IMMERSIVE VOICE AND AUDIO SERVICES (IVAS) WITH ADAPTIVE DOWNMIXING STRATEGIES Russian patent published in 2024 - IPC G10L19/08 G10L19/24 H04S5/00 

FIELD: means for encoding and decoding an audio signal.

SUBSTANCE: input audio signal represents an input audio scene and comprises a main input audio channel and additional channels. Type of the downmix coding scheme is determined based on the input audio signal. Based on the type of the downmix coding circuit, calculating one or more input downmix gains, to be applied to the input audio signal to construct the main downmix channel. Downmix input gains are determined to minimize total prediction error on additional channels. One or more downmix scaling gains are determined for scaling the downmix main channel. Downmix scaling gains are determined by minimizing the energy difference between the reconstructed representation of the input audio scene from the main downmix channel and the input audio signal.

EFFECT: high efficiency of encoding an audio signal.

23 cl, 8 dwg