Title :
A Low-Complexity Design for an MP3 Multi-Channel Audio Decoding System
Author_Institution :
Dept. of Exp. Psychol., Univ. of Cambridge, Cambridge, UK
Abstract :
In this paper, a low complexity design for a DFT-based MP3 multi-channel audio decoding system is proposed. By utilizing the proposed decoding system, the advanced MP3 multi-channel decoder can play high-quality multi-channel audio with low processing power. The MP3 multi-channel system is composed of two components, a parametric multi-channel decoder, and an MP3 decoder. A novel convolution-synthesis method in the transform domain can replace the pseudo-quadrature mirror filter (PQMF) module in the MP3 decoder and several small size discrete Fourier transform (DFT) modules can be used instead of a large-size DFT module in the multi-channel decoder. This combination noticeably reduces computing power without significant loss in the decoded audio signal quality. The effect of the error in the decoded audio signal is analyzed by both objective and subjective methods.
Keywords :
audio coding; convolution; decoding; discrete Fourier transforms; encoding; pulse height analysers; quadrature mirror filters; DFT-based MP3 multichannel audio decoding system; audio signal quality; convolution-synthesis method; discrete Fourier transform; low processing power; low-complexity design; pseudoquadrature mirror filter module; Complexity theory; Decoding; Digital audio players; Filter banks; Materials; Transforms; Convolution-synthesis method; MP3; MP3 multi-channel; low-complexity decoder; parametric multi-channel;
Journal_Title :
Audio, Speech, and Language Processing, IEEE Transactions on
DOI :
10.1109/TASL.2011.2161081
