SPEECH CODING SYSTEM WHEREIN NON-PERIODIC COMPONENT FEEDBACK TO PERIODIC SIGNAL EXCITATION SOURCE IS ADAPTIVELY REDUCED

摘要

A speech signal coding apparatus comprising: a periodic excitation signal source (1); a non-periodic excitation signal source (2); a synthesizing unit (3) for synthesizing a periodic excitation signal and a non-periodic excitation signal output from the above sources (1,2) to generate an excitation sound source signal; a filter (4) for regenerating an input speech signal from the excitation sound source signal; and an error evaluating unit (6) for controlling the periodic excitation signal source (1) and the non-periodic excitation signal source (2) so that these units output periodic and non-periodic excitation signals which minimizes a difference of the regenerated signal from an input speech signal. The above excitation sound source when the error is minimized is supplied to a feedback excitation signal modifying unit (7), which reduces a non-periodic component in the above supplied signal according to a relative mount of the non-periodic component in the supplied signal so that the non-periodic component is reduced more when the relative amount of the periodic component is greater. The modified signal is fed back to the periodic excitation signal source unit (1) to modify the content thereof.