摘要 |
PROBLEM TO BE SOLVED: To enable changing freely features of a text synthetic voice by regulation synthesis. SOLUTION: This is a method and device synthesizing a voice by regulation synthesis using a real voice as auxiliary information, rhythm information for a phoneme system of each word of a word system obtained by analyzing an input text is set referring to a word dictionary, and a phoneme waveform system is obtained from a phoneme system of each word referring to a voice waveform dictionary 16. On time other hand, rhythm information is extracted form an input real voice, either of set rhythm information and extracted rhythm information is selected, a voice waveform system is controlled by any of selected rhythm information, and synthetic voice is obtained. |