A speech encoder with a low data rate, and low delay.

摘要

A vector quantizing speech coder wherein the originally sampled speech signal s(n) is, prior to being quantized at a low bit rate into a device (23), first decorrelated into a residual signal r(n) using Short-Term-Predictive filter (10) the coefficients of which are dynamically derived from a reconstructed residual signal r min (n).