SONIC ARCHITECTS

SONIC ARCHITECTS

Sound can be composed as a finished piece. It can also be designed as a system that listens, changes and responds.

Code lets us work in that second mode. Instead of fixing every event on a timeline, we define relationships: how a sound reacts to movement, how long a state can last, what may change and what should remain recognisable.

The result is not less composed. It is composition extended into behaviour.


BUILDING SOUND WITH RULES

Digital tools give composers direct access to pitch, level, timbre, timing and spatial position. Code adds another layer: conditions.

A sound may begin when someone enters a zone. Its colour may change with distance. Several musical layers may combine differently depending on how many visitors are present. A system may remember an earlier interaction and let that memory return later.

These decisions create a framework rather than one fixed playback sequence. We compose the material, but also the rules that determine how the material can meet the world.

Visual programming environments such as Max/MSP are useful here because musical, interactive and technical relationships can be developed together. Sensors, audio processing, spatial playback and media control become parts of one instrument built for the project.

Working with code changes the scale of the task.

The composer still shapes melody, rhythm, harmony, texture and form. At the same time, someone has to decide how the system behaves: which inputs matter, how quickly it responds, what happens when no one interacts and how variation can remain coherent over many hours.

This is why we think of ourselves as sonic architects. The term does not replace composition or sound design. It describes the work of giving sound a structure in which it can develop.

The structure must be understandable to listeners even when they never see the code. If a response feels arbitrary, the interaction loses meaning. If every outcome is predictable, the system soon feels lifeless. The craft lies between those extremes.


REAL-TIME SOUND IN PRACTICE

Our early experiments often focused on one relationship at a time. Sensors allowed sound to respond to movement. Non-linear and vertically layered compositions allowed intensity to change without abandoning the musical identity of a piece.

MOMENTUM, created with SCHNELLE BUNTE BILDER, connected custom software with a music-driven installation. I AM NOT A ROBOT explored participation inside a three-dimensional loudspeaker environment. In both cases, visitors were not simply an audience placed in front of a finished work. Their presence affected what the work became.

These projects also taught us that interaction needs restraint. A system should not react just because it can. The response must support the larger spatial and emotional arc.

LIGHT CLOUD: a system with states

LIGHT CLOUD, developed with TAMSCHICK MEDIA+SPACE, brought many of these ideas together in a large sound and light installation.

The work behaves like a changing organism. It moves through states such as sleeping, dreaming, remembering, feeling and thinking. Each state has its own audiovisual character and its own level of responsiveness.

Visitors may observe one moment and influence the next. Their movement becomes one input among several, so the installation can react without turning into a simple one-to-one trigger machine.

More than eighty soundscapes provide the musical material. Code selects, combines and transforms them according to the current state and the activity in the room. The work can therefore remain recognisable while never unfolding in exactly the same way.

In LIGHT CLOUD, sound is not added after the visual design. The sound engine also drives aspects of the light.

Rhythm, intensity and sonic events can shape visual behaviour in real time. This creates a shared composition rather than two parallel tracks. Visitors experience sound, light and movement as one responsive field.

The interesting part is not the technical connection itself. It is the change in authorship inside the installation: musical decisions influence the visual scene, visitor movement influences the system, and the system chooses how those inputs become a moment.


CODE AS A CREATIVE COLLABORATOR

CHASING WATERFALLS explored another kind of system: artificial intelligence as both subject and participant in a music-theatre production.

The AI generated text and musical phrases that appeared alongside human performers. Its role was not to provide a futuristic sound effect. The production used the uncertainty of machine-generated material to examine authorship, control and the uneasy relationship between human intention and computational output.

Real-time generation introduced results that could not be planned note by note. That unpredictability was useful because the work was about the limits as much as the possibilities of collaboration with a machine.

Code did not replace artistic decisions. It changed where those decisions were made: in the training material, the rules, the selection, the response to unexpected output and the way human performers framed it on stage.


DATA AS MUSICAL MATERIAL

Sonification opens a related field. Here, data is mapped to audible parameters such as pitch, rhythm, density, timbre or spatial position.

The mapping can reveal a pattern, but it can also become part of a composition. Data does not arrive with one correct sound attached to it. The designer decides which relationships should remain accurate, which can be interpreted and what listeners need in order to understand the result.

This again turns code into an instrument for relationships. It connects an external process—climate, movement, biology or social activity—to a sonic system that can make change audible over time.


COMPOSING BEHAVIOUR

Code expands sound design from objects to behaviours.

It allows installations to respond, evolve and connect with light, data, movement and architecture. But the technology only becomes meaningful when the listening experience remains clear.

The central work is still composition: choosing material, shaping time, creating contrast and deciding what a moment should feel like. Code gives those decisions a space in which they can continue to act after the composer has left the room.


RELATED PRACTICES AND TECHNICAL NOTES

Generative Sound

How rules and systems create evolving sound environments that remain coherent over long durations.

Interactive Sound

How sound responds to movement, touch, presence or other live inputs.

Data Sonification

How data can be mapped to pitch, rhythm, timbre or spatial behaviour without confusing interpretation with fact.

Spatial Audio Production Workflows

How real-time engines, prototypes and installed systems are connected and tested in production.


BUILD A SOUND SYSTEM THAT CAN RESPOND

We develop generative, interactive and spatial sound systems in which composition, code and real-time behaviour work as one. Explore projects where sound listens, changes and responds.