摘要 |
PROBLEM TO BE SOLVED: To provide a voice synthesizer capable of generating a synthesized voice having high natural voice feeling, while holding stable feeling by a plurality-of-elementary-piece selection integration system. SOLUTION: This voice synthesizer 4 is provided with a periodical component integration part 46 and a nonperiodical component integration part 47, the periodical component integration part 46 integrates periodical components of a plurality of voice elementary pieces per a segment selected by an elementary piece selecting part 44, and the nonperiodical component integration part 47 integrates nonperiodical components thereof. The voice synthesizer 4 is provided further with an addition part 48, adds the periodical components and the nonperiodical components of the integrated voice elementary pieces, by the addition part 48, and conducts edition and connection to generate a synthesized waveform. COPYRIGHT: (C)2009,JPO&INPIT
|