The first step towards a digital model is in representing the tube as a series of concatenated tubes of constant cross-sectional areas, as in Figure 1.1(b) (where =8). The tubes are assumed to be of equal length ; if is the total length of the vocal tract, we have . In the limit as becomes small, the shape of the approximation of the series of tubes will converge to that of continuous vocal tract shown in Figure 1.1(a).

Stefan Bilbao 2002-01-22