| 摘要 |
PROBLEM TO BE SOLVED: To synthesize natural voice, even when the number of phonemes which is made into a database is fewer than in a conventional method, in a voice synthesis of a phoneme connection system. SOLUTION: On the basis of a musical piece data for indicating notes for constituting a musical piece and a lyrics data for indicating its lyrics, a singing synthesis score (voice synthesis direction) for indicating a plurality of phonemes used for synthesizing singing voice of the musical piece, its generation time and a pitch of its singing voice, are created. When the phoneme indicated by the singing synthesis score includes a silence vowel, a waveform of a predetermined silence pronunciation in the vowel of an object language of voice synthesis is converted into one including a spectrum envelope similar to the spectrum envelope of the phoneme in which the vowel is pronounced with voice, and the waveform after the conversion is used for synthesis of the singing voice as the waveform of the phoneme including the silence vowel. COPYRIGHT: (C)2010,JPO&INPIT
|
