A spatial audio workflow carries a listening idea from the first sketch into a real room.
That journey includes composition, previews, prototypes, system planning, content preparation, integration and on-site tuning. The tools may change along the way, but the central question should not: what should visitors hear, and how should that sound behave in this place?
A good workflow connects seven stages:
Previews help teams make decisions before the room is ready. They are valuable working models, but they are not proof of the final result. The installed space remains the reference.
Before choosing software or loudspeakers, we define what sound should do.
It may guide attention, connect rooms, give an object a voice, create a surrounding atmosphere or respond to movement. These are different tasks. Each leads to different decisions about content, timing, system design and testing.
This first step also identifies what must be protected. Clear speech may need distance from an interactive station. A subtle composition may need a quiet threshold. A responsive installation may need immediate feedback so visitors understand that their movement matters.
Early sketches make an invisible idea discussable. A spatial score can show when and where events occur. Zone plans can show where sound should remain local. Simple multichannel mock-ups can test movement and layering.
Binaural previews are useful when a team needs to review direction, proximity and general motion over headphones. Real-time engines help when sound reacts to position, gesture, data or changing visual scenes. VR and XR prototypes can place that behaviour inside an early model of the environment.
The aim is not to imitate the final room perfectly. It is to answer specific questions while change is still easy.
Choose the production chain
Once the key behaviours are clear, the team can choose how spatial information should travel through production.
A known permanent installation may use fixed multichannel files. A work that needs to adapt may keep sources as objects until rendering. Ambisonics may be useful for soundfield recording or variable playback. A headphone experience may be binaural from the start.
Renderer, playback system and content format should be chosen together. The workflow also has to connect with show control, network audio, sensors, projection and any other system that affects timing or interaction.
Use previews for the right decisions
Every preview leaves something out.
Headphones leave out the shared room. A stereo mock-up leaves out spatial depth. A simulation may simplify reflections, materials, noise and bodies. A real-time prototype can show how a system responds but not exactly how that response will feel in the final architecture.
This does not make previews unreliable. It makes their purpose specific. A useful review begins with one question: which decision can this model support?
Previews also improve communication. Architects, curators and clients cannot read a complex routing diagram as a listening experience. A short headphone sketch, annotated score or interactive mock-up can make movement and timing much easier to discuss.
When the installed setup becomes available, the spatial material must be mapped to it. This may mean decoding Ambisonics, rendering objects, preparing fixed loudspeaker feeds, configuring zones, routing network audio or connecting sensors to show control.
Translation is part of composition. A movement that feels clear in headphones may blur in a reflective room. A dense layer may mask speech. A sound may need a wider position or a slower transition to remain legible while visitors walk.
The right response is not to defend the preview. It is to adjust the score so that the intended behaviour survives the real conditions.
Tune on site
Only the room reveals the full relationship between loudspeakers, architecture, materials, neighbouring media and moving listeners.
We listen from different positions and walk the actual visitor route. We check thresholds, balance speech and music, test interaction, compare quiet and busy moments and observe how sound behaves when several exhibits run together.
On-site tuning includes level, delay, localisation, frequency balance, dynamics, coverage and spill. It also includes duration: a sound that is exciting for two minutes may become tiring over a long visit or a full day of operation.
For a permanent installation, the workflow continues after the mix is approved. File naming, routing, calibration, startup behaviour, monitoring, presets and failure handling all need to be documented.
Good handover material lets the operating team understand what normal behaviour looks like, identify a problem and restore the intended state without rebuilding the system.
A synchronised narrative unfolds across several rooms and listening positions. The workflow had to connect one timeline with changing distance, architectural spill and a moving audience.
Interaction and a 24-loudspeaker system form one responsive installation. Prototyping, mapping and on-site testing were inseparable from the composition.
A four-hour multichannel composition was implemented in an underwater tunnel exhibition. The long duration and the acoustic conditions of the site were as important as the original musical material.
We build the workflow around the critical listening decisions of each project. That may involve binaural previews, multichannel sketches, real-time prototypes, spatial scores, interaction tests or on-site mock-ups.
Each tool has a defined job. The final playback condition stays visible throughout the process, so composition, system design and integration do not drift apart.
The wider practice that defines how sound is placed and shaped around listeners.
How spatial and temporal intentions are mapped before the production chain is fixed.
Spatial Audio Formats & Methods
How channel-based, object-based, Ambisonics and binaural approaches carry spatial information through production.
Spatial Audio Systems & Renderers
How the playback environment and renderer turn the workflow into an audible result.
How real-time behaviour, sensors and visitor input become part of the production process.
Can a binaural preview replace a loudspeaker test?
No. It can communicate direction and movement over headphones, but it cannot reproduce the shared acoustics, coverage and spill of the final room.
When are real-time engines useful?
When sound responds to movement, position, interaction, data or changing media. They let the team test behaviour rather than only a fixed timeline.
Why is on-site tuning necessary?
Because materials, reflections, loudspeaker placement, visitor routes and neighbouring media change how sound is perceived.
When should workflow planning begin?
Early, together with architecture, exhibition design, media planning and interaction. That is when the technical chain can still follow the listening concept.
We develop workflows that carry sound from concept and studio preview into the real conditions of an exhibition, installation or immersive environment.
