Editor: Jacques Steyn
This book is about Music Informatics - how ICT (Information and Communication Technology) is used in the domain of music, and more specifically, using XML languages to label the semantics of the structures of the information architecture of music.
Very crudely, ICT is about data, process and actor, where actor could be a human agent or another IT system. XML is about the data of ICT, not so much about process or actor, although the data or information about processes and actor could be expressed in XML. Structured data enables computers to find and manipulate specific data. The tables of structured data are semantically ignorant. Today, with much more powerful computing power, data searches and manipulation does not require structured data, as such processes could be performed on unstructured data. But to give sensible results, semantics is required, at which computers are notoriously poor. Markup languages is an attempt to add semantics to unstructured data. The very method of applying markers to information imposes structure on the data, which could then be used effectively by computers. Markup languages could of course also be used to mark the information of structured data. Markup languages are good not only at structuring information, but also at clarifying relationships between such structures. This is achieved by declaring hierarchical relationships, where vertical nodes imply inheritance, while horizontal nodes imply peer relationships.
The architectural structure of a building can be used as analogy to explain information architecture. A building has important structures, such as beams, foundations, walls, doors and windows. These could be aesthetically designed in many different ways. While building architecture is about both structure and aesthetics, information architecture is concerned only with structure. Markup languages describe information structures and their relationships, while visual information aesthetics is achieved by applying style sheet languages to the information structures.
The hierarchical relationships of information structures need to be constructed by designers. The discipline of ontology, defined not as in philosophy, but as in Information Systems, investigates how such relationships could be defined, while the results of such an investigation constitute a particular information architecture. Information architecture is about information structures and their relations within the information space.
This book is thus about structuring music, and marking those structures so that the music data can be manipulated with ICT. As is evident from the different domains of music covered by the different authors in this book, music, or the components of music regarded as important to describe, contains a vast variety of possible information structures.
In the chapter, The information architecture of music, Steyn explains the relevance of ontology and information architecture to determine what should be marked. The word "music" has many different meanings, as is true of so many other words used in human language. It is possible to design an XML language to describe and mark any of the domains of meanings of music. The investigator will first need to demarcate the field that needs to be marked, and then analyze the information structural components of the field and indicate their relations. The final step is to design a particular markup language around the structures.
The chapter The Physics of Music, by Jyri Pakarinen, approaches music from the perspective of acoustics of the soundwaves of music, and regards the fundamental frequency as one of the most important parameters for the design of a music markup language. Not only the acoustics of instruments are important, but also the acoustics of the room in which the music is performed, as that is where listeners process the music. This investigation into the internal structures of music can be regarded as an analysis of the intrinsic features of music.
In Expressing Musical Features, Class Labels, Ontologies and Metadata Using ACE XML 2.0, Cory McKay and Ichiro Fujinaga approach music from an extrinsic point of view. Automatic retrieval of music related information is very challenging, while classification can never be absolute. The definitions of a piece of music might be arbitrary and different people might class it in different genres. Yet most classification engines restrict classification into single genres. One of the design goals of ACE XML is to address this shortcoming. The generality and modularity of ACE might also allow its functionality to be extended beyond just being a classification tool. It might also be possible to address the symbolization of intrinsic music events with this system.
Antoine Allombert and Myriam Desainte-Catherine's chapter, Towards an encoding of Musical Interaction, covers interactive scores. Traditional scores are passive, closed entities, and applications typically merely render this static form in yet another static form. For example, if a score is in the MIDI format, tempo variations cannot be introduced by the musician during real-time performances - only during recording. Their proposal is to allow for such static scores to follow the dynamics and tempo variations of real-time performance.
Wijnand Schepens and Marc Leman's chapter, Chronicle: XML-representation of symbolic music and other complex time structures, create a universal timeline used for synchronizing information from many different sources in addition to music related events, including poetry, choreography, and task scheduling.
The chapter Representing music as work in progress, by Gerard Roma and Perfecto Herrera is about collaborative composition over networks, specifically for the domain of social networking. By using a visual interface, networked users can compose songs jointly by selecting among audio samples.
Alexander Refsum Jensenius developed a Gesture Description Interchange Format (GDIF) to capture the body movements during performance. His chapter, Structuring music-related movements, investigates taxonomies of the different types and layers of movements, as these and not technical matters were the issues in the project.
The chapter Expressiveness in music performance: analysis, models, mapping, encoding, by Sergio Canazza, Giovanni De Poli, Antonio Rodà and Alvise Vidolin, is about emotional expression during performance, especially when needed to be captured for machines that lack emotions. Music is presented in multiple layers at different levels of abstraction - multidimensional spaces - as music needs to be understood as an integration of many subsystems. Emotional expression is one such subsystem.
Michael Good created an XML application (MusicXML) for Common Western Music Notation (CWMN), which has been a commercial success. In his chapter, MusicXML: The First Decade, he gives an overview of this development and some of the challenges encountered during this journey.
In the chapter Universal information architecture of acoustic music instruments, Jacques Steyn proposes a high-level generic description to account for the acoustic properties of music instruments, not from the acoustic soundwave point of view, but from the point of view of the structures that play a role in the creation of the unique sounds of different acoustic music instruments.
Music is a very complex entity, consisting of many domains. The information architecture of any of these domains, and in fact, any of the subdomains of each of these domains, can be expressed in some or other form and information architecture. In the past most XML applications of music focused on CWMN. In this book, apart from Good's historical overview of MusicXML, none of the chapters is about CWMN, but about other aspects of music. There is a vast field that awaits cultivation of information architecture descriptions, while an integrated strategy should perhaps rather be followed. Perhaps in the first phase efforts should focus on establishing ground rules for a generic music ontology - so that we know which concepts would form the core around which XML languages could be developed to better capture the semantics of the information architecture of music.