Generative graphics & music

GENIUS-MIDI

A generator of aligned MIDI parts: rhythm, percussion, bass, harmony, melody and controller automation for further DAW work.

Algorithmic composition · Python / MIDI / Seeded generation / Music theory

The product problem

Create a controllable compositional starting point with recorded parameters, not a finished audio-track promise.

Stage

Generator source is present; byte-level reproducibility needs a pinned version and environment check.

Concept illustration: GENIUS-MIDI
A concept illustration, not a screenshot of a running product.

From parameters to DAW material

An illustrative intended journey, not a demonstration of the current live release.

  1. Choose genre, tempo, scale and length.
  2. Record the seed and generation settings.
  3. Export aligned MIDI parts and a manifest.
  4. Continue instrumentation, arrangement and sound work in a DAW.

System components

  • Seven MIDI parts: main/glitch drums, polymetric percussion, bass, chords, topline and dynamics.
  • Genre profiles, tempo, bar count, key and scale.
  • Number patterns and musical heuristics shape the structure.
  • A seed and manifest preserve each generation’s settings.

Architecture & ownership

A Python generator and MIDIUtil export MIDI events and a manifest. The DAW supplies instruments, synthesis, mixing and final audio.

Where does MIDI-event generation end and the composer’s artistic decision begin?

What to validate before use

Number theory is a composition principle, not a guarantee of artistic quality or listener effects. Third-party instruments are not bundled.

This describes the development scope. This page does not run the product, process payments or verify real-time availability.

Related products, distinct problems

Next step

Discuss a GENIUS-MIDI workflow

Describe the problem, current system, constraints and desired outcome. We agree scope and deliverables before cost and launch.

Discuss the task on Telegram ↗

The link opens a human contact; it does not send a message automatically.