Speech coding

Abstract

The variable bit rate speech coding method determines for each subframe a quantization vector d(i) comprising a variable number of pulses. The excitation vector c(i) used to excite the LTP and LPC synthesis filters is obtained by filtering the quantization vector d(i) with a gain value g c is determined to scale the pulse magnitude excitation vector c(i) such that the scaled excitation vector represents the weighted residual signal ∴ that remains after removing redundant information in the subframe speech signal by LPC and LTP analysis In the sub-frame speech signal. The predicted gain value ∴ is determined from previously processed subframes, and when the magnitude of the vector residual signal c(i) is scaled according to the number of pulses m in the quantization vector d(i), ∴ is contained in the excitation vector c The energy E in (i) c The function. Quantized gain correction factor γ gc Gain value g can be used c and predictive gain value ∴ determined.

Description

technical field [0001] The present invention relates to speech coding, and more particularly to coding speech signals in discrete time frames containing digitized speech samples, but is particularly applicable, though not necessarily, to variable length bit speech coding. Background technique [0002] In Europe, the accepted standard for digital cellular telephony is known by the acronym GSM (Global System for Mobile Communications), the most recent version of the GSM standard (GSM2; 06.60) resulting in a new voice known as Enhanced Full Rate (EFR) The details of an encoding algorithm (or codec). Like conventional speech codecs, EFR is designed to reduce the bit rate required for individual voice or data communications. By minimizing this bit rate, the number of independent calls that can be multiplexed into a given signal bandwidth can be increased. [0003] A usual diagrammatic illustration of a speech coder structure similar to that used in EFR is in figure 1 given in....

AI Technical Summary

Problems solved by technology

Because the energy prediction of equation (4) is based on previous subframes, the prediction value may be poor in the case of a large reduction in the number of excitation pulses

This leads to a prediction gain relatively large error in , causing the gain correction factor to vary widely over the entire speech signal

Images