METHOD AND APPARATUS FOR HIERARCHICAL ENCODING AND DECODING AUDIO Russian patent published in 2014 - IPC G10L19/24 G10L19/28 

FIELD: information technology.

SUBSTANCE: method for hierarchical encoding audio includes dividing frequency domain coefficients of an audio signal after modified discrete cosine transform (MDCT) into a plurality of coding subbands, quantising and encoding values of the amplitude envelope of the coding subbands; distributing bits into each coding subband of the basic level, quantising and encoding frequency domain coefficients of the basic level to obtain encoded bits of the frequency domain coefficients of the basic level; calculating the amplitude envelope value of each encoding subband of the residual signal of the basic level; distributing bits into each encoding subband of the extended level, quantising and encoding the encoding signal of the extended level to obtain encoded bits of the ecoding signal of the extended level; multiplexing and packing the encoded bits of the amplitude envelope value of each encoding subband, which consists of frequency domain coefficients of the basic level and the extended level, encoded bits of the frequency coefficients of the basic level and encoded bits of the encoding signal of the extended level, followed by transmission to the decoding side.

EFFECT: high efficiency of hierarchical encoding/decoding audio.

22 cl, 12 dwg, 9 tbl