摘要
為了提高通訊管道的使用效能，音訊的壓縮技術在過去這20年來也發展及演進。在成功制定MPEG-1音訊壓縮標準後，ISO/IEC MPEG委員會繼續開發下一代的音訊壓縮標準MEPG-2 AAC (Advanced Audio Coding)於1997年。AAC是一個音訊/語音編碼技術的大家族，它有著9種不同的規格以因應各種不同的場合，而依照每種規格的不同也能自由選擇各種不同的工具以滿足低位元率及高位元率的傳輸環境。 SLS (Scalable to Lossless Coding)系統是MPEG-4音訊壓縮技術的延伸，它通常發展於AAC的壓縮技術之上儘管它的基本編碼架構有別於其他技術。如同它的名稱，它可以達到所謂的無損(lossless)壓縮，若是我們需要有損(lossy)音訊檔案時，也能夠從位元串之中提取出一部分的資料。換言之，一個編碼過後的位元串可以提供不同位元率下音訊檔案的解碼。在本篇論文中，使用了現有的AAC編碼器並套用SLS，之後並試著提升SLS在低位元率下的編碼性能，而且重新排列了MDCT係數的優先順序來達到這個目標。
並列摘要
To increase the efficiency of channel usage, the compressed audio techniques have been developed and improved in the past 20 years. After defining the successful MPEG-1 audio compression standards, the ISO/IEC MPEG Committee continued to develop the next generation MPEG-2 AAC (Advanced Audio Coding) standard in 1997. AAC is a big family of audio/speech coding technology. It has nine profiles to meet a variety of application scenarios. It has many different tools to meet the diversified requirements of both low-rate and high-rate transmission environments. The SLS (Scalable to Lossless Coding) system is an extension of the previous MPEG-4 audio techniques. It is often built on the top of the AAC codec although its base-layer codec can be the other technology. Its name, Scalable to Lossless Coding, implies that it can produce a bit stream that achieves the lossless compression. However, if we extract a portion of this bit stream according to its syntax, we can play back a lossy version of the original audio. In other words, a single bit stream can provide decoded audios at various bit rates (and audio quality). In this study, we propose an implementation of SLS using an existing AAC codec. And then, we try to improve the SLS coding performance at low bit rates. Rearranging the MDCT coefficient priority in the SLS encoding (and decoding) process, we are able to achieve this goal with little computational overhead.