Re-Targeting Expressive Musical Style using a Machine-Learning Method
Simon Lui, Andrew Horner · 2009
Expressive musical performing style involves more than what is simply represented on the score. Performers imprint their per-sonal style on each performances based on their musical under-standing. Expressive musical performing style makes the music come alive by shaping the music through continuous variation. It is observed that the musical style can be represented by appropri-ate numerical parameters, where most parameters are related to the dynamics. It is also observed that performers tends to perform music sections and motives of similar shape in similar ways, where music sections and motives can be identified by an auto-matic phrasing algorithm. An experiment is proposed for produc-ing expressive music from raw quantized music files using ma-chine-learning methods like Support Vector Machines. Experi-mental results show that it is possible to induce some of a per-former’s style by using the music parameters extracted from the audio recordings of their real performance. 1.