In the study of musical performance and practice it can be difficult to precisely analyse the playing and practicing process. Traditional methods of data collection have relied upon participant feedback and video or audio recording of the performance. Traditional methods of data analysis have relied upon the researcher’s comparison of the performer’s interview transcription with the actually observed event. This paper describes how a new, improved method of data analysis, the Note-Time Playing Path (NTPP), can provide a high degree of detail of the musician’s performance and use audio to MIDI software to precisely analyse the data. Sound recordings of practice sessions and performances are converted by high precision audio to MIDI conversion software allowing precise tagging and measurement of note-onset times of pitches and any other musical events of interest. Notes and other events are labelled sequentially according to the cumulative beat number they occupy. The labels and their corresponding timings are then exported into a spreadsheet, with timings converted into appropriate units and graphed as a scatterplot. This NTPP visually displays the time (y-axis) at which each note (sequential note number in original piece on x-axis) was played. The performance representation commences at the top of the y-axis and unfolds down the page. The method can be combined with other established research techniques such as error analysis, video coding, think-alouds, participant interview and data analysis, to provide a rich combination of modes of data. Visual observation of these NTPPs reveal tempo variations in musicality and can indicate technical accuracy as they provide precise information about: pitching errors, the speed of performance, repetitive practice behaviours, tempo (including expressive) variation in performance, chunking of material and actual time spent on aspects of learning (e.g., playing, not playing). Potential applications of this method are in the area of music psychology (cognitive processing), music education (pedagogy) and expressive timing research.
Authors: Deborah de Graaff and Emery Schubert
Event: SF08: Search and Information Extraction from Audio Data Workshop