The image-source method, and where it breaks down

August 3, 2026

The image-source method is the cleanest way to compute early reflections in a room with flat, parallel surfaces: mirror the source across each wall, and every mirror image becomes a phantom source whose straight-line distance to the listener gives you a real reflection’s delay and, with the wall’s absorption folded in, its level.

It is exact for a rectangular room with hard, flat surfaces, which is also exactly the kind of room that sounds harsh in real life. The moment a room has a curved wall, a coffered ceiling, or a diffusing surface, the method’s core assumption — that every reflection is a perfect mirror bounce — stops holding, and you need either a much more expensive ray-tracing model or a statistical, diffuse-field model for the later part of the response.

Our homepage simulator uses image-source for exactly the region it is honest about: the early, specular part of the reflection pattern in a simple rectangular room. Past that window it switches to a statistical decay matched to the room’s own Sabine reverberation time, which is the same split full auralization software makes for the same reason.

Bring us in before the drywall does.

The earlier acoustics enters a project, the less it costs to get right. Tell us what you are building.