FIELD: information technology.
SUBSTANCE: invention relates to means of exciting a mixed codebook for speech coding. Method of encoding an audio/speech signal involves determining a vector of the mixed codebook basing on an incoming audio/speech signal, herewith the mixed codebook vector includes a sum of the first codebook record from the first codebook and the second codebook record from the second codebook. Method additionally includes generating an encoded audio signal basing on the defined vector of the mixed codebook and transfer of a coded excitation index of the defined mixed codebook vector.
EFFECT: technical result is increased perceptible quality of the speech signal in comparison with coding systems using only pulse excitation or only noise excitation.
26 cl, 17 dwg
