Abstract

This paper addresses problems and misconceptions pertaining to the notion of the musical surface, a notion that is commonly thought to be relatively straight-forward and is often taken as a given in computational and cognitive research. It is suggested that the musical surface is comprised of (complex) musical events perceived as wholes within coherent musical streams- the musical surface is not merely an unstructured sequence of atomic note events, such as score notes or a piano-roll representation. Additionally, it is maintained that the emergence of the musical surface involves rather complex mechanisms that require, not only multi-pitch extraction from the acoustic signal, but, the employment of cognitive processes such as beat-tracking, metre induction, chord identification and stream/voice separation. Such processes do not come into play after the surface has been formed, but are, rather, an integral part of the formation of the musical surface per se.