
You can perform this analysis by selecting one or more Sound objects and choosing the appropriate command to generate an LPC.
The acronym LPC stands for Linear Predictive Coding.
In the LPC analysis one tries to predict x_{n} on the basis of the p previous samples,
x′_{n} = ∑ a_{k} x_{nk} 
then {a_{1}, a_{2}, ..., a_{p}} can be chosen to minimize the prediction power Q_{p} where
Q_{p} = E[ x_{n}  x′_{n}^{2}]. 
Several different algorithms exist for minimizing Q_{p}:
© djmw, January 26, 1997