Augmented Violin

Adaptive Musical Instruments

This project is motivated by the question how does one consider learning when designing a new instrument? How does the performer progress from novice to expert? How do we design an instrument that is both accessible and challenging? In order for an instrument to become useful, it must be practiced and how can we, during instrument design, encourage a novice's desire to practice? Is it possible to build an instrument that optimizes its own difficultly in basic performance to match the user's skill level? Can this "adaptive instrument" be designed in a way that compliments end technique while also providing the beginner satisfying musical results, thus improving learning motivation?

The Violin...

The violin is example of an established virtuosic instrument with a motivationally poor learning curve. It takes a long time before a beginner can play pleasantly play a song. Basic violin technique requires a reasonable level of mastery of both pitch determining finger placement (without frets to help out), and also proficiency with the bow, another tricky tool to control. The level of complexity faced when playing even the most simple song is enough to turn many potential violin students away; with less time, and patience, very few adults take up violin from scratch. It just takes too long to build the skills required before a student can play fun music. This excessive early complexity makes the violin an excellent instrument to study the potential for adaptive instruments. It is a known great instrument, but the reward for the amount of effort during early learning is anything but great.

We are looking at whether we can improve the learning experience on a violin by using a combination of sensors and audio analysis to simplify the instrument and reduce learning complexity. For instance, can we use a variant of auto-tune on an electric violin to improve the ability of a student to play in tune? What is the right amount of help to give? How does "the right amount of help" vary with student skill?

Our Augmented Violin

Targets for an augmented violin to support enhanced learning include:
  • Real-time responsive- all sensing and sensor processing complete in less than 10 ms.
  • Inexpensive to build and mount- under $100.
  • Non-intrusive for both the instrument and the player- does not require custom violin or bow.
  • Low-latency high accuracy pitch detection.
  • Low-latency accurate bow pressure and bow position detection.
For a PDF version of the below, please click here.

This project is funded by the UK Engineering and Physical Sciences Research Council (EPSRC) under grant EP/K032046/1 (2013-14).

References (Selected)

  1. L. Pardue and A. McPherson. Near-field optical reflectance sensing for violin bow tracking. Proc. New Interfaces for Musical Expression, Seoul, South Korea, 2013. PDF
  2. L. Pardue, D. Nian, C. Harte and A. McPherson. Low-latency audio pitch tracking: a multi-modal sensor-assisted approach. Proc. New Interfaces for Musical Expression, London, UK, 2014. PDF