Encoding, music and multimedia
- Shlomo Dubnov
Many encoding formats exist today to represent music, such as DARMS, NIFF, and MusicXML for score typing and publishing, Csound, MIDI, and SASL/SAOL for computer-generated performances, and AAC, MP3, and MPEG for audio and video recordings. These formats capture specific aspects of music but are unable to encode all of these aspects together.
First proposed in 2001, the IEEE Standard 1599 has been developed to allow interaction with music, such as notes and sounds in video applications, and in ad hoc interactive devices by providing a technological framework that makes prerecorded music and related media content navigable and interactive. This is achieved by the use of layers that combine encodings of music with structural and logical representations to allow alternative versions and random access within the piece. These layers are logically organized and synchronized by XML files consisting of symbols that represent an event, referring and pointing to different instances of the same event in the various layers.
IEEE 1599 employs six different layers. The General Layer contains catalog information, such as genre classification, and other ancillary documents. The Logic Layer is the core of the format; it provides mandatory symbolic description with information about timing and synchronization of music events. The Structural Layer deals with identifying music objects and their relationships and allows representation of different kinds of musicological analyses through manual of automated analyses, chord progressions, and musical algorithms using a formalism called Petri Nets. The Notational, Performance, and Audio Layers contain pre-recorded multimedia aspects of music using existing encoding formats as described above.
Envisioned applications include music publishing, education, musicological research, computer-aided composition, computer games, and entertainment-oriented applications. The standard is also designed to accommodate research results concerning automatic feature extraction, alignment algorithms for layer synchronization, and Music Information Retrieval Models for searching music collections.