Speech analysis using discrete spectral modeling

A method for the parametric modeling of spectral envelopes when only a discrete set of spectral values is given, e.g. voiced speech, is introduced. This method, which is called discrete pole-zero (DPZ) modeling, uses a discrete version of the Itakura-Saito distortion measure as its error criterion. An efficient iterative algorithm for the estimation of the envelope parameters which is guaranteed to converge to a locally optimal model is introduced. How DPZ modeling can be applied to the estimation of real and synthetic speech spectral envelopes is shown.<<ETX>>