Visualization feedback for musical ensemble practice: a case study on phrase articulation and dynamics

Abstract

We consider the possible advantages of visualization in supporting musical interpretation. Specifically, we investigate the use of visualizations in making a subjective judgement of a student's performance compared to reference "expert" performance for particular aspects of musical performance-articulation and dynamics. Our assessment criteria for the effectiveness of the feedback are based on the consistency of judgements made by the participants using each modality, that is to say, in determining how well the student musician matches the reference musician, the time taken to evaluate each pair of samples, and subjective opinion of perceived utility of the feedback. For articulation, differences in the mean scores assigned by the participants to the reference versus the student performance were not statistically significant for each modality. This suggests that while the visualization strategy did not offer any advantage over presentation of the samples by audio playback alone, visualization nevertheless provided sufficient information to make similar ratings. For dynamics, four of our six participants categorized the visualizations as helpful. The means of their ratings for the visualization-only and both-together conditions were not statistically different but were statistically different from the audio-only treatment, indicating a dominance of the visualizations when presented together with audio. Moreover, the ratings of dynamics under the visualization-only condition were significantly more consistent than the other conditions.