ISCA Archive SSW 1994
ISCA Archive SSW 1994

On the use of a sinusoidal model for speech synthesis units

M. A. Rodríguez-Crespo, P. Sanz-Velasco, L. Monzón-Serrano, J. G. Escalada-Sardina

This paper discusses the adequacy of a model that represents the speech signal as a sum of sine waves to the requirements of concatenative speech synthesis in TTS. This model allows wide range and high quality prosodic modifications, and produces smooth transitions where the speech segments are joined. A preference test has been carried out between speech synthesized using a pitch synchronous LPC (PS-LPC) synthesizer and the sinusoidal synthesizer.