MPEG-2先進音訊編碼理論在ARM處理器之實現

Abstract

MPEG-2 Advanced Audio Coding (MPEG-2 AAC)是由Moving Picture Experts Group (MPEG)所發展的音訊壓縮標準，其主要是利用感官式音訊編碼（Perceptual Audio Coding）技術作為其主要的架構，提供多聲道高品質的環場音效。 ISO國際標準組織已在1997年4月將AAC制訂為國際標準(ISO 13818-7)。 AAC捨棄與MPEG-1/2 Audio的相容性。沒有向後相容的限制，它可以改善編碼時的效率。AAC的壓縮架構主要是根據MPEG-1/2 Layer-3的架構，經過一些修改，另外增加一些新的壓縮機制，使其在高壓縮率時能維持聲音的音質。根據MPEG官方測試，96kbps AAC比128kbps MP3有更好的聲音品質。 在本篇論文中，我們將介紹AAC的編碼架構，並在ARM7 32位元處理器上實現AAC即時解碼器。

MPEG-2 Advanced Audio Coding (MPEG-2 AAC) is a high-quality perceptual audio coder developed by Moving Picture Experts Group (MPEG). It provides high-quality multi-channel surround audio. It has been standardized as ISO 13818-7 in April 1997. AAC gives up the backward compatibility with MPEG-1 Audio. Without any compatibility restrictions, it can improve coding efficiency. AAC follows the same basic coding scheme as MPEG-1/2 Layer-3, but improves on Layer-3 in a lot of details and uses new coding tools for improved quality at low bit-rates. MPEG formal listening tests have demonstrate AAC is able to provide better audio quality at 96 kbps than MP3 at 128 kbps. In this thesis, AAC algorithm will be discussed and a real-time MPEG-2 AAC decoder on ARM7 32-bit RISC is implemented.