Extraction of the Spectrum Envelope for Voiced Speech Segments
Paul G. Mermelstein · The Journal of the Acoustical Society of America · 1967
A previously suggested method for smoothing the narrow-band spectra of voiced speech segments by “low-pass filtering” the logarithm of the power spectrum [M. R. Schroeder and A. M. Noll, “Recent Studies in Speech Research at Bell Telephone Laboratories. I,” Paper A21 in Proceedings of the Fifth International Congress on Acoustics, Liége, Belgium, 1965, D. E. Commins, Ed. (Georges Thone, Liége, 1966)] has been found inadequate for high-pitched female voices. Low-pass filtering, with a cutoff “quefrency” (independent variable in the Fourier-transform domain of power spectra) below the pitch-period value, may result in elimination of peaks due to the formant structure. An approximation to the spectrum fine structure (due to the quasiperiodic excitation of the vocal tract) is generated by requiring its Fourier cosine transform to match that of the log power spectrum of the speech signal in the immediate neighborhood of the pitch peak, but to have no other significant quefrency components. By subtraction of this approximation to the fine structure from the log power spectrum of the speech signal, an improved spectrum envelope is obtained. A formant-tracking algorithm, operating on spectral-envelope data so derived, yielded correct results where otherwise serious errors would have been committed. Several examples for male and female speech are given.