From music symbolic information to sound synthesis : an XML-based approach

This paper deals with the automatic generation of computer-driven performances and related audio renderings of music pieces encoded in symbolic format. A particular XML-based format, namely MX, is used to represent the original music information. The first step is illustrating how symbolic information can originate music performances. The format we have chosen to represent performance information is Csound. Then, an audio rendering is automatically produced. Finally, we will show how the aforementioned computer-generated information can be linked to the original symbolic description, in order to provide an advanced framework for heterogeneous music contents in a single XML-based format.