A space parameter quantification and entropy coding method for 3D audio signals and its system architecture

Abstract

The invention discloses a space parameter quantification and entropy encoding method for a stereo voice-frequency signal, and a system structure thereof. Firstly, the space parameter quantification for the stereo voice-frequency signal uses different quantization tables in different frequency bands to perform the nonlinear scalar quantification by table search; then quantification indexes of each obtained space parameter are combined into a vector which executes a finite difference with the vector of the former quantification indexes; finally, the obtained finite difference quantification index vector acts the Huffman entropy encoding according to a Huffman code table corresponding to each present frequency band. The invention utilizes the dependency relationship between the hearing and esthesia of the space parameter and the frequency bands as well as the pertinency among the space parameters, thus the subjective and objective redundancy of the space parameter is effectively removed, the tone of the space parameter stereo code is enhanced under the same encoding rate, or the encoding rate is reduced under the same encoding tone.

Description

technical field [0001] The invention relates to a spatial parameter quantization and entropy coding method of a stereo audio signal and a system structure used therefor, and belongs to the field of audio coding. Background technique [0002] Spatial parameter coding of stereo audio signals is an audio coding method that extracts stereo spatial information and parametrically represents it. Spatial parameters include interchannel time difference (Interchannel Time Difference, ITD), interchannel intensity difference (Interchannel Level Difference, ILD), and interchannel correlation (Interchannel Coherence, IC). Since the bit rate of the parameter is usually much lower than the bit rate of the traditional coding of each channel alone, the structure of the traditional coding of the downmix channel plus the stereo space parameter coding can greatly reduce the bit rate while maintaining the stereo sound effect, so the stereo space parameter coding becomes An indispensable technolo...

AI Technical Summary

Problems solved by technology

As mentioned above, the problems caused by the use of uniform quantization by BCC, and the problems of sound quality degradation or bit rate increase caused by the use of a single quantization table in the full frequency band of PS and MPEG Surround, do not fully take advantage of the nonlinearity of the human ear's perception of spatial parameters. Frequency dependent properties

In addition, spatial parameters are also interrelated. For example, ILD and ITD depend on the spatial position of the sound source, while PS and MPEG Surround quantize each parameter separately, ignoring the redundancy between parameters, to a certain extent Reduced coding efficiency

Images