Abstract
B. S. Atal and J. R. Remde (Proc. ICASSP 82) described a method for determining the multipulse excitation for LPC speech synthesis. We find that, even for reasonably periodic speech waveforms, the secondary pulses in the multipulse excitation do not show similar periodic behavior. What is the influence of this lack of periodicity in the excitation on the quality of synthetic speech? We have investigated this question by introducing controlled periodicity into the multipulse excitation. The excitation was first segmented into individual pitch periods. For every sequence of n pitch periods, the multipulse pattern from the pitch period in the middle of each sequence was then used to repeat the excitation pattern in the other n − 1 surrounding pitch periods. In this process, the overall amplitude envelope of the utterance was left unchanged. We find that as many as 15 periods may be repeated without introducing significant audible distortion in the synthetic speech signal. However, repetition of a single multipulse pattern across all voiced segments of an utterance produced noticeable coloration in synthetic speech. We will discuss further implications of these results in this paper.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.