Multi-Track Music Machine (MMM)

We propose the Multi-Track Music Machine (MMM), a generative system based on the Transformer architecture developed by Jeff Enns and Philippe Pasquier, and designed to support co-creative music composition workflows. The system generates multi-track music by providing users with a fine degree of control of iterative resampling driven by machine learning.

MMM supports the infilling of musical material on the track and bar level, and can condition generation on particular attributes including: instrument type, note density, polyphony level, and note duration. In order to integrate these features, we employ a different type of representation for musical material, creating a time-ordered sequence of musical events for each track and concatenating several tracks into a single sequence, rather than using a single time-ordered sequence where the musical events corresponding to different tracks are interleaved. In what follows, we outline the real world applications ofMMM , including the production of musical albums, and collaborations with industry partners that explore integrating MMM into real-world products.

Model

Check out the MMM model page for more on the scientific research and publications on this project.

Software Integrations

Cubase Plugin: A prototype in Cubase that integrates MMM into a popular DAW workflow

Calliope: Calliope is an interactive environment using MMM for symbolic music generation in computer-assisted composition.

OP-Z: A synth by Teenage Engineering

Ableton Live Plugin: Multi-Track MIDI Music Generation Machine for Ableton Live

PreGLAM-MMM: Using MMM in video games

Music Produced

See Music with MMM for more

Phil Tremble Album [produced using Ableton Live Plugin
(https://philtremble.bandcamp.com/album/blockhouse-dy20-electro)

Ableton Live Templates [produced using Ableton Live Plugin]
Chris Anderson produced templates for Ableton Live using the Ableton Live Plugin version of MMM:

Galactic Defense [produced using Calliope]
Cale Plut created an interactive score for the game 'Galactic Defense' using MMM in Calliope. The stems for the interactive score are available here. The game is available for download. An article has been published which explains the development of the generative score.

Steinberg Cubase Plugin User-Study
Renaud Bougueng Tchemeube conducted a user study to evaluate an MMM plugin in Cubase. The following are examples of the music composed by participants.

Papers and Posters

Ens, Jeff, and Philippe Pasquier. "MMM: Exploring Conditional Multi-Track Music Generation with the Transformer." arXiv preprint arXiv:2008.06048 (2020).

Plut, C., Pasquier, P., Ens, J., & Tchemeube, R. (2022, September). PreGLAM-MMM: Application and evaluation of affective adaptive generative music in video games. In Proceedings of the 17th International Conference on the Foundations of Digital Games (pp. 1-11).

Tchemeube, Renaud Bougueng and Ens, Jeffrey and Plut, Cale and Pasquier, Philippe and Safi, Maryam and Grabit, Yvan and Rolland, Jean-Baptiste. Evaluating Human-AI Interaction via Usability, User Experience and Acceptance Measures for MMM-C: A Creative AI System for Music Composition. Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, 5769-5778, 2023

Forthcoming: Evaluating Human-AI Interaction in Computer-Assisted Music Composition with MMM-Cubase v2



Previous
Previous

DJ-MVP 2022

Next
Next

MMM-synthrider