Acoustic analysis on VOT of plosives, affricates in Uyghur

Based on the research needs of the application of phonetics such as speech synthesis and speech recognition, this work mainly focus on VOT(Voice-Onset Time)in Uyghur based on the text analysis module, which is also the first systematic empirical work. The purpose is to enhance the naturalness of Uyghur speech synthesis and speech recognition system. The experimental analysis is based on an acoustic database recorded by two speakers, one male and one female. This database contains VOT values of 556 disyllabic words including voiceless plosives and voiceless affricates. In this paper, the concepts of aspirated and unaspirated consonant are introduced for the first time from the view of experimental phonetics. It is concluded that: voiced and voiceless plosives, affiricates can be effectively distinguished by using VOT features; word-initial voiceless plosives and affiricates sometimes are aspirated and sometimes unaspirated, but they are unaspirated only in word-medial position. There are three levels for the word-final position:weakly aspirated, moderately aspirated and strongly aspirated.