At higher sample rates (16000+), use asymmetric LF-inspired glottal waveform
with sharp closure (sharpness=3.0) for richer harmonics and "grainier" Eloquence-like
character.
This is based on the Liljencrants-Fant (LF) glottal model from 1985 - the same research that informed Klatt88 and likely Eloquence's DSP!
André Polykanine
in reply to Luna Rose • • •