Abstract:

A model for fricative sound generation based on aeroacoustic theory is implemented in an articulatory synthesizer. The aeroacoustic model is based upon Howe's reformulation of the acoustic analogy, which allows the source to be specified in terms of the jet formed at a constriction, the shape of the vocal tract walls, and the acoustic loading due to the lip termination. As such, the model uses existing acoustic models for the vocal tract combined with an approximate model for the behavior of the jet based on turbulence theory and experiment. This allows the fricative source strength, impedance, and temporal characteristics to be specified using the lung pressure, the constriction diameter, and the shape of the vocal tract downstream of the constriction. Evaluation of the source model is presented. Computed estimates of fricative source characteristics are presented and compared to source models used by other investigators. A comparison of synthetic fricative sounds generated using these models is also presented. [Research supported by NSF/ARPA IRI-9314946 and ARPA DAST 63-93-C-0064.]