Generalization for damping in the acoustic tube model of speech production
A. H. Gray, John D. Markel, H. Wakita · The Journal of the Acoustical Society of America · 1976
The widely used method of vocal tract modeling proposed by Wakita [IEEE AU-21 pp. 417–427] is a lossless model which places all possible damping at the glottis. In addition, the lip radiation impedance for this model is, by definition, zero. Elementary recursion relationships allow unique transformations between the inverse filter A(z) and the acoustic tube area functions. In this paper, the acoustic tube model is generalized to allow for arbitrary real attenuation within each acoustic tube section and to allow for the lip radiation impedance form described by Flanagan [Speech Analysis (Springer-Verlad, Berlin, 1972), 2nd ed. p. 36]. The approach suggested in Markel and Gray [Linear Prediction of Speech (Springer—Verlag, Berlin, 1976), p. 88] is pursued here and it is shown that a recursive expression for a new inverse filter can be obtained in terms of the area functions, the glottal termination, the internal losses, and a realistic lip radiation impedance. Theoretical and experimental results will be discussed.