Improving Musical Expressiveness by Time-Varying Brightness Shaping

Former studies show that removing brightness temporal variations of tones in an expressive clarinet performance induces a decrease in the appreciation judgments among listeners. This study aims to investigate the reciprocal process where a time-varying brightness pattern is extracted from a clarinet performance and added to MIDI sound sequences generated by sampler-based instruments. Performances with and without brightness temporal variations have hereby been generated using samples from three sustained instruments (piccolo flute, clarinet and cello) and two non-sustained instruments (guitar and sitar). 20 listeners were asked to rate the dissimilarities between the performances in terms of interpretation, and then to select the performance they preferred. Results show that the application of the brightness temporal variations extracted from the clarinet performance increases the appreciation judgment of the performances for almost all the instruments used in the experiment.