“Noise performer” 23N (a.k.a. Stéphane Perrin) has created a performance called Nunk on Droise.

Simply put, the performance consists in dynamically generating noise music from the alcohol drunk by the performer during the performance.
During the performance, the performer drinks alcohol and several breathalyzers are used to generate sounds and interacts with the visual. In addition, the abuse of alcohol inevitably leads to uncontrollable results and the body of the performer becomes itself a musical instrument.

Several alcohol sensors output each a voltage that depends on the alcohol content in the breath of the performer. These voltages are measured by an Arduino board and sent to a program written under OpenFrameworks that processes them and sends them through OSC (Open Sound Control) to a Pure Data patch. The patch dynamically generates sounds from the received data. In addition, the use of a microphone allows the sound emitted by one of the (un)desirable effects of the consumption of various alcohols in a very short time, namely vomiting, to be processed too by the Pure Data patch.

See Perrin’s Flickr page for more images, as well as YouTube videos of his Nunk on Droise performances: part one and part two.