Empirical Methods for Music Theorists Symposium Details
Sunday November 1
- 2:00-2:15 Introductions
In the last thirty years, the technology to record listeners' experience of music continuously in time has become more and more accessible, developing from a few handcrafted sensors to ipod applications. Now the question is how to analyse these piles of time series data to get statistics relevant to the musical experience. With examples of subjective familiarity and emotional intensity ratings, we will discuss how to handle this highly variable data, common challenges and pitfalls, and how we can look for relationships to musical form and other dynamic musical feature.
Motion capture (MoCap) is a series of techniques that allow for the (very) accurate recording of human movements. Commonly available MoCap systems are based on (active and passive) infrared, electromagnetic, ultra-sound, mechanical, inertial and hybrid technologies. Such systems vary tremendously in complexity and price depending on the technique and the size of the capture space, from a few thousand to several hundred thousand dollars, basically limiting their use to research laboratories. But independent of their price, another limitation for the widespread use of MoCap systems to the analysis of musical performer users is, in many cases, their lack of accurate synchronization between audio and MoCap data. Such a feature is essential in the analysis of music performance if one wants to accurately relate the acquired movement to a musical event (e.g. a note played). In this talk, I will review various MoCap technologies and discuss ways to circumvent synchronization limitations in various MoCap systems available at the Input Devices and Music Interaction Laboratory at McGill. Examples will be provided illustrating the use of MoCap to the analysis and synthesis of performer movements.
- 4:00-4:30 Johanna Devaney Techniques for extracting performance data from audio recordings
There is a wealth of information available in recording performances. Interest in studying this information dates back almost as far as the birth of recordable media. While early studies were extremely arduous and entirely manual, today there are numerous options for semi-automatic and automatic extraction of performance data. This talk will survey a number of such tools and will consider the various challenges that arise when extracting and studying performance parameters.
Monday November 2
- 9:00-10:00 Caroline Palmer The role of sensory feedback in performance
Musicians perform sucessfully under a wide variety of feedback conditions due to changes in musical instruments, architectural environments, and sound transmission issues such as wearing headphones in a studio versus live conditions in a concert venue. The sensory feedback (auditory, visual, tactile/proprioceptive) available to performers differs across these conditions, and in fact is rarely the same from one performance to another. I will review recent research that demonstrates performers' sensitivities and insensitivities to changes in sensory feedback, and how their performances are altered in both solo and ensemble performance. Examples will be provided from motion capture experiments that demonstrate changes in resulting sound and motion of performers due to changes in sensory feedback.
- 10:00-12:00 David Huron From Idea to Experiment
Music theorists have a long history of identifying and framing highly interesting and sometimes vexing problems surrounding music as heard and conceptualized. Studying these problems in a rigorous fashion, however, is another matter. In this 2-hour session, symposium attendees will work through three small-group activites, tracing out a general Question-to-Theory-to-Conjecture-to-Hypothesis-to-Protocol procedure.
Organizers
- Johanna Devaney, Schulich School of Music of McGill University
- Peter Martens, Texas Tech University School of Music