A new model for the source of vowels based on the vortex sound

This paper presents a new model for the source of vowels based on the vorte x sound generated at the glottis by an interaction between the kinetic vibr ation of the glottal muscle and the airflow through it. This new model dist inguishes itself from previous models in presenting analytical (mathematica l) expression of the waveform generated at the glottis, because all of the previous models regard the vocal tract as the primary source for determinin g the waveform of the sound, whereby simply assuming some arbitrary wavefor m of the sound at the glottis. Our new model states that the airflow through the glottis generates vortice s, which subsequently generates acoustic sound, as Lighthill's general theo ry of aerodynamic sound has stated and that, in some cases, the acoustic so und causes a significant change in the airflow pressure affecting, in turn, the glottal movement. Based on the model presented here, some vowels have actually been synthesiz ed by computer programs written in C which generate WAV-format fries on MS Windows. Furthermore, using the new model, reasonable explanations are provided for both the issue of duty ratio of the glottis' opening and closing movement a nd the nonlinear nature of the vocal phenomenon, in particular the mechanis m of the falsetto. (C) 1999 Elsevier Science Inc. All rights reserved.