We report a music analysis study of Atmosphères (1961) from György Ligeti, combining symbolic information retrieved from the musical score and audio descriptors extracted from the audio recording. The piece was elected according to the following criteria: (a) it is a music composition based on sound transformations associated to motions on the global timbre; (b) its conceptual creative intercourse makes direct references to electronic music and sound/timbre techniques from the ancient Renaissance Music; and (c) its sonorities are explored by means of variations on the timbre contrast. From the symbolic analysis perspective, Atmosphères’ timbre content can be discussed considering the entanglement of individual characteristics of musical instruments. The computational method approaches the musical structure from an empirical perspective and is based on clustering techniques. We depart from previous studies, and this time we focus on the novelty curve calculated from the spectral content extracted from the piece recording. Our findings indicate that novelty curve can be associate with five specific clusters, and regarding the symbolic music analysis, three leading music features can be argued: (a) instrumentation changes; (b) distinct pitch chromatic set locations and (c) intensity dynamic fluctuations.
Analysis of Ligeti’s Atmosphères by Means of Computational and Symbolic Resources
