Speech synthesis by concatenation of Digital waveform fragments

A method to rule-synthesize speech by concatenation of digital waveform fragments at a subphonemic level is presented. No special hardware is needed to implement this software synthesizer other than a D/A converter and an ordinary audio system. Computer software for an on-line analysis-by-synthesis process was developed. Phonetic cues, such as characteristic waveform fragments, durations of each quasi-steady state and the transition motion of a certain number of phonemes were extracted and stored. Classifications of phonetic cues were possible and necessary to reduce the storage requirement and to obtain rules for synthesis. An interpolation scheme was developed to generate transient waveforms to eliminate the discontinuities at the concatenated junctions. Pitch variation was found to be the most influential factor for creating intonation in polysyllable utterances and was achieved by a pitch modification routine included in the synthesis program. Test procedures and results are reported in which a comparable vowel recognition rate for synthetic words is 93% vs. the 94% of digitized natural words in the first test. Further studies are needed to generalize the method to synthesize unrestricted text. The findings of the phonetic cues could be applied to speech recognition in future work.Applied Science, Faculty ofElectrical and Computer Engineering, Department ofGraduat