How Echo actually calculates this.

Echo and Echo Studio work out how long sound lingers in your room using two well-established statistical methods — Sabine and Eyring — evaluated separately for each octave band from 125 Hz up to 8 kHz. They're both built from the same two ingredients: how big the room is, and how much of its surface actually soaks up sound rather than bouncing it back.

The two methods normally agree closely. Sabine is the simpler, older formula, and it starts to overestimate reverberation once a room is quite absorbent already; Eyring corrects for that. Echo reports whichever is the better fit for your room rather than always defaulting to one — and if the two disagree by more than about 15%, that's flagged on your report directly, because it usually means the room is absorbent enough overall that the choice of formula starts to matter. That's a different thing from the absorption being unevenly spread across particular surfaces (a lot of soft furniture on one wall, bare everything else) — this flag can't tell you that on its own; it's a prompt to treat the single figure as a range, not a diagnosis of where the absorption sits.

Alongside reverberation, Echo works out the Speech Transmission Index (STI) — a 0–1 score for how intelligible speech will actually be, combining reverberation and background noise rather than looking at either alone. The method follows IEC 60268-16's modulation-transfer approach: each octave band's reverberation time and signal-to-noise ratio feed into a per-band modulation index, weighted and combined the way that standard sets out.

C50 — the ratio of early sound energy (the first 50 milliseconds, which the ear still uses to understand a word) to everything that arrives after — is reported too, calculated directly from the reverberation figure. It's a statistical estimate, shown as an indicator rather than a measurement.

Every calculation needs to know how much sound each surface in your room actually absorbs. Echo uses its own compiled reference set of octave-band absorption coefficients for common finishes — carpet, plaster, glazing, acoustic panel classes and so on — drawn from published acoustics references and manufacturer test data (ISO 354 chamber tests), not invented or guessed. Where you've told Echo the specific product you plan to use and the manufacturer's tested absorption data is available, Echo uses that figure instead of the representative one.

This is the part of any acoustic prediction that carries the most real-world uncertainty — not the arithmetic, which is checked against the published formulas it implements, but whether the assumed absorption matches your room's actual surfaces.

Sabine and Eyring both assume a reasonably even, diffuse sound field: they treat a room as one averaged space, not as a specific point where someone's standing making noise and another specific point where someone's trying to listen to them. That's a very good assumption for most ordinary rooms, and a weaker one for a room with very unevenly distributed absorption or an unusual shape — which is exactly the gap Echo Pro's geometric ray-tracing exists to close (see below).

One honest consequence, worth stating plainly rather than leaving you to discover it: today, Echo and Echo Studio's treatment recommendations are a quantity of absorption, worked out purely from the room's size and how far its reverberation misses the target — not a position tied to where your sound source actually is. Moving the sound-source marker in the 3D view changes what you see on screen; it doesn't currently change the calculation. Position matters as well as quantity, and building that in properly — panel placement that actually follows the source, not just the room — is real work we're doing next, tier by tier.

Echo Pro doesn't use Sabine or Eyring at all. It builds your room as an actual three-dimensional shape — including a genuinely pitched roof, not an approximation of one — and fires a large number of individual sound rays from your source position, tracking every bounce off every real surface: which wall, which patch of ceiling, at what angle, losing energy according to that specific surface's own absorption every time it reflects.

That's what lets Pro report seat-by-seat figures rather than one number for the whole room — the back-left corner behind a pillar and the front row by a glass wall can have genuinely different speech clarity, and a statistical model has no way to tell you that; a geometric one does, because it's actually tracing the path sound takes to get there.

For every seat position, Pro's engine already identifies which specific reflection — off which specific wall or ceiling area — is doing the most damage to that seat's clarity. Today that calculation only drives the ray-paths drawn in the 3D view; surfacing it as an actual "treat this surface" recommendation, for every seat, is the next piece of work landing on Echo Pro — because the data to do it properly already exists inside the engine, it just isn't shown to you as guidance yet.