An efficient low complexity encoder for MPEG advanced audio coding

MPEG advanced audio coding (AAC) is the current state of the art in perceptual audio coding technology standardized by MPEG. Nevertheless the real-time constraint of reference software implementation for the AAC encoder provided by MPEG leads to a heavy computational bottleneck on today's portable devices. In this paper we review briefly the algorithms specified in the MPEG AAC standard and analyze the algorithmic performance. Then we carry out algorithm level optimizations in order to get a low complexity as well as high quality audio encoder that can be easily ported onto multiple platforms. In psychoacoustic model, a MDCT-based psychoacoustic analysis and block switching algorithm in time domain and a tonality detection measure are presented. Moreover, we employ improved pre-processing and adjustment of scale factor in quantization module. Experimental results show the optimized encoder can save about half of overall computational complexity. According to subject evaluation result, the perceptual audio quality of improved encoder is better than the AAC LC reference encoder