Method and device for binaural audio signal processing

Abstract

A method and device for processing binaural audio signals. The method comprises: subtracting original audio signals of a left sound channel and a right sound channel and then performing middle-frequency band-pass filtering thereon to obtain a middle-frequency background sound signal; superposing the original audio signals of the left sound channel and the right sound channel and then performing low-frequency band-pass filtering thereon to obtain a low-frequency background sound signal; respectively performing random delay processing on the middle-frequency background sound signal and the low-frequency background sound signal and superposing same, and outputting a first left sound signal in a forward direction and outputting a first right sound signal in a reverse direction; respectively convoluting the first left sound signal and the first right sound signal with a head-related transfer function to perform virtual locating processing so as to respectively obtain corresponding left and right surround 3D environmental background sounds; and superposing the left surround 3D environmental background sound and the original audio signal of the left sound channel, and the right surround 3D environmental background sound and the original audio signal of the right sound channel to form left and right output sound signals having good surround stereoscopic sensations. By means of the above-mentioned method and device, left and right output sound signals having good surround stereoscopic sensations can be obtained.

Description

technical field [0001] The present invention relates to the technical field of sound effect processing, in particular to a method and device for processing binaural audio signals. Background technique [0002] After natural sound is recorded, the original spatial information will be lost when it is played back. Stereo audio technology solves the technical problem of keeping the original spatial information as much as possible when the recorded natural sound is played back. [0003] The sound field is a subjective perception of the human ear to the sound spatial information, which can be divided into two types: localized sound field and diffuse sound field. The localized sound field is the spatial range formed by the human ear's positioning of multiple sound sources, which represents the actual spatial position of the sound source. The diffuse sound field is the spatial range formed by the ambient background sound. It has no sense of reverberation, but has a sense of envelo...

AI Technical Summary

Problems solved by technology

However, the output sound signal obtained by this algorithm only produces a relatively large-scale diffuse sound field feeling in the senses, but this output sound signal does not match the actual listening habits, and the effect is not good when using ordinary headphones to play back. , the sound is not natural

Images