摘要
在文本语音转换的过程中,两个或多个声音文件的无缝拼接是经常需要处理的操作·从分析微软通用声音文件WAV的内部格式着手,以8bit位长,8MHz采样率声音文件为例,设计了对多个WAV文件的拼接算法,避免了实时拼接过程中产生的语音迟滞现象,为实现在文本语音转换中应用波形拼接技术奠定了良好的基础·本算法已实现并已在满文语音合成系统中成功应用·
The seamless connection operation between two sound wave files is absolutely necessarily operation during the text-to-speech process. Analyses the detail format of wave file which is Microsoft general sound file type, then takes the sound file which parameters value is 8bit and 8MHz as an example and designs the connection algorithm between two sound wave files. This method avoids the delay phenomenon of real-time sound connection and lays a foundation for realization of text-to-speech. This algorithm has been realized and applied in Manchu text-to-speech system.
