Digital audio compression and decompression is important in multimedia applications. Dolby AC-3 is widely used as an international standard of audio compression. The AC-3 digital compression algorithm can encode the audio signal of 5.1 channels, whose compressed bitrates range from 32Kbps to 640Kbps. Typical applications are satellite and terrestrial audio broadcasting, delivery of audio over metallic or optical cables, and storage of audio on magnetic, optical, and other storage media.
The KAIST Image Computing System (KICS) is developed to realize the real- time MPEG-2 encoding and decoding utilizing the parallel architecture of multiple general purpose MVPs (Multimedia Video Processor, TMS320C80). This system employs the Dolby AC-3 for audio coding algorithm.
In this thesis, an implementation of the modified DCT is presented, which is a core part of the Dolby AC-3 algorithm. The specific features of the MVP is analyzed and the MDCT is fully optimized to the MVP processor. The processing time required for real-time MDCT and inverse MDCT is evaluated and the simulation results for SNR are presented when the processing is performed in fixed point.