A nested elastic cube in a blue-violet dark room, with four speakers surrounding a participant at its centre.

Sonic Elasticity

An experiment in sound and spatial perception

2025–26 / Bartlett School design project · Collaborative installation

Sonic Elasticity

A perceptual environment constructed through sound and spatial transformation

Sonic Elasticity is an interactive installation exploring how controlled sensory conditions construct and reshape spatial perception. Drawing on the geometric logic of the tesseract, it uses time as a dimension of perceptual change. An elastic cube expands and contracts alongside a four-channel soundscape, shifting judgements of distance, scale, and boundaries through synchronisation and deviation.

My contribution: Installation design and fabrication within a collaborative project
Project stage: Research, model iterations, physical installation, and audiovisual programming

Prediction and perception

When our predictions of space fail, how is perception reconstructed?

The brain continuously predicts sensory input rather than passively receiving information. Repeated rhythms, scales, and spatial relationships establish stable expectations. Mismatches between prediction and incoming stimuli prompt perception to reorganise.

Can a room feel larger than its physical dimensions? Can time seem to accelerate or slow down? These questions position illusion as a means of revealing and intervening in perceptual reconstruction through subtle but perceptible changes.

A complete concept board of predictive perception, space, and time — From sensory reception to active prediction: questions of spatial and temporal perception.

Repetition, expectation, and deviation

Repetition establishes regularity and a pattern that participants use to anticipate what happens next. Disruption makes previously reliable cues inconsistent, requiring a new interpretation of the environment.

A connected flow from repetition through prediction to disruption and illusion — Repetition establishes a pattern; deviation prompts reassessment.

Making perceptual change spatial

Visual and auditory cues become a controllable sequence of change. Transforming frames, light points, and boundaries allow participants to encounter the formation and revision of their spatial expectations.

Four authored concept views of changing frames, light points, and space — Four spatial transformation studies from the concept stage.

Research and cognitive framework

ITPRA: the temporal structure of expectation

Huron’s Sweet Anticipation introduces the ITPRA model: Imagination, Tension, Prediction, Reaction, and Appraisal. It informs how repetition establishes expectations and how deviation produces immediate responses and subsequent reassessment.

Space and time provide direct sensory cues. The project translates this cognitive framework into changes in rhythm, distance, and scale, introducing controlled deviation within a stable pattern.

The complete ITPRA diagram showing responses before and after an event — Huron, Sweet Anticipation, 2006.

How sound shapes time and space

Three research references for rhythm, soundscape, and spatial sound — From rhythmic repetition to continuous soundscapes and spatial scale conveyed by echoes.

Rhythm and unexpected events

Repeated rhythms establish temporal expectations. Unexpected events become salient and may alter judgements of duration.

Soundscape

R. Murray Schafer’s soundscape concept informs a continuous environment of layered sounds in which listeners can establish reliable expectations.

Echoes and spatial scale

Echo timing, intensity, and decay provide cues to distance and spatial extent. Altered delay and reverberation suggest expansion or compression.

From fixed illusions to dynamic space

Ames room

From a fixed viewpoint, irregular geometry appears regular, revealing the role of visual references in judgements of scale.

Spazio Elastico

Gianni Colombo’s elastic lines, dark room, and movement destabilise boundaries and turn viewing into a bodily, dynamic experience.

Tesseract

The three-dimensional projection of four-dimensional geometry reconfigures boundaries and volumes, informing relationships between inner and outer frames.

Translating cognition into space

Active prediction, statistical regularities learned through repetition, and reorganisation after violated expectations become a three-stage experience: establish a predictable audiovisual relationship, gradually separate sound from space, and invite reassessment through new cues.

Establish a pattern

Repeated rhythms and scale relationships establish stable cues.

Form expectations

Participants learn to anticipate subsequent changes in sound and space.

Controlled deviation

Alter their correspondence so existing judgements require revision.

Perceptual experiments and physical model

The sound experiment used two stone-impact tracks. One accelerated before becoming irregular; the other added further sound elements. Participants estimated elapsed time, producing varied subjective durations that informed subsequent rhythm studies.

