Michalakos, NIME 2023 Music Proceedings · Performance paper · Read 2026-08-10
1. What they did
A two-player competitive game-piece. Two improvising musicians use their instruments as game controllers, duelling through randomised mini-challenges. Instrument signals are analysed live by machine learning in Max/MSP; parameters go over OSC to Unity, which drives the game and the image; DMX lighting maps game state so that the physical and digital spaces fuse. The performers face each other on either side of a split screen, the two halves at ninety degrees.
The core mechanic: a player has to win the game and make good music at the same time, and the friction between those is the performance. The game becomes the score of the improvisation.
2. Method
- No formal research method — this is a Music Proceedings paper, performance documentation rather than research.
- Structure: Program Notes → Project Description → Performance Notes → Equipment List → Media Links → References.
- Stack: instrument → Max/MSP (ML audio feature analysis) → OSC → Unity → screen + DMX.
- Mini-challenges are randomised across different musical affordances — pitch challenges favour the violinist, dynamics challenges favour the percussionist.
- Performers: Michalakos (drums/electronics) and Dimitris Papageorgiou (violin).
- Built on eight years of practice: Pathfinder (2016) → ICARUS (2019) → Death Ground (2019) → SPLT/SCRN (2023).
- Four references, all the author’s own prior work.
3. Findings and contribution
- Game as score. Game mechanics can replace notation as the structuring device for improvisation. The game supplies rules, randomness and visible feedback; the music emerges from playing it.
- Friction as feature. The tension between winning and sounding good is not a bug. The double constraint produces a drama conventional improvisation does not have.
- Transparency. A game engine makes visible what is normally opaque to an audience in live electronics — why did the player go silent, why is that sound being processed that way.
- Light as a second channel. In some challenges the lighting is the primary driver and the screen is secondary; performers get enough information from colour and intensity without looking.
- Sabotage. Players can interfere with each other — the drummer can bury the violinist’s pitch challenge under volume. Most improvisation is cooperative; this is competitive.
- Signal flow is instrument → Max → Unity: audio drives visuals. The opposite paradigm to controller → Unity → Max.
4. What it gives me
- A direct template for the ICLC submission. Program Notes → Project Description → Performance Notes → Equipment List is exactly what my live performance proposal needed.
- Academic identity, made explicit. This paper is what let me say it: I am a game designer making music, not a musician making games. The signal runs the other way, and the community is different — TEI/DIS/CHI PLAY rather than NIME/ICMC.
- “Game engine as instrument” is not “game controller as instrument.” The former is what I do — Unity’s 3D space, physics and collision as the instrument. The latter is everywhere in NIME. Mine is the smaller, newer, more defensible niche.
- Lighting and multimodality. The screen is not the only output. That widens the boundary of “screen-based installation” in the thesis to screen plus light plus sound plus vibration.
- Game as score is citable if the thesis argues that rules are the structural skeleton of an experience rather than a way to make it fun.
- NIME publishing culture: Music Proceedings are not research papers. No literature review, no user study, no theoretical contribution required — reviewed on performance novelty, technical reliability and relevance.
5. Questions it left me
- SPLT/SCRN is competitive, player against player. Mine is exploratory. Could I get gameful tension from person-against-environment or person-against-system instead of person-against-person?
- Sabotage is interesting inverted: what if two performers must cooperate to progress, and failure to cooperate collapses the music?
- Their performers are professional musicians. If the performers are ordinary visitors — which is my situation — does the game-piece still hold? Where would the improvisational skill come from?
- Is the lighter Music Proceedings format actually better for a first submission than a short paper?
- Michalakos spent eight years iterating one direction. What would my eight years look like?