Network audio mixing processing method and device

Abstract

The invention discloses a network audio mixing processing method and device, relating to the technical field of audio processing. The method comprises a step of performing symbolic accumulation on each audio data, a step of performing segmented compression on the accumulated audio data, a step of determining an attenuation factor of each segment, and a step of acquiring an outputted mixed audio signal. According to the mixing processing method, the method of segmental compression is employed for mixing, that is, the operation efficiency is high, and the mixing is easy to realize in a single-chip microcomputer. Through the actual waveform observation, the method is synthesized, there is no overflow and topping phenomenon after obtaining the mixed sound of the method, the data is compressednear upper and lower limits, the volume is large, the remaining intelligibility is high, and there is no obvious noise interference in listening. A system used in the device of the present invention can realize the function by only using the single-chip microcomputer with an ARMCortexTM-M4 architecture or higher, an existing LWIP protocol stack, the system can support the basic transmission protocol in TCP / IP, since only one single-chip microcomputer for processing is required, an FPGA can be omitted from a circuit, and therefore, the cost is saved.

Description

technical field [0001] The invention relates to the technical field of audio processing, in particular to a network audio mixing processing method and device. Background technique [0002] At present, the common audio mixing processing methods on the market all adopt the existing "normalized audio mixing". The advantage of this algorithm is that it is easy to understand and the operation expression is concise. However, the disadvantages are also obvious. It is necessary to use floating-point calculations or expand floating-points into integers before performing calculations, so it is difficult to use a single-chip microcomputer to complete the calculation efficiency; and the mixing data at the current moment will be affected by the mixing data at the previous moment. If packet loss occurs. [0003] In addition, in the industry, the FPGA+SCM architecture is used for conference audio transmission. For example, Patent No. 201010593939.5 (a multimedia digital conference system...

AI Technical Summary

Problems solved by technology

[0004] This method of transmission is stable, but the disadvantages are also obvious

[0005] (1) The cost is high, this is because the system uses dual chips, FPGA+SCM;

[0006] (2) The transmission protocol is inflexible, because the network audio part is processed by FPGA, and the logic of FPGA is poor, so the system does not have a complete TCP / IP protocol stack;

[0007] (3) The development time is long, and the FPGA and single-chip microcomputer need to be developed separately, and relevant agreements and long-term debugging need to be formulated between the two chips;

[0008] (4) There are many developers, and professional engineers with FPGA and single-chip microcomputer are needed to complete the design of the entire software; and when a problem occurs, one cannot handle and modify the software BUG alone

Images