RESEARCH — ACOUSTICS ALGORITHM
APC
Audience Parallax Compensation
APC aligns the system to the audience, not to the speakers. It computes the delay, level and tone each loudspeaker needs so the sound arrives correct where people actually are — turning a rig split between ground fills and flown arrays into one coherent image for the crowd.
THE PROBLEM
The sweet spot is not where the audience is
Spatial systems are usually aligned at the array: delays and levels set so the speakers agree with each other. But the audience is somewhere else — spread across a floor, under flown boxes, in front of ground fills — and from those seats the geometry is different. A source that lines up perfectly at the rig arrives smeared and mislocated out in the crowd.
The problem is parallax: a low front-fill and a high delay hang sit at different angles and distances to every listener, so the timing that makes them coherent for the system pulls them apart for the audience. Align to the boxes and you optimise the empty air above the crowd.
It gets worse the larger the audience. A single reference alignment can only be right in one place, so the front, the middle and the back of the crowd each hear a slightly different, slightly wrong version of the show. On a big floor, or a system split across several stages of loudspeakers, that gap between what the rig was tuned for and what people receive is exactly where clarity and localization are lost.
HOW IT WORKS
How it works
Aligned to the listener
Delay, level and tone are computed toward the audience — the first listener in each speaker's line of fire — not toward the other speakers.
Ground and air, made one
Front fills and flown arrays are reconciled for the crowd, so a split system reads as a single coherent image.
Zone-aware
The audience is described as real areas, and each speaker group aligns to the part of the crowd it actually serves.
UNDER THE HOOD
Going deeper
APC re-references the whole alignment to the audience. Each loudspeaker's aim is intersected with the audience area to find the listener it actually serves first, and the delay, level and high-frequency balance are computed toward that point rather than toward the rest of the rig. The speakers closest to the crowd are held back in time and level so that everything arrives together, and correctly toned, where people are standing — the far boxes no longer race the near ones to a listener neither of them was aimed at.
The audience itself is part of the model. Listening areas are described as oriented zones with their own density and shape, and each speaker group is tied to the zones it covers, so a system with distinct floor and balcony areas aligns each of them on its own terms instead of averaging them into a compromise. High-frequency air loss over distance is compensated per speaker, so the front and the back of the crowd share the same tonal balance rather than the back going dull.
It is a deployment tool as much as a rendering one: the same field the renderer produces is delivered to the people in the room, correcting the parallax between a system's geometry and its audience — the gap that usually eats hours of manual time-alignment at every venue.
None of this asks the operator to model the crowd by hand. Draw the audience areas, tie each group of loudspeakers to the zones it covers, and the alignment falls out of the geometry — recomputed for every new venue instead of dialled in delay by delay. What used to be a slow, partial soundcheck ritual becomes a property of the design that travels with it.
AT A GLANCE
At a glance
| Input | Speaker layout + audience areas |
| Output | Per-speaker delay, level and HF trim |
| Latency | Zero added — alignment layer |
| Implementation | C/C++ · integrates with the renderer |
| Availability | Ripl beta · OEM licensing |
PARAMETERS
Hands on the algorithm
| Audience zones | Position, size, orientation, density |
| Group assignment | Speaker groups ↔ the zones they serve |
| Delay alignment | Per-speaker, to the served listener |
| Level alignment | Per-speaker, to the served listener |
| HF compensation | Per-speaker air-loss trim |
POSITIONING
Compared to the alternatives
vs aligning at the array
System-referenced alignment optimises the sweet spot; APC optimises the seats, where the audience actually is.
vs manual time-alignment
Hand-tuning delays per venue is slow and partial; APC derives them from the geometry of the room and the crowd.
vs a single system delay
One global alignment can only be right in one place; APC resolves timing per speaker and per zone, so the whole audience — not one row — hears it right.
APPLICATIONS
Where it fits
Touring & festivals
Reconcile ground fills, out-hangs and main arrays for the field in front of them, at every new site, without a night of manual alignment.
Immersive venues
Keep a spatial image coherent across a whole audience area rather than a single reference seat, so every section hears the intended placement.
Worship & theatre
Distributed and delayed loudspeakers align to the congregation or the house, keeping speech localised to the stage and intelligible throughout.
Installed sound
Permanent systems aligned to where people stand and move, holding tone and timing across the covered area.
INTEGRATION
Built to live inside your product
| Delivery | C · C++ · MATLAB · .dsp — full source code |
| Platforms | macOS · Windows · Linux |
| DSP platforms | Flow DSP · Audio Weaver — solutions in preparation |
| Documentation | Whitepaper — every algorithm explained, in the clear |
OEM LICENSING
- ■ One-time payment per brand
- ■ Full source code — C, C++, MATLAB, .dsp
- ■ Whitepaper — all algorithms explained
- ■ Integration support included
- ■ Free updates
- ■ Volume discounts on multiple licenses
SHIPS IN