Voiceless Stop Consonant Modelling and Synthesis Framework Based on MISO Dynamic System

Abstract

AbstractA voiceless stop consonant phoneme modelling and synthesis framework based on a phoneme modelling in low-frequency range and high-frequency range separately is proposed. The phoneme signal is decomposed into the sums of simpler basic components and described as the output of a linear multiple-input and single-output (MISO) system. The impulse response of each channel is a third order quasi-polynomial. Using this framework, the limit between the frequency ranges is determined. A new limit point searching three-step algorithm is given in this paper. Within this framework, the input of the low-frequency component is equal to one, and the impulse response generates the whole component. The high-frequency component appears when the system is excited by semi-periodic impulses. The filter impulse response of this component model is single period and decays after three periods. Application of the proposed modelling framework for the voiceless stop consonant phoneme has shown that the quality of the model is sufficiently good.