Speech analysis–synthesis using sinusoidal representations: A review

Abstract

Various speech analysis–synthesis representations have been suggested in the literature, and the more well-known ones are explored in this chapter, specifically, sinusoidal representation, harmonic/noise representation, and adaptive sinusoidal representations. Hence, the main objective of this chapter is to give a tutorial review of speech analysis–synthesis representations, by highlighting major improvements over these representations. It would be a desirable representation of speech that is relatively simple, flexible, high quality, and robust in re-synthesis. Emphasis will be given in adaptive sinusoidal representations, since they seem to be more promising and robust representations of speech.