SeedBloom is an experimental generative MIDI instrument system I am currently developing in Python, inspired by the ideas of adaptive and evolving music systems explored by artists such as Brian Eno. Rather than functioning as a conventional sequencer, SeedBloom acts more like a collaborative musical organism, responding to live MIDI input, harmonic gestures, timing, velocity and evolving rule sets to create continuously shifting musical behaviour.

The system is designed around live interaction with hardware synths, DAWs, MIDI devices, sensors and data sets. Incoming MIDI notes can be used to seed the harmonic language of the engine, allowing the performer to influence and steer the system in real time. Multiple independent generative voices operate simultaneously, each with its own density, mutation behaviour, rhythmic activity, register and probability model. The result is somewhere between composition, improvisation and systems design.

Current development includes live MIDI routing, probabilistic sequencing, adaptive phrase generation, evolving harmonic pools, chord and rhythm analysis, MIDI CC modulation and reactive performance behaviour. Future plans include microphone and audio analysis, OSC support, MIDI clock synchronisation and eventual migration toward a dedicated JUCE-based instrument and plugin architecture.

The project sits at the intersection of generative music, sound design, live performance systems and experimental instrument design. It is also feeding into my wider work exploring custom MIDI controllers, sensor-based interfaces and alternative methods of interacting with music technology systems.