How SOONOS Live Club Simulation Works
Live Club Simulation places your set inside a modelled room. You choose a venue, a moment of the night and a crowd size; the model reads the energy and structure of the mix and shows how simulated clubbers — placed at different distances from the booth — could respond over time. The clubbers are software agents, not people, and the result is a rehearsal tool, not a prediction of a real audience.
Published by SOONOS · Last updated 9 September 2026
Venue
Three room types are modelled: underground club, beach club and festival. Each one changes the geometry of the simulation — width and depth, how the audience spreads sideways versus in depth, and whether the crowd forms one block or several pockets. The same set therefore reads differently from one room to another.
Moment of the night
Warm-up, building, peak time and after. The moment sets what the modelled room is waiting for. When the simulation is driven by a real analysed mix, the moment is read from the position inside the set rather than chosen by hand.
Crowd profile and crowd size
Crowd size drives the density of the floor and the number of rendered agents, which are a sample of the room rather than one dot per person. Each agent carries a listening profile — early adopter, energy seeker, groove listener, melodic listener, casual clubber or genre expert — and reacts to the same moment differently.
Mix energy and structure
The simulation is fed by the energy curve and the structure of the set: where it builds, where it drops, where a breakdown lands and where attention is likely lost. On an analysed mix these come from the SOONOS analysis of the audio. On the demo set on the homepage, they come from a clearly labelled synthetic curve.
Virtual clubber profiles
Each agent has a position in the room, a distance from the booth, a zone (front, mid, back — left, centre, right) and simulated values for perception, energy response, engagement, attention and movement. Selecting a clubber opens those values. They are computed by the model; they are not observations of a human being.
Possible audience response patterns
- Room engagement rising or falling across the timeline.
- Peak response: how strongly the modelled room answers the biggest moments.
- Attention retention over long stretches without contrast.
- Energy match between what the set does and what the moment calls for.
- Strong and weak zones: where in the room the response holds and where it drops.
Ibiza sunset, underground club, festival peak time
A beach-club sunset room is wide and shallow: attention is spread, the audience is not waiting for maximum intensity, and long patient builds hold. An underground club is deep and dense, focused on the booth, and rewards groove and contrast over spectacle. A festival at peak time splits into pockets across a very large floor, where distant zones follow the biggest gestures far more than the fine detail. The same mix produces three different response shapes for that reason.
What is simulated
- Room geometry, density and audience placement.
- Per-agent perception, engagement, attention and movement over the timeline.
- Zone-level and room-level response derived from the mix energy and the chosen moment.
What is not measured
- No real crowd, venue, sound system or acoustic measurement is involved.
- No camera, sensor, ticketing or attendance data is used.
- Nothing here is validated against a recorded reaction from a real night.
- The agents are not people and carry no personal data.
Why this is not a prediction
A real dancefloor depends on the city, the line-up, the hour, the sound system, the weather and the mood of a few hundred strangers. None of that is knowable from an audio file. The simulation is a way to rehearse structural decisions — where to build, where to hold, where to release — before you play the room. Treat it as a rehearsal surface, never as a forecast.
Real analysed examples
Public reports whose measured energy and structure feed this kind of simulation: