Information embedded scalable to lossless (SLS) audio

• Main Author: Hoo, Qing Yu.
• Other Authors: Shen Zhiqi
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Computer science and engineering::Data::Coding and information theory
• Online Access: http://hdl.handle.net/10356/17747
• Institution: Nanyang Technological University

I2R, A*STAR, has recently collaborated with German Research Institute Fraunhofer to develop a new lossless MPEG-4 SLS (Scalable to Lossless) audio codec. SLS is able to achieve lossy to lossless and fine granular bit-rate scalability, and allows encapsulation of side information such as lyrics, album art, or even low-resolution music video. SLS is also backward compatible with MPEG-4 coding standards when coupled with a MPEG-4 AAC core. This project explores the audio coding techniques used in SLS. However, before that, background knowledge on the lossy coding techniques used, particularly in the case of AAC, would be studied. This provides basic understanding in the structure of audio coders before zooming into the novel spectral analysis, error mapping, and entropy coding used in SLS. Subsequently, a Graphic User Interface capable of supporting SLS playback is implemented on a Windows platform using C++. The developed music player, Moomba, is able to provide basic play, pause, stop functionalities, and displays the lyrics embedded as side information in the SLS file. An input .wav file could also be encoded in SLS format with options available for the end-user.