• providing two-channel first audio signals defining sound images which are to be respectively localized on left and right sides of a listener, a center-channel audio signal defining a sound image to be localized at a center, and nonlocalization audio signals, by one of reproducing from a medium and receiving from an outside;
• conducting filtering processes respectively corresponding to transfer functions of paths from a virtual point in a three-dimensional acoustic space to left and right ears of the listener on the two-channel first audio signals, to generate two-channel second audio signals defining a sound image to be localized at the virtual point;
• conducting a phase shifting process on the nonlocalization audio signal to generate two-channel nonlocalization audio signals separated in phase by 180 deg. from each other;
• adding the center-channel audio signal commonly, and the two-channel nonlocalization audio signals respectively, to the two-channel second audio signals to generate two-channel third audio signals;
• and
• conducting a crosstalk canceling process on the two-channel third audio signals to generate two-channel fourth audio signals, so that, when the two-channel fourth audio signals are respectively generated by two loudspeakers positioned in front of the listener, the two-channel fourth audio signals reach the left and right ears of the listener without producing crosstalk.