Abstract

In intervocalic voiced stop consonants, pitch decreases then increases; in voiceless stops it decreases, skips up during silence, then falls back again. Here pitch depends upon glottal pressure drop, hence upon supraglottal constriction, glottal opening, and intervening cavity volume. An experiment is reported showing that these pitch contours can cue the voicing distinction in stops. Down-up-down contours can occur with voiced fricatives; as supraglottal pressure mounts (i) closure formants fade, (ii) voice-modulated friction onsets, and (iii) pure friction takes over. Oscillograph experiments show that occurence of down-up-down contours depends upon whether Phase (iii) is reached. This varies with place of articulation, and freely. Thus, pitch changes could not rival closure formants before or during friction as chief voicing cue in fricatives. As friction amplitude varies directly, and pitch and voice amplitude vary inversely, with supraglottal pressure, all segmental excitation data can be derived from time course of supraglottal pressure, and hence from glottal and supraglottal constriction parameters. In articulatory synthesis by rule, therefore, only these parameters need be specified for each phoneme; an algorithm to generate excitation data is outlined.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call