Modification of Velvet Noise for Speech Waveform Generation by Using Vocoder-Based Speech Synthesizer

Masanori MORISE  

IEICE TRANSACTIONS on Information and Systems   Vol.E102-D   No.3   pp.663-665
Publication Date: 2019/03/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2018EDL8179
Type of Manuscript: LETTER
Category: Speech and Hearing
noise generation,  white noise,  velvet noise,  zero-frequency component,  speech analysis/synthesis,  

Full Text: FreePDF(294.1KB)

This paper introduces a new noise generation algorithm for vocoder-based speech waveform generation. White noise is generally used for generating an aperiodic component. Since short-term white noise includes a zero-frequency component (ZFC) and inaudible components below 20 Hz, they are reduced in advance when synthesizing. We propose a new noise generation algorithm based on that for velvet noise to overcome the problem. The objective evaluation demonstrated that the proposed algorithm can reduce the unwanted components.