Lag windowing

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Lag windowing is a technique that consists of windowing the auto-correlation coefficients prior to estimating Linear prediction coefficients (LPC). The windowing in the auto-correlation domain has the same effect as a convolution (smoothing) in the power spectral domain and helps stabilizing the result of the Levinson-Durbin algorithm. The window function is typically a Gaussian function.

External links[edit]

See also[edit]