The idea for the audiocubes first came to mind when I was discussing tangible user interfaces with some friends who worked as recording and sound engineers. Later on I was researching how this interface could be applied in a scientific context (human-computer interaction) at the DSSP research group.

Stephane Bronckers, Peter Vuchelen who were EE students at the VUB and Jimmy Verhulst, an engineering student of Groep T worked on early prototypes of the hardware and software for the interface, in the context of project assignments or bachelor's thesis.

If you have any suggestions for this project or if you would like to collaborate, please feel free to send me an email. Input from musicians and artists is essential in this intense development process. If you would like to use material from this website, please email me first. You can find more information about me and my work here.


  This is a musical interface I have been working on over the past years. It is a tangible user interface (TUI) consisting of a number of cubes made out of a plastic material. Each cube contains a digital signal processor (DSP) with optical sensors and emitters (infrared, red, green and blue LEDs). The sensors and emitters receive and send audio signals which are generated or processed by the signal processor in the cube. Each cube is powered by a rechargeable battery pack. By positioning the cubes relative to each other and moving them around, a signal processing network can be created. A musician can use this interface to learn a new way of interacting with sound and music.

  On the left you can see the components of an AudioCube. The cube is based on a TMS320LF2407 DSP, which controls 6 RGB LEDs or 18 seperate LEDs, as well as 6 infrared LEDs for audio transmission. Infrared phototransistors are used to receive the audio signals. The cube is powered by a battery pack, which can be charged using the charging circuit in the cube. To give the performer even more control over the algorithm running on the DSP, different sensors could be added, such as an orientation sensor, shock sensor, position sensor, etc.

  The figure on the left shows a possible scenario. In the audio cube network, there can be one or more cubes with an audio input jack and one or more cubes with an audio output jack. The jacks are used to get audio in and out of the system.The yellow cube could be filtering the incoming audio signal from the line in cube. The filtered output of the yellow cube could be received by the blue cube on the left which could then apply another transformation and emit a new signal from the same face, which could then be received by the green cube below which connects to a headphone (if it had a headphone driver of course).
  In parallel with the hardware and software prototyping process which is constantly happening, I am using a variation on the paper prototyping technique to determine how musicians and artists or people in general would use the cubes in their performances.  I hope to extract some valuable usability information from this paper prototype for future publication.
  As a first milestone in the AudioCube project, I built a prototyping system which you can see on the left. The system consists of a hole island eurocard (100mmX160mm) containing analog filters and drivers for the LEDs and phototransistors in the small cubes on the table. The eurocard connects through 2 flatcables to a development board from Spectrum Digital for the TMS320F2812 DSP, which is a more powerful DSP than the 2407, but is code compatible. Up to 6 cubes can be connected to the eurocard, through small flatcables. Each small cube contains an RGB LED and an infrared LED/phototransistor. So the system you see on the left consists of all the electronics I would like to fit inside the AudioCube mechanical design I am currently working on (see below).
  April 2004. The electronics will probably be divided among several small PCBs, which are stacked using board-to-board connectors. There will be a PCB for the battery pack, for the DSP, memory, in-system programming, and for some analog filters and LEDs, as well as audio and power connectors. On the left you can see the mechanical design in SolidEdge.

Click here to view the 3D model in VRML (you can download a VRML viewer here).

  AudioCube Scenario #1: Two cubes (green and red) are emitting a sawtooth wave at frequencies 140 Hz and 280 Hz respectively. The blue cube receives and mixes both signals.

The cubes you see in the video are actually smaller than the complete AudioCube will be. Each of the smaller cubes contains the functionality for one face of the AudioCube. The smaller cubes are connected to a prototyping board (which you can see above).

Download the video here: 3MB Windows Movie or 4MB QuickTime Movie.

Video notes: The audio track for the video was recorded from the prototyping system, to which the blue cube is connected. The noise you can hear in the video before the cubes start to move or when the system is idle comes from the crosstalk between the long cables and the fact that all boards were hand-soldered without proper shielding and separation of analogue and digital circuits.


    Check out some new sound demos: saw tooth (1MB WAV) and noise (1.2MB WAV) oscillators with frequency being adjusted by the intensity of the received light.