发明名称 VOICE SYNTHESIZER
摘要 PROBLEM TO BE SOLVED: To obtain a high-quality synthetic voice by providing an element piece producing method which has both merits of an one pitch-equivalent waveform element piece producing method and a symmetrized element piece producing method while limiting a processing quantity and a holding information quantity to slight increase as compared with the conventional symmetrized element piece producing method and in which connections are satisfactory and the degrading of higher harmonic components is small. SOLUTION: In the element piece producing circuit 105, however element data are produced and they are written in an element piece dictionary 106 in an element piece writing circuit 112, the ratio of powers of high harmonic components is very large with respect to the power of a voice signal in the vicinity of voiced consonant parts being in phonemes including voiced consonants and the voice signal is not suited for a Fourier analysis. Positions being in phonemes are applied to the voice signal previously and when the voice signal is voiced consonant parts including an explosive or an explosively frictional sound (/b/, /d/, /g/, /j/, /z/), one pitch-equivalence waveform element piece data are produced to be written in the dictionary 106 and in parts other than them, symmetrization element piece data are produced to be written in the dictionary 106.
申请公布号 JPH1173198(A) 申请公布日期 1999.03.16
申请号 JP19970234578 申请日期 1997.08.29
申请人 OKI ELECTRIC IND CO LTD 发明人 IWAKI TAKESHI
分类号 G10L13/06;(IPC1-7):G10L5/04 主分类号 G10L13/06
代理机构 代理人
主权项
地址