DRIVING THE VOTRAX SPEECH SYNTHESIZER FROM A WIDE PHONETIC TRANSCRIPTIONWITH HIGH-LEVEL PROSODIC MARKERS

Abstract

Previously developed procedures for controlling the pitch and rhythm of synthetic speech by markers in the input phonetic text have been adapted for the Votrax ML-I speech synthesizer. The aim is not necessarily to advocate these procedures, but rather to show that existing knowledge about prosodic features of speech can be adapted to the Votrax device. Also described are rules for translating International Phonetic Alphabet transcriptions into Votrax sound segments, which incorporate many of the hints for allophone adjustment that are given in the Votrax manual.