Sub-band encoding and decoding method and device based on silk codec

Abstract

The invention discloses a sub-band codec method and device based on a SILK codec, belonging to the field of audio codec. The method includes: obtaining a full-frequency time-domain signal corresponding to the current audio frame; decomposing the full-frequency time-domain signal into a low-frequency time-domain signal and a high-frequency time-domain signal; performing SILK encoding processing on the low-frequency time-domain signal to generate a low-frequency time-domain signal Corresponding low-frequency parameters; and encoding the high-frequency time-domain signal according to the low-frequency parameters to generate high-frequency parameters corresponding to the high-frequency time-domain signal; quantizing and compressing the low-frequency parameters and high-frequency parameters to generate a bit stream corresponding to the current audio frame. The present invention allocates more bit resources to low-frequency signals and uses relatively less bit resources to encode high-frequency signals, thereby realizing more reasonable bit resource allocation. It can effectively improve the coding efficiency and preserve the harmonic structure in the high-frequency signal, so as to achieve better listening effect under the same bit rate setting.

Description

technical field [0001] The present invention relates to the field of audio coding and decoding, in particular to a subband coding and decoding method and device based on a SILK codec. Background technique [0002] With the development of the Internet, the demand for voice communication continues to grow. The VOIP (VoiceOver Internet Protocol) technology based on voice packet switching is more and more popular with users for its low cost, easy expansion and excellent call quality. [0003] The mainstream coding method in VOIP technology is SILK coding. The coding method is: model the speech signal at the coding end, disassemble the signal into different system parameters through the speech model, and transmit these parameters to the decoding end through the channel , the decoder solves the relevant parameters, and then restores the speech signal according to the same speech model. [0004] In the process of realizing the present invention, the inventor finds that there are a...

AI Technical Summary

Problems solved by technology

[0005] In the human voice, the high-frequency signal is usually not as rich as the low-frequency signal, and the SILK encoder processes the high-frequency and low-frequency signals separately according to the preset bit resources, so the bit allocation is not efficient enough when encoding wideband speech, making the Reduced efficiency of low frequency signal processing

Images