Decomposing a Composition: On the Multi-layered Analysis of Expressive Music Performance
Esther Coorevits, Dirk Moelants, Stefan Östersjö, David Gorton, Marc Leman · Lecture notes in computer science · 2016
In our engagement with music, not only the physical experience of sound is important. Also the interplay between body movements, musical gestures and the cognitive processes of performers and listeners is part of our experience. Yet, this multimodal aspect is not always fully considered when analyzing music performance. In this paper, we want to establish a framework for a multi-layered analysis of music performance, building on data retrieved from quantitative and qualitative procedures and involving the perspectives of composer, performer and musicologist. The performance of a classical guitarist was analyzed in detail, using both a ‘bottom-up’ approach (audio-analysis and motion-capture) and a ‘top-down’ perspective (annotations from video-footage, perceived phrasing and the composer’s, performer’s and researcher’s perspective). These different analytical layers were compared and evaluated, which pointed out that multiple perspectives can reinforce each other in understanding musical intentions and can help detecting mismatches between qualitative and quantitative data. The analytical framework developed could be an important step in the coupling of performer’s intentions with the expressive enactment of a musical score.