Audio Signal Processing and Coding

11.1 INTRODUCTION

The emergence of high-end storage formats such as the DVD-audio and the super-audio CD (SACD) created opportunities for lossless audio coding schemes. Lossless audio coding techniques yield high-quality audio without any artifacts. Lossy audio coding (LAC) results, typically, in compression ratios of 10:1–25:1, while the lossless audio coding (L²AC) algorithms achieve compression ratios of 2:1–4:1. Therefore, L²AC techniques are not typically used in real-time storage/multimedia processing and Internet streaming but can be of use in storage-rich formats. Several lossless audio coding algorithms, including the SHORTEN [Robi94], the DVD [Crav96] [Oome96] [Crav97], the MUSI-Compress [Wege97], the AudioPaK [Hans98a] [Hans98b] [Hans01], the lossless transform coding of audio (LTAC) [Pura97], and the IntMDCT [Geig01] [Geig02] have been proposed. The meridian lossless packing (MLP) [Gerz99] and the direct-stream digital (DSD) techniques [Brue97] form a group of high-end lossless compression algorithms that have already become standards in the DVD-Audio [DVD01] and the SACD [SACD02] storage formats, respectively. In Figure 11.1, we taxonomize various lossless audio coding algorithms.

This chapter is organized as follows. In Section 11.2, we describe the principles of lossless audio coding. A survey of various lossless audio coding algorithms is given in Section 11.2.2. This is followed by more detailed descriptions of the DVD-audio (Section 11.3) and the SACD (Section ...