School of Electronic and Information Engineering, Tianjin University, Tianjin, China

---

Abstract—The third part of the AVS family, AVS–P3 is an audio coding standard of China own intellectual property. It is drafted according to the requirement of the advanced audio compression technology with high quality and efficiency in the field of DAB, DVB, DSM, ISM and MMC. In this paper, main algorithm modules of AVS-P3 are analyzed in detail, including Transient Detection, MDCT, Multi-Resolution Analysis in Frequency Domain, Quantization, Post Quantization Square Polar Stereo Coding, and Context Based Coding. Next, AVS P-3 audio Codec is implemented on DSP platform. The demo system could capture original audio input, encode/transmit/decode with AVS-P3 format, and play back in real time. Finally, the result of the subjective test is presented, its quality is a little better than the quality of LAME MP3.

Index Terms—Audio Coding, AVS, Codec, Algorithm.

[PDF]

Cite: Zhang Tao, Dong Yuxi, Zhang Wen, "AVS-P3: Algorithm and Implementation," Journal of Software vol. 5, no. 8, pp. 858-865, 2010.