An alternative for the LF model

Abstract

An alternative for the Liljencrants-Fant (LF) glottal-pulse model is presented. This alternative is derived from the Rosenberg model, so we have caned it the Rosenberg++ model. It is described by the same set of Tor R parameters as the LF model but has the advantage over the LF model that it is computationally simple, which allows its use in real-time speech synthesizers. The Rosenberg++ model is compared with the LF model in a psycho-acoustic experiment, from which we conclude that in a practical situation it is capable of producing synthetic speech which is perceptually equivalent to speech generated with the LF model.