An XML Multi-layer Framework For Music Information Description

Music is a rich and complex kind of communication, which can be represented in a number of forms and conveyed through different media. In the digital age, common file formats for music encoding are usually aimed at a detailed description of only one aspect at a time: logical, structural, aural, interpretative, notational, typographic, or graphical. It is the case of JPEG and TIFF files for score scans, or MP3 and WAV files for performances. In this work we will analyze the heterogeneous contents involved in a comprehensive description of music, organizing them according to a multi-layer structure. Layers are means to provide a top-down approach in music description, by beginning at symbolic layer (i.e. music symbols in a score) and working its way down to audio layer. In other words, each layer corresponds to a different degree of abstraction in music information. In particular, our approach arranges music contents in six layers: General, Logic, Structural, Notational, Performance, and Audio. Each layer can contain one or more instances, i.e. objects of the same type: for example, in Audio layer a number of audio files can be described. In order to reflect the aforementioned, multi-layer structure, we are developing MX, a new format based on Extensible Markup Language (XML). In a single MX file, music symbols, printed scores, audio tracks, computer-driven performances, catalogue meta-data, and graphic contents related to a music piece can be linked and mutually synchronized within the same encoding. Among other advantages, the multi-layer structure allows us to gather and organize heterogeneous contents leaving them encoded in well-known and commonly used formats aimed at music description. For example, MX supports: EPS, JPEG, and TIFF formats for notational aspects; MP3,WAV, and WMA files for audio tracks; MPEG, Quicktime and WMV for videoclips. MX inherits all the peculiarities of an XML-based format: e.g., it is open, free, easily readable by humans and computers, and editable by simple and common software applications. Moreover, most features of an XML format well suit to music: XML is strongly structured, as well as most music scores; an XML format can be extended, supporting new notations and music symbols; XML provides a means of music interchange over the Net. All the characteristics we mentioned before represent themselves valid reasons to adopt an XML-based encoding for scores. But our XML format, namely MX, presents further advantages. First, the music description provided by an MX file is flexible and potentially very rich, as regards both the number and the type of media involved. In fact, thanks to our approach a single file could contain one or more descriptions of the same music piece in each layer. For example, if we had to provide a description of an operatic aria, the MX file could house: • The catalogue metadata about the piece, its author(s) and genre; • The corresponding portion of the libretto; • The scans of the autograph version and of a number of printed versions; • Audio files containing different performances; • Related iconographic contents such as sketches, on-stage photographs, and playbills. MX is aimed at a comprehensive description of music; as a consequence, software applications based on MX format allow an integrated enjoyment of music in all its aspects. This is a very relevant point of our research activity, and the last part of the text will describe the development and testing of software prototypes illustrating MX potential. Another key feature from both a theoretical and an applicative point of view is the full synchronization among objects within a layer (intra-layer synchronization) and among objects from different layers (inter-layer synchronization). Synchronizations take place both in time and in space dimensions. These characteristics on the one side let the user enjoy different aspects of music simultaneously, and on the other side let the user switch from a media object to another in real-time. For instance, it is possible to compare the vocal performance of Enrico Caruso to the one of Luciano Pavarotti while following the same score, as well as it is possible to compare the music symbols notated on the autograph score to their translation in a printed version while listening to the same audio clip. Moreover, other advanced features are supported by implementations based on the format: for example, the user can view and manipulate the structure of the piece he/she is listening to. As indicated by the title of this work, the features of MX format constitute an XML-based framework for music information description, in its most comprehensive meaning, and MX itself is the base to build applications for an advanced music fruition.