Spatial trials placed fluorescent frames in a dark room and combined fixed observation, bodily movement, and manual adjustment. Model tests explored how elastic structure, viewing position, and lighting affect the visibility of changing boundaries.

A complete twelve-image collage of listening trials and perceived-duration notes — Temporal experiment: altered impact rhythms and participants’ duration estimates.

A complete four-image collage of fluorescent-frame spatial experiments — Spatial experiment: observing changing elastic boundaries in a dark room.

1:5 physical model

A scale model made with wood, elastic cord, fluorescent paint, a UV lamp, and metal nails tested connections, tension, and transformation. It translated geometric proposals into an operable physical structure.

A high-resolution photograph of a participant operating the fluorescent elastic scale model — Testing elastic boundaries and structural transformation with a physical model.

Four design iterations

From a multisensory maze to a controlled relationship between space and sound.

A complete comparison of four proposals, from dense grids to a nested cube — Overview of the four iterations.

Iteration 1 · Visual dominance

The first proposal was a maze of fluorescent grids combining moiré patterns, grating imagery, 3D mapping, material weaving, knots, and sound.

Competing stimuli lacked a clear hierarchy. Participants could not establish a stable pattern or readily interpret the source of change. The next iteration reduced visual competition and sought clearer relationships between cues.

The complete maze proposal with weaving, mapping, and optical-effect studies — Connections between layout, form, materials, and visual effects.

Iteration 2 · Visual–auditory balance

The space became a linear corridor with a central “loom of time”, where stone impacts shifted from regular to irregular rhythms. Textured knots, vibrating lines, and sounding lines supported the experience, with sound becoming the primary cue.

Sound still lacked a continuous relationship with the overall structure. The link between rhythmic and spatial changes remained unclear, motivating an integrated sound field within one space.

The complete corridor and time-loom analysis with material and sound studies — From a dense maze to a linear space and rhythm experiment.

Iteration 3 · Auditory integration

The proposal became a single cube whose scale changed through cord angle and tension. Four corner sound fields shared an underlying rhythm, while visually similar floor zones had different textures to create sensory mismatches.

The experience depended on movement between zones, dividing changes into local events. A continuous formation and disruption of expectation required spatial and auditory transformations that could unfold over time.

The complete single-cube proposal with floor zones, sound fields, and scale changes — A single space connects visual scale with four corner sound fields.

An early full-scale fluorescent elastic frame with a participant — An early frame setup in G16.

Four states of an elastic frame at different heights and angles — Model studies of changing height and angle.

Iteration 4 · Pattern and disruption

The final proposal paired a fixed outer frame with a transformable inner frame, with the participant remaining at the centre. Sound initially matched spatial change and gradually diverged, allowing one position to experience different visual scales and acoustic distances.

Redundant floor textures and complex visual effects were removed. A central white sphere provided a reference, while a shared temporal sequence coordinated visual and auditory change to establish and then disrupt expectations.

The complete analysis of fixed and elastic frames, a central reference, and sound relationships — Final proposal: changing scale within a fixed reference.

A complete diagram of retained, removed, and added elements across four iterations — Reducing sensory competition to control the relationship between space and sound.

Interaction between space and sound

Within the stable outer frame, the elastic inner cube changes size. Rhythm, delay, reverberation, and spatial distribution alter acoustic scale. The project first establishes synchronisation and then separates physical and acoustic space.

The complete three-track timeline of rhythm, physical space, and acoustic space — Interaction timeline: synchronisation establishes regularity; deviation disrupts it.

From four sound zones to a continuous field

Early sound studies included a mechanical clock, water clock, temple bell, and bell tower. A shared underlying rhythm provided continuity, while varying echoes and reverberation suggested spaces from compact to expansive.

A complete analysis of four sound zones, waveforms, and spatial conditions — Four-zone sound study from iteration 3.

Correspondence and mismatch

The comparison shows congruent and incongruent physical and acoustic scales: a small visual space can carry an expansive bell sound, while a larger space can contain a compact sound. The final version makes these relationships unfold continuously rather than requiring movement between zones.

