Source- lter model of speech production

continuous signals. In practice, speech is analyzed over a very brief duration, say 20 or 30 ms. That is, the ... to be de ned over the whole of Z. For a signal f[n] that is de ned on a nite domain D, ... discrete lter. Throughout, the vocal tract is modeled as a discrete lter; hence, we restrict our attention to discrete lters in the sequel ... ................
................