发明名称 METHOD FOR ENCODING SPEECH AND DEVICE THEREFOR
摘要 PROBLEM TO BE SOLVED: To reduce a wave distortion caused by selecting a fixed code book when an input widely changes in aural frequency components in encoding used by changing over an adaptive code book and a fixed code book to/from each other for a selection such as PSI-CELP(pitch synchronous innovation-code excited linear prediction). SOLUTION: An output from an adaptive code book 21 or an output from a fixed code book 22 is selected by a selection switch 26 and is added to outputs from noise code books 23, 24, for being sent to a linear prediction synthesis filter 16. A switch control circuit 19 controlling the selection switch 26 discriminates a voiced sound according to a prediction gain, the ratio of a linear prediction residual energy to an initial signal energy from a linear prediction analyzing circuit 14, when this prediction gain is smaller than a prescribed threshold, and controls the selection switch 26 to forcibly select the output from the adaptive code book 21.
申请公布号 JPH1020891(A) 申请公布日期 1998.01.23
申请号 JP19960179178 申请日期 1996.07.09
申请人 SONY CORP 发明人 MAEDA YUJI
分类号 G10L19/08;G10L19/00;G10L19/04;H03M7/30;H03M13/00;(IPC1-7):G10L9/14;G10L9/18 主分类号 G10L19/08
代理机构 代理人
主权项
地址