Belfast: a first collaboration session

Performance Fingerprints started with a simple question: what would it mean to treat a musical performance as something you can draw with?

Not in the “music visualiser” sense — not graphics pasted on top of sound — but something closer to a drawing practice that has to listen. We’re interested in marks that can register the specific character of a performance: timing, intensity, articulation, rhythm, texture.

Some of the earliest experiments in music responsiveness were encouraged by flautist Carla Rees. Two of her works — Nightsong and Tregenna — became the first real subjects we returned to as the system took shape.

Those early runs were messy (as they should be), but they were enough to show that the approach could hold up outside the studio. A gilded work on canvas made from a Nightsong drawing was later bought by a London collector, which gave us a quiet but important nudge: this wasn’t only an interesting process — it could also make a finished object that someone wanted to live with.

At the centre of the work is a live computational drawing system that responds in real time to acoustic instruments. It runs continuously, and we can take snapshots directly from that live process — freezing particular moments that can then be re-made as physical work.

From live drawing to physical works

Those snapshots are still born out of performance, but once they’re held still they give us something we can slow down, revisit, and translate into different materials. So far that’s meant a small family of formats:

• large colour triptychs made from photographic prints mounted onto panels,
• sequences of ink drawings on paper (sometimes cut and collaged into a left-to-right timeline),
• drypoint print series made from multiple states of the same plate, where some marks fade through repetition and others return when the plate is re-etched.

Two early translations with silver leaf: a print and a detail.

Silver leaf has become a recurring material. Part of it is straightforward: it catches light, changes as you move, and refuses to behave like ink or pigment. But it’s also literal reflection, which ties the work back to the viewer and the space around it — something we recognise from live music.

All of this was useful groundwork, but it still left the relationship mostly one-way: music drives the drawing.

The more interesting direction is a loop: musicians responding to the drawing in real time, and the drawing responding back.

Belfast, February 2025

In February 2025 we held a first proper collaboration session in Belfast with violinist and performer Ioana Petcu-Colan and trombonist and collaborator Ross Lyness.

We worked at Digital Arts Studios (DAS). The setup was intentionally simple: a studio room, a large screen for the drawing, and time to try things without needing to “produce” anything finished.

Ioana and Ross improvised while watching the drawing develop. We treated the whole thing as a conversation: play a little, watch a little, stop and talk, then try again.

A few things became clear quickly.

1) We needed a shared language

The system could respond, but it wasn’t always obvious what it was responding to.

That’s not just a technical problem — it’s a collaboration problem. If you’re going to improvise with something, you need some confidence about what your actions mean inside the other system.

So we talked about how to build a shared language between musicians and drawing: a small set of musical actions that can be recognised reliably (density, register, pulse, intensity, articulation), mapped to changes in drawing that are readable in the moment.

2) The loop was exciting — and chaotic

When the mapping isn’t clear enough, the loop can get busy fast: the drawing reacts strongly to moment-by-moment change, the performers react to the drawing, and the system reacts again.

In Belfast we felt that directly. The energy was good, but sustaining it over time was difficult because we needed a better balance between responsiveness and stability. This isn’t a request for the drawing to be calm; it’s more like musical dynamics — intensity is useful, but so is space, contrast, and the ability to hold a texture.

3) The drawing needed a bigger vocabulary

Professional musicians have a huge range of techniques available instantly, as part of expression. If the drawing is going to feel like a true partner, it needs a broader mark-making vocabulary to match that range.

That isn’t about visual novelty. It’s about giving the system more ways to behave so that the drawing can shift character when the music shifts character.

Two early “fingerprints” from the Belfast session. Left: a response to melodic violin playing. Right: a response to extended techniques on trombone. Even with a limited visual language at this stage, the contrast suggested we could make distinguishable traces from different performances.

What we brought home

We left Belfast with hours of audio recordings and filmed material.

Those recordings matter in two ways: they’re documentation of the collaboration itself, and they’re raw material for development — especially around bringing video responsiveness back into the system, so the drawing can pay attention not only to sound but also to aspects of performance gesture and movement.

Next steps

The immediate work is a mix of technical and artistic tasks, but they’re tightly linked:

• make the audio-to-drawing mapping more legible in the moment,
• design drawing behaviours that can be responsive without becoming constantly agitated,
• expand the mark-making toolbox so the system can shift between different visual “modes”.

The longer-term goal is simple to say and hard to do well: build a performance situation where the drawing isn’t an output, and the music isn’t an input, but both are participants.

Belfast was the first day that felt like we were genuinely in that territory.