ABSTRACT

Timescale modification is the general term applied to various methods for altering the temporal structure of speech without altering other aspects of the signal such as its pitch or the voice quality of the speaker. Timescale modification can be applied globally to change the perceived speaking rate of an utterance by uniformly increasing or decreasing the durations of all the syllables in the utterance. Several methods have been developed for timescale modification with low distortion. There are a variety of reasons one might wish to alter the timescale of speech. Turning from socially valuable contributions of timescale modification to purely economic contributions, another application of time compression is to adjust the length of movies, commercials, and other prerecorded material to fit within a restricted time frame. The procedure for speech timescale modification can be formulated in terms of the process of copying an input waveform buffer or file to an output waveform buffer or file.