Live Granular Engine

„I let the chord ring and tell the software with a gesture to endlessly repeat those exact 500 milliseconds. Then I access the elementary parts of the tone and transform them into something new. I keep on playing and combine the newly generated sound with familiar listening habits up to the point, where I gesture the software once again to endlessly repeat those exact 283 milliseconds.“
– Ralph Rubin

Ralph Rubin

Composition and performance of a sound manipulated piano piece!


Video anzeigen

Beim Anzeigen des Videos wird eine Verbindung zu Vimeo hergestellt.

Work Master thesis by Ralph Rubin in the degree course "Interactive Mediasystems"
University University of Applied Sciences in Augsburg
Reader Prof. Robert Rose and Prof. Dr. Michael Kipp
Written in MaxMSP (v.7)
Used Hardware Leap Motion V2 sensor, self-built 32-pin single-coil pick-up, Oktava Mk 012-01 MSP2 Pair, MOTU Traveler MK3, Behringer FCB 1010

What the hell is going on there?

117934~ is an experiment to extract new sounds from a classical upright piano via sound manipulation. The audience should be presented with a composition of familiar listening habits and newly created sounds. The sound is picked up by a self-built 32-pin single-coil pick-up and two condenser microphones, gets manipulated in a piece of software and is finally presented to the listener.

This work is focused on the development of a software to edit live-recorded sound material in real-time. The MAX/MSP-based software allows the access of the latest 12 seconds of played audio. With the help of a gesture recognition sensor it is possible to choose any sector of this sequence and let it repeat endlessly. The chosen sector gets buffered and can be manipulated by granular-synthesis in real-time. The program flow and parametrization of the granular-synthesis is visualized by a specially developed Graphical User Interface and is controlled by a foot-controller.

117934~ Interface


Here you can choose the audio-device, the inputs, the midi-port and the driver.


This buffer constantly records the last played 12 seconds without any hearable acoustic breakpoints.


As soon as you have catched 12 seconds from the Circular-Buffer, it gets automatically copied into the Temporary-Buffer. There you do the exact selection of your piece of audio by gesture control via the leap motion sensor.


Now you can copy your audio-selection into one of four Grain-Buffers to do granular synthesis. You can manipulate your audio-selection by changeing the following parameters: Grainsize, quantity, amount, speed, stereo-spread, fade and position-variation. With B.I.T. (”BangInTime”) you can force the audio-selection to start from the beginning.


Finally you can put the manipulated audio-selection into a step-sequencer. There you can create rhythmic patterns by controlling the step-sequencer`s interface with the leap motion sensor
Used Max-Externals in 117934~ Akamatsu, Masayuki: aka.keyboard und aka.leapmotion.
http://akamatsu.org/aka/max/objects/ (CC 3.0)
Dobrian, Christopher: tap tempo.
Muir, Chris: EveryNumber.
Pasquet, Oliver: op.buffercopy.
Rozendal, Timo: grainstretch~ (v20120126).

The Live-setup in detail

117934~ Technik Aufbau

One more thing:The 32-pin single-coil pick-up.

A further experiment in 117934~ is a self-built 32-pin single-coil pick-up. A lot of gain is used to push it loud and the sound is quite special and reminds a bit of the sound of a Yamaha CP-80 but convince yourself and watch the video. The sound you will hear is solely recorded with the pick-up, no microphones were used.


Video anzeigen

Beim Anzeigen des Videos wird eine Verbindung zu Vimeo hergestellt.

117934~ 32-pin single-coil
Special thanks to Gerhard Stachulla, Julia Barton, Claus Hoffmann, Professor Robert Rose, Professor Dr. Michael Kipp, Elias Naphausen, Benjamin Paska, Dominik and Valentin Scherer, Jakob Nicklbauer and YEAH.

Kein Tracking – nur Inhalte

Wir verwenden kein externes Tracking, aber binden Videos von YouTube und Vimeo ein. Mehr Informationen dazu in unserer Datenschutzerklärung.