SPEECH AUDIO ENCODING DEVICE, SPEECH AUDIO DECODING DEVICE, SPEECH AUDIO ENCODING METHOD AND SPEECH AUDIO DECODING METHOD Russian patent published in 2018 - IPC G10L19/32 G10L19/35 G10L21/388 G10L19/02 

FIELD: physics.

SUBSTANCE: invention relates to audio encoding and decoding devices. Time-domain speech input signal is transformed into a frequency-domain spectrum. Frequency region of the spectrum in an extended band is divided into a plurality of bands. Limited band is set for a respective divided band, when a difference between a first frequency with a first maximum amplitude in a spectrum of the divided band in a preceding frame and a second frequency with a second maximum amplitude in a spectrum of the divided band in a current frame is below a threshold, a width of the limited band in the current frame being narrower than the divided band, and the limited band including the first frequency. Spectrum in the limited band in the current frame is encoded for transmitting to a decoder side, and a spectrum outside the limited band within its respective divided band in the current frame is not encoded.

EFFECT: technical result is reducing the number of bits necessary to encode the spectrum, while maintaining sound quality.

12 cl, 24 dwg