Ledotron

Our tool for creating and driving light · 2015 – 2026 · From Mexico to China

Our role Software design and development.

Ledotron is the tool we wrote for our own installations. It has run more than forty of them since 2015, from OX to Retour aux Sources. LED strips, moving lights, video walls and the venue itself are drawn in 3D; the show is composed on a timeline, and you see the result before a single cable is pulled.

Three ways to play. On loop, a written timeline runs during opening hours, locked, and restarts by itself if anything goes wrong. By draw, the Brain composes a new show at every press, every hour or every message, from the weather, the season and the date: every visit is unique. Live, the Brain listens to the music and makes the image evolve as much as the music does.

The AI that listens to music comes from Cyril Laurier’s PhD thesis on automatic music mood classification at the Music Technology Group of Universitat Pompeu Fabra, Barcelona (2011). Beats and onsets are detected with aubio, the open-source library born from Paul Brossier’s thesis at Queen Mary University of London.

Inputs: live audio, local weather and climate, clock and calendar, OSC, buttons and keys, Chataigne. Outputs: Art-Net and DMX, video walls, motors, OSC to Ableton Live. Twenty-three generative effects, all able to react to sound or weather. On site: auto-start, locked screen, logging, automatic reconnection and restart. That is what lets us promise a client that an installation will hold for ten years without a technician on site.

Ledotron can be rented or bought, for an installation, a tour or a permanent venue: write to us for a demo or a quote.

Credits

Used on

Photos

Ledotron, Our tool for creating and driving light
Ledotron, Our tool for creating and driving light
Ledotron, Our tool for creating and driving light
The timeline. One track per part of the installation; each block is an effect with its duration. The eye switches a track off without erasing it, the padlock freezes it.
All of Ledotron on one screen. The show timeline at the top, the 3D scene in the middle, clips and effect settings on the left, the Show panel on the right: you compose and check the result in the same place, before being on site.
Realistic rendering. Each lamp is drawn in place with its halo and its haze, as in a room with a fog machine: you see what the audience will see without switching on a single real lamp.
The scenography. Every LED strip or ring is placed in space with its network address and its shape; here ten lamps are selected and moved in one gesture.
Three ways to play. In a loop, by draw (a button, the Brain composes a show, back to standby) or live (the music leads). Launch, lock and volume settings are in the same window.
The Brain. A version is a chain of steps; at each step the Brain draws a clip among those that fit (category, season, nature of the day). Everything is set in menus, without touching the code.
A Brain per project. Each project declares what its Brain uses: a musical project has no seasons, a heritage site follows the calendar and the weather.
The Inputs table. A button, an OSC message or a key, and the action it triggers: start a version, play a clip, mute the sound, switch off. “Listen” learns the next message received.
Listening to music. Level, energy, tempo, mood (calm, happy, sad, excited), dissonance, frequency bands, spectrum and song sections, measured live: the data the live Brain and the sound-reactive effects use.

Videos

Team

  • Maya Benaïnous · Co-founder. Artist and engineer, creator of Ledotron.
  • Cyril Laurier · Co-founder. PhD in music AI, Sony CSL researcher.
  • Paul Brossier · PhD, music developer

Expertise