ENCODER, DECODER AND METHODS FOR BACKWARD COMPATIBLE SPATIAL ENCODING OF AUDIO OBJECTS WITH VARIABLE AUTHORIZATION Russian patent published in 2018 - IPC G10L19/08 G10L19/02 

FIELD: speech analysis or synthesis; speech recognition.

SUBSTANCE: invention relates to means of encoding and decoding audio signal. Decoder comprises an unmixing information determiner for determining unmixing information by receiving the first parametric auxiliary information about one audio object signal and the second parametric auxiliary information about one audio object signal, frequency resolution of the second parametric auxiliary information is higher than the frequency resolution of the first parametric auxiliary information. In addition, the decoder comprises an unmixing module for applying unmixing information to the downmix signal indicating downmixing of one audio object signal, to obtain an audio signal that has been unmixed, comprising said plurality of unmuted audio channels. Unmixing information determiner is configured to determine unmixing information by modifying the first parametric information and the second parametric information for obtaining modified parametric information.

EFFECT: technical result is higher quality of a coded audio signal.

18 cl, 21 dwg