EFFECTIVENESS OF A PERIODIC AND APERIODIC DECOMPOSITION METHOD FOR ANALYSIS OF VOICE SOURCES

Decomposition of speech into periodic and aperiodic components is usef ul in analyzing and describing the characteristics of voice sources, S uch a decomposition is also useful in controlling the excitation sourc e for synthesis, This paper addresses the issue of decomposition of sp eech into periodic and aperiodic components in the context of speech p roduction, The effectiveness of a recently proposed algorithm for deco mposing speech into these components is examined for analysis of voice sources, Synthetic signals are generated using formant synthesis. Dif ferent sources of aperiodicity encountered in normal speech production are considered, using a set of parameters to control the synthetic si gnals, The sources of aperiocity studied are 1) additive pulsed or con tinuous random noise, and 2) modulation aperiodicities due to variatio n in the fundamental frequency, jitter, and shimmer, Three types of me asures are used to characterize these voices: ratio of energies in the periodic and aperiodic components, perceptual spectral distance, and spectrogram, The results demonstrate the effectiveness of the periodic -aperiodic decomposition algorithm for analyzing aperiodicities for a wide variety of voices, and point out the limitations of the algorithm .