Marlow Hall is a shoebox because the site is a shoebox. The plot is a forty-six metre strip of reclaimed quay between a listed bonded warehouse and the river wall, and by the time we were appointed the plan was already fixed by the party walls either side of it. That is usually bad news. Here it was not: the proportion the site forced on the building — long, narrow, and tall enough to be uncomfortable to build — is close to the proportion a symphonic room wants anyway. Our first report said so in one sentence and then spent nine pages on the section, which was the part still open to argument.
The room finished at 44 m long, 23 m wide and 17 m to the underside of the ceiling. That is 17,204 m³ for 1,800 seats, or 9.6 m³ a seat. A hall of this size has to sit somewhere between eight and eleven cubic metres a seat if it is going to hold two seconds with a full audience in it, because the audience is the absorption; below eight you are fighting your own seating, above eleven you are heating a volume nobody hears. The 17 m ceiling is the number we defended hardest and the one that came closest to being lost.
The surface schedule is unremarkable, and that is the point. The seating block is 1,080 m² of upholstered seats and does the great majority of the work. Above and around it are 2,400 m² of timber panelling on battens — a panel absorber tuned by its mass and its cavity, chosen because it takes out bass where wool does not and because it can be shaped. The remaining 500 m² of plaster and 320 m² of parquet are acoustically almost inert. Computed by Sabine at 500 Hz the schedule gives 2.01 s, which is where a room asked to hold Bruckner and Debussy in the same week has to be. Measured occupied it came in at 1.99 s, and 1.61 s at 4 kHz, the difference being air absorption over a forty-metre path rather than anything on the walls.
We lost the reflector canopy. It was in the Stage 3 drawings as fourteen suspended elements over the platform, it was priced, and it went in the second value-engineering round in favour of a rebate on the seating. What we got instead was the side-wall modelling in the lower two-thirds of the room, which returns early lateral energy to the stalls and does most of what the canopy would have done for the audience — but nothing at all for the players, who still cannot hear the back desks as well as they should. We said so at the time, we say so now, and the trust has a costed scheme in a drawer for when there is money.