Mathematical models of vocal tract with distributed sources

Frequcncy dornaiii models for tlic vocal t,ract with distributcd sources are introduced. Pressure and voluinc velocity distributed sources arc inodcled within iiicreniciital lossy cylindrical pipes i n a manner similar to that of inodcliiig electrical cornponciits. Equations for transmission matrices of uniforn-area vocal tract scctioiis with distributed sources are dcrivc:d. Transfer finictioris for pressure and volume velocity at various scgmcnts of t,he vocal tract are coriiputcrl. Cornparison of distributed sourcc niodcls with point sorirce models revcal sevcral iniprovcincnts 011 the traditional limped source rriodclirig rncthod and shows that tlic effects of t,lic finitc irripcdcncc coilstriction arid hack cavity cam iiot bc adcquatcly modclod using poiiit soiirces. The distributed sourcc vocal tract framework is important for building articulatory alii1 acoustic models for fricativt: soiinds and othcr plioiiciric categories. This paper provides sucli a rnathcrriatical framework for tlie first, tirnc.