“Music theory” might be thought of as a collection of hypotheses about the effect a musical passage may have on listeners in general or certain classes of listeners (e.g. parallel fifths sound bad). Computational music theorists may be interested in testing these hypotheses in some objective sense - determining the musical conditions under which they hold, and for which types of listeners in which types of listening experiences, for example. Measuring how a listener (i.e. their physiology) is affected by music, however, seems to be a fairly scattered discipline, as many methods are costly, noisy, and invasive to the listening experience.
In this project I'd like to share ideas, work on code, and test methods, as time permits, to advance the implementation of an assembly of low-cost, minimally invasive methods of measurement (and, equally important, data analysis) to extract a maximal amount of information about how a listener is physiologically affected by a piece of music (or other media). For example, I have found the $100 Neurosky Mindwave EEG headset (which I will bring) to provide low-noise, usable data, as well as a simple $25 Arduino-based light-based pulse meter. These may provide even more useful data when analyzed in certain ways - for example, extracting heart rate variability from the pulse signal, or, say, alpha wave variability from the EEG signal.
A vague distant goal: a file repository, which anyone can access/contribute to, containing a collection of files in a standard file format containing a) the media file listened to and b) recorded biological signals, each stamped by time, some kind of listener id, and information about the listening set-up.
I'll bring devices and Java code I began developing awhile ago.