Illustrative worked example. This is a realistic brief run through Echo using the same method a real report uses — not an account of a named past client. See every case study for more, including which of these are real, permissioned projects as they're added.
The brief
An independent small-plates restaurant had a review pattern the owner could set a clock by: five stars for the food, and a noise complaint in nearly every review mentioning a table conversation. The open kitchen pass — the centrepiece of the room — was the one thing not up for discussion.
Dining room with open kitchen pass
16 × 8 × 3.8 m, about 486 m³ — polished concrete floor, painted brick walls, exposed steel-truss ceiling, full-height glazing to the street.
Concrete, brick, steel and floor-to-ceiling glass left almost nothing to absorb the room's own noise, putting it at roughly 4.3 seconds — genuinely like being inside a drum — and the reason a full room made conversation difficult almost anywhere you sat, not just by the second course.
The standard, and the target
There's no statutory target for a restaurant, so the brief was guest comfort: not a library, but a room where two people across a table can hear each other without leaning in — in practice, at or below 0.8 seconds, a different world from the untreated 4-plus it started at.
The constraint
The open kitchen pass was the restaurant's signature feature and explicitly off-limits to anything that would block sightlines to the chefs, and the street-facing glazing drove passing footfall too heavily to curtain over. That eliminated the two largest, most obvious surfaces before the brief reached a shortlist — and left the ceiling to do most of the work.
The decision
Ceiling-suspended acoustic baffles were positioned across roughly two-thirds of the ceiling, in the sections that don't sit over the pass, upholstered banquette seating replaced the loose chairs along the one uninterrupted brick wall, and felt-faced panels went on the single wall with neither glazing nor kitchen frontage — extra margin on top of the ceiling fix, not the thing doing the heavy lifting.
Every route considered, vendor-neutral — we don't sell any of it.
- Suspended ceiling baffles, clear of the passabout 82 m² (roughly two-thirds of the ceiling)
Positioned section by section to leave every sightline to the kitchen open — the single biggest lever in a room this dominated by hard, reflective surfaces.
- Upholstered banquette seating22 covers
Replaced loose bar-style chairs along the one clear brick wall — extra margin on top of the ceiling fix.
- Felt-faced wall panelsabout 14 m²
The only wall in the room without glazing or kitchen frontage to work around — a finishing touch, not the main event.
The outcome
The model predicts the three changes would bring the room from roughly 4.3 seconds to about 0.8, inside the guest-comfort range the brief called for, with the open kitchen left as visible as before. This is a design-stage prediction for an illustrative room, not a measured result.
Indicative figures for this example, from the same method a real report uses. Not a measured survey and not a compliance certificate — where one is required, that's a survey to BS EN ISO 3382-2.
Now do it for your room.
Four questions, about a minute. Free to see whether you have a problem.
Check my room →Questions people ask
Won't leaving a third of the ceiling untouched cost you the target?
It's close — the room's other hard surfaces (concrete floor, brick, the street glazing) mean the ceiling has to do most of the work, which is why the treated zone covers roughly two-thirds of it rather than a token strip.
Would this make the room quieter, or dead?
The model predicts quieter, not dead. 0.8 seconds is still on the lively side, deliberately. The brief was never silence; it was getting back into a range where a table of four can hear itself.
Is 0.8 seconds an official hospitality standard?
No — there's no regulation to point to for a restaurant, which is worth being honest about. It's a design judgement based on what comfortable dining rooms typically measure at, not a cited figure.