InTune: A System to Support an Instrumentalist’s Visualization of Intonation

One of our most beloved music teachers emphasized the importance of “facing the music,” by which he meant listening to recordings of our playing. As with the first hearing of one’s voice on a recording, many of us were both surprised by and suspicious of this external perspective. Despite sometimes revealing more than we are ready to hear, this exercise has the long-term effect of helping us to “hear ourselves as others hear us.” Thus armed, we initiate practice habits that, perhaps over many years, move our music-making toward a state we would admire hearing from another player.

The “face the music” approach begins by accepting that most of us are not born able to judge ourselves objectively, but can learn to do so when given the proper external perspective. We adopt this approach in this article, still in the service of music education, though we use visual, in addition to aural, feedback. Although a visual representation of audio is necessarily an abstraction, it has the advantage that the observer can “visit” the image at will. For instance, the observer may see a note having a certain undesirable (or desirable) property; find the same trait in another note of the same pitch; formulate a hypothesis of systematic error (or accuracy); and validate or refute this theory on subsequent notes. In contrast, audio data must be digested nearly at the rate it comes into the ear.

We apply the “face the music” approach to the practice of intonation—the precise tuning of frequencies corresponding to different musical pitches. Although good intonation, “playing in tune,” is often neglected in the earliest years of musical practice, it is as essential a part of technique as the playing of fast notes or the control of emphasis. Intonation is also central to what some see as the illusion of tonal beauty—that is, for a sound to be beautiful it must (among other things) commit clearly to the “correct” pitch. We introduce a system that allows musicians to visualize pitch in ways that leverage the centuries-long tradition of music notation and are intuitive to the non-scientist.

The electronic tuner is, without doubt, one of the most widely used practice tools for the classically oriented musician, thus justifying efforts to improve this tool. The tuner provides an objective measurement of the pitch or frequency with which a musician plays a note, which can be judged in relation to some standard of correctness (say, equal-tempered tuning at A = 440 Hz). Though the tuner has been embraced by a large contingent of performing musicians, it does have its weaknesses, as follows. The tuner gives only real-time feedback, requiring the user to assimilate its output as it is generated. The tuner takes time to respond to each individual note, making it nearly impossible to get useful feedback when playing notes that are only moderately fast. The tuner cannot handle simultaneous notes, such as double stops. (This is actually part of the reason the tuner fails on fast notes, because past notes linger in the air, confusing the device.) Perhaps most significantly, the tuner does not relate its output through the usual conventions of notated music, thus hiding tendencies and patterns that show themselves more clearly when presented as part of a musical score. Our program, InTune, seeks to overcome these weaknesses by presenting its observations in an intuitive and readily appreciated format.

In what follows, we present our system, InTune, describing the three different views of audio the program allows, as well as the backbone of score-following that distinguishes our approach from others. We consider other approaches to this problem and place ours appropriately in this context. Finally, we present a user study, giving reactions to our effort from a highly sophisticated collection of users. The program was developed in close consultation with Allen and Helga Winold, professors emeriti of music in the Jacobs School of Music at Indiana University, and it is freely available for download at www.music.informatics.indiana.edu/programs/InTune.

Past Work

Two recent examples from...