Simulations of voice onset–offset hysteresis in /aha/ utterances

Abstract

The subcritical Hopf bifurcation model for voice onset–offset hysteresis [J. C. Lucero, J. Acoust. Soc. Am. 105, 423–431 (1999)] is investigated considering the glottal area as control parameter. A two-mass model of the vocal folds coupled to a two-tube approximation for the vocal tract is used for producing oral airflow simulations. The glottal rest area is varied in an abduction–adduction pattern to reproduce smoothed oral airflow records measured during /aha/ production [L. L. Koenig and R. S. McGowan, J. Acoust. Soc. Am. 100, 2689(A) (1996)]. The resultant simulations show good agreement with the unsmoothed measured records, with a clear hysteresis effect: vocal-fold oscillation stops during glottal abduction at a higher value of the glottal area than the value at which it starts during the adduction. The dependence of the onset–offset hysteresis and voice parameters on vocal-fold stiffness (Q factor), subglottal pressure, and male–female laryngeal configurations is analyzed and illustrated with bifurcation diagrams. [Work supported by CNPq, Brazil, and NIH.]