The Great Bee Symphony: Translating Waggle Dances into Live Orchestral Music

The Great Bee Symphony is an ambitious cross-disciplinary project to capture the spatial and temporal data encoded in honeybee waggle dances and render them as musical scores performed by human musicians. A camera-equipped observation hive streams video to a computer vision system that tracks each dancer's angle, duration, and intensity. These parameters are mapped to a set of musical rules: dance angle determines pitch interval (relative to a tonic), dance duration determines note length, and waggle frequency maps to dynamic articulation. The output is a live-generated notation displayed on tablet music stands in front of an 8-piece chamber ensemble (flute, clarinet, violin, viola, cello, piano, marimba, and percussion). The piece never repeats — it's a literal snapshot of the hive's foraging priorities at any moment. Rehearsals would involve a 'bee whisperer' (an entomologist) who interprets the hive's mood and adjusts the mapping algorithm for emotional resonance. The absurdity lies in the earnestness: we are giving bees a voice in a concert hall, and expecting humans to follow their directions.

Current methods of studying animal communication are analytical and rarely produce aesthetic experiences. Meanwhile, human art often explores nature's patterns (birdsong, whale calls) but seldom uses real-time, non-anthropomorphized data. The hive's waggle dance is incredibly rich: it contains distance, direction, quality of food, and even social excitement. Yet it remains siloed in scientific papers. This project bridges the gap by inviting both scientists and musicians to listen to bees in a new way — and to ask if we can jam with them.

• Build a computer vision system that reliably tracks multiple waggle dances in real time.
• Develop a mapping algorithm that translates dance parameters (angle, duration, intensity) into musical notation with human- playable constraints.
• Design a digital sheet music display that updates live as dances occur.
• Compose a flexible 'frame' composition that accommodates the bee-generated material.
• Rehearse and perform a 30-minute live piece with an 8-piece chamber ensemble.
• Record the performance and publish a paper on the artistic-scientific methodology.

• Not claiming that bees 'compose' music intentionally.
• Not attempting to communicate with bees through music.
• Not aiming for commercial music release.
• Not replacing the entomologist's role — human interpretation remains key.
• Not requiring any modification to the bees or hive (all observational).

• Observation hive with glass walls and infrared lighting.
• High-resolution USB cameras (4K, 60 fps) with macro lenses.
• Computer vision software: OpenCV + custom Python scripts for tracking bees and decoding waggle runs.
• Raspberry Pi cluster for edge processing.
• MIDI-to-notation engine: custom Max/MSP patch maps parameters to MusicXML.
• Digital sheet music app (e.g., forScore on iPads) receiving updated PDFs via network.
• Musical instruments: flute, clarinet, violin, viola, cello, piano, marimba, percussion (glockenspiel, crotales).
• Microphones and recording gear for performance capture.

The system is divided into three stages: (1) Hive Observation, (2) Dance Interpretation, (3) Musical Rendering. In stage 1, cameras capture the dance floor of the observation hive. Video is sent to the Raspberry Pi cluster, which runs a tracking algorithm that identifies the dancing bee and follows her path. We use a modified version of existing bee-tracking software (e.g., BEEtag) to output a series of (angle, duration, intensity) tuples per dance. In stage 2, the Max/MSP patch receives these tuples via OSC. Angle is mapped to scale degree (e.g., 0° = tonic, 90° = perfect fifth, etc.) clamped to a diatonic set. Duration is mapped to note length (log scale: 0.5s waggle = eighth note, 4s = whole note). Waggle frequency (intensity) maps to dynamics (ppp to fff) and articulation (legato vs staccato). The patch also handles polyphony: if multiple dances overlap, they become different voices in the score. In stage 3, the patch generates MusicXML in real time and sends it to the iPads via a server. A human conductor (the entomologist) can intervene with a control panel to transpose, adjust tempo, or mute voices if the bees get too chaotic. The composition framework is 'Meadow' by a human composer, which provides a loose structure (A-B-A form) but fills in the melodic/harmonic content from the bees.

• Computer vision fails in low light or when bees cluster.
• Dance-to-music mapping may sound terrible (but that's part of the art).
• Bees may stop dancing during rehearsal or performance.
• Human musicians struggle to follow an unpredictable, real-time score.
• Entonologist's intervention biases the output too much.
• Audience may not appreciate the conceptual depth.

• How do we handle the 'silence' when no bees are dancing? (Use a drone pedal or pre-recorded bee hum.)
• Should we include other hive sounds (buzzing, piping) as ambient texture?
• How do we ensure ethical treatment of bees (stress from observation lights)?
• Can we crowd-fund a bee-inspired concert without being mocked?
• Is the score meaningful to humans or just a gimmick? (We'll never know.)

The project was planned in 2019 as a collaboration between a composer and an entomologist, but funding fell through when a grant was denied. The COVID-19 pandemic ended in-person rehearsals, and by 2021 the team had moved on to other projects. The idea remains a what-if — a joyful absurdity that still sparks wonder.

Inspired by the work of Karl von Frisch and later artists like Marcus Maeder (Plants, Algae). We wanted to extend the idea to social insects and live performance. The absurdity is central: treating bees as composers is both ridiculous and profound.