A new cosine series antialiasing function and its application to aliasing-free glottal source models for speech and singing synthesis

摘要

We formulated and implemented a procedure to generate aliasing-freeexcitation source signals. It uses a new antialiasing filter in the continuoustime domain followed by an IIR digital filter for response equalization. Weintroduced a cosine-series-based general design procedure for the newantialiasing function. We applied this new procedure to implement theantialiased Fujisaki-Ljungqvist model. We also applied it to revise ourprevious implementation of the antialiased Fant-Liljencrants model. Acombination of these signals and a lattice implementation of the time varyingvocal tract model provides a reliable and flexible basis to test fo extractorsand source aperiodicity analysis methods. MATLAB implementations of theseantialiased excitation source models are available as part of our open sourcetools for speech science.