A quantitative, parametric model of musical tension

Abstract

This thesis presents a quantitative, parametric model for describing musical tension. While the phenomenon of tension is evident to listeners, it is difficult to formalize due to its subjective and multi-dimensional nature. The model is therefore derived from empirical data. Two experiments with contrasting approaches are described. The first experiment is an online test with short musical excerpts and multiple choice answers. The format of the test makes it possible to gather large amounts of data. The second study requires fewer subjects and collects real-time responses to musical stimuli. Both studies present test subjects with examples that take into account a number of musical parameters including harmony, pitch height, melodic expectation, dynamics, onset frequency, tempo, and rhythmic regularity. The goal of the first experiment is to confirm that the individual musical parameters contribute directly to the listener's overall perception of tension. The goal of the second experiment is to explore linear and nonlinear models for predicting tension given descriptions of the musical parameters for each excerpt. The resulting model is considered for potential incorporation into computer-based applications. Specifically, it could be used as part of a computer-assisted composition environment. One such application, Hyperscore, is described and presented as a possible platform for integration.