Eight interaction views comparing matching and mismatched physical and acoustic scales — Scale comparisons inform a continuously changing interaction.

Structure and fabrication

From a test frame to a fixed structure

The first full-scale test used tripods, elastic cord, and four speakers. Fluorescent paint improved visibility but altered the material appearance, so the later setup used white elastic cord under controlled dark-room lighting.

The first full-scale test with tripods, four speakers, and elastic cord — Testing scale and sound before refining the fixed outer frame.

Modules and connections

A fixed timber frame provides a stable reference. Corner connections and winding mechanisms transform the inner elastic cube. The design combines modular assembly, timber joints, cord nodes, brackets, and motor-driven rollers.

The complete component board with plans, elevations, modular joints, cord nodes, and rollers — Component relationships and joint details.

Fabrication process

Fabrication progressed from frame prototypes and joint tests through timber cutting, tab-and-slot connections, elastic cord installation, roller production, drilling, and assembly. The nine-step process board connects structural drawings with workshop photographs.

A complete construction board with structural drawings and nine fabrication photographs — Fabrication of the frame, joints, rollers, and assembly.

Completed physical framework

The completed outer structure forms a stable boundary in the black-box environment. Elastic lines, corner joints, and winding paths constitute a system for continuous, smooth changes in scale.

The completed timber outer frame and elastic inner cube — The overall relationship between outer and inner frames.

A high-resolution close-up of a corner cord connection — Corner connection.

The elastic inner cube and cord winding paths — Elastic frame and winding paths.

Transformation system and control

Letting sound drive spatial transformation

Max/MSP sends transformation commands to Arduino, which controls clockwise and counterclockwise stepper-motor rotation. Winding and unwinding cord at four corner rollers expands or contracts the inner cube.

NEMA 17 stepper motors, DRV8825 drivers, flange couplings, and 3D-printed rollers form the mechanism. Serial commands connect motor actions with the audio program so audiovisual relationships can be adjusted within a shared timeline.

The complete control flow linking Max, Arduino, motor drivers, couplings, and rollers — The control chain from sound programming to physical motion.

A complete board of Arduino code, motor tests, and roller production — Control code, hardware connections, and workshop testing.

Sound design and audiovisual integration

From regular to irregular rhythm

The final sound design uses stone impacts. The complete audio sequence lasts 300 seconds, with regular rhythm for the first 45 seconds before becoming irregular. Clear attacks support added echoes and comparisons of reflection and decay across simulated spaces.

Compact acoustic spaces use more direct, dry sound, while expansive spaces add echo layers and longer decay. Rhythm affects temporal expectation and the sound field affects spatial judgement, with correspondence and deviation coordinated within one sequence.

The complete 300-second waveform, regular-impact detail, and reflection analysis — The first 45 seconds establish rhythmic expectations before irregular changes.

Max/MSP version 1

Version 1 imported four sounds and used Spat5 to position virtual sources before speaker output. Source positions were adjusted manually, and spatial cues relied mainly on volume, without integrated reverberation, automated paths, or physical actuation.

The complete first Max/MSP patch with two source-position interfaces — A spatial-audio prototype with four sources.

From audio prototype to structural integration

The upgraded program uses a single source, adds echoes and reverberation, defines a default source path, and connects Arduino through serial communication. One program coordinates source movement, acoustic scale, and elastic-frame transformation.

A complete flow comparison of original and upgraded programs — Adding echoes, a default path, and Arduino integration.

Max/MSP version 2

Version 2 adjusts sound according to virtual source distance: within four metres it uses more direct sound and higher volume; beyond that threshold it adds echoes and reduces volume. This is a distance in the programmed acoustic environment, not the dimensions of the physical frame.

Yellow annotations identify added functions and blue annotations indicate the original logic. The Arduino connection allows acoustic and physical scale to align or gradually diverge as designed.

The complete second Max/MSP patch annotated with added and original functions — Integrating a single source, spatial positioning, echoes, and the physical structure.

Three scale states

Small, medium, and large states connect Spat5 source positions with the physical frame. Audiovisual agreement establishes expectations, while gradual disagreement invites spatial judgements to reorganise around new cues.