发明名称 VOICE SYNTHESIZER
摘要 PURPOSE: To generate a waveform parameter with less memory capacity and capable of also warranting continuity of connection on a spectrum by using a vocal cords voice source wave and synthesizing a sentence voice. CONSTITUTION: A spectrum slope of a syllable stored in a syllable storage pattern 407 is connected by a syllable pattern connection part 409 to be converted to a waveform parameter by a waveform parameter conversion part 410. The vocal cords voice source wave is generated using the waveform parameter, and a synthesis voice is generated through a vocal tract filter. In such a manner, since the spectrum slope of the vocal cords voice source wave is the parameter directly corresponding to the spectrum, the spectrum continuity is warranted even after the spectrum slopes stored in individual storage pattern are connected with each other. Since the spectrum slope is a single parameter, a required memory is reduced to 1/4 compared with the case when four kinds of waveform parameters such as SK, OQ, C, D are stored.
申请公布号 JPH08202381(A) 申请公布日期 1996.08.09
申请号 JP19950012903 申请日期 1995.01.30
申请人 SUMITOMO ELECTRIC IND LTD 发明人 YAMAGUCHI MIKIO
分类号 G10L13/08;G10L13/04 主分类号 G10L13/08
代理机构 代理人
主权项
地址