摘要 |
PURPOSE:To select music when the words of a song is known even at the time of selecting music whose name is not known by using the words of the song inputted from a microphone as music selection information. CONSTITUTION:Information on the words of the song corresponding to a musical performance registered in a KARAOKE (recorded instrumental accompaniment) device is previously stored in a registration information storage part 6. A sound recognition part 7 sets information on the words of the song which is previously stored in the registration information storage part 6 to be a standard sound pattern, refers the standard sound pattern to music selection information stored in a music selection information storage part 5 so as to recognize the words of the song inputted as music selection information. When reference results agree, it outputs recognition information. Then, a music selection part 9 selects music based on recognition information and the selected performance among performances registered in a performance registration part 10 is outputted to an output processing part 11. When the words of the song corresponding to music is known even in the selection of music whose name is not known, the music can be selected, and the music selecting operation can easily be executed at a place apart from a music selector. |