More than half the enquiries this practice receives contain the same confusion, and it is entirely reasonable that they do, because the trade uses one word for two unrelated problems. “Soundproofing” is asked for; what is described is sometimes insulation and sometimes room treatment, and the materials that solve the two have almost nothing in common. Buying one when you needed the other is the commonest and most expensive mistake in the field, and it is made every week.
Two different questions
Insulation asks: how much sound gets from this room to that one? It is a question about a partition, and it is answered in decibels of transmission loss.
Treatment asks: how does sound behave inside this room? It is a question about reverberation and reflections, and it is answered in seconds and in absorption coefficients.
They share no physics beyond the fact that both involve sound. A room can be superbly insulated and horrible to be in. A room can be beautifully balanced and audible three doors away. Improving one does essentially nothing for the other, and one popular material actively makes the second problem worse while appearing to address the first.
What insulation actually requires
Transmission loss through a single solid partition is governed by the mass law: roughly 6 dB for every doubling of surface mass, and 6 dB for every doubling of frequency. It is a punishing relationship. Going from a 100 mm block wall to a 200 mm one — twice the material, twice the weight, twice the foundation — buys 6 dB. Real insulation is bought with three things, in this order:
Mass, because there is no substitute for it. A dense partition simply moves less when sound strikes it.
Decoupling, because a break in the structural path defeats the mass law’s economics. Two leaves of 15 kg/m² separated by a wide cavity outperform a single leaf of 60 kg/m², and the improvement grows with the cavity width. This is why a genuine studio is a box inside a box, and why anything rigid bridging the cavity — a joist, a conduit, a mistakenly fixed batten — throws away most of the benefit through one small contact.
Sealing, because sound follows air. A one per cent open area in an otherwise excellent partition limits the whole assembly to about 20 dB. The letterbox, the socket back-to-back with its neighbour, the gap where the duct passes through: these are the reason real partitions underperform their laboratory figures, and hunting them is most of the site work.
Porous absorbent inside the cavity helps, by damping the resonance of the air spring between the leaves. It contributes perhaps 5 dB in a construction that is already correct. It contributes nothing at all on its own.
What treatment actually requires
Treatment is porous material, panel absorbers, and shape. It is thin, it is light, and it goes on the room side of the surface. A hundred millimetres of mineral wool with a stretched fabric face is superb treatment. Its transmission loss is approximately zero: put it on a party wall and your neighbour will hear exactly as much as before, though your own room will sound less harsh, which is precisely the illusion that sells acoustic foam by the pallet. The room got quieter to be in. Nothing crossed the wall any differently.
The reverse confusion happens too, and produces worse buildings. A meeting room built as two layers of dense board on resilient bars, fully sealed, is properly insulated and — with no absorption inside — reverberant, harsh, and worse to hold a meeting in than the open plan it was carved out of. That is a room that met its specification and failed its purpose.
How to tell which one you have
Two questions settle it almost every time.
Where is the sound you object to being made? If it is being made in another room, another building or outside, you have an insulation problem and no amount of material inside your room will fix it. If it is being made in the room you are standing in, you have a treatment problem.
Does it get worse when the room is empty? Reverberation does — carpets, furniture and people are absorption, and an empty room rings. Transmitted noise does not care.
We ask both on the first phone call, and we say plainly which of the two the caller has, including on the occasions when the answer is that they need neither and should move the printer. It costs us a small amount of work each year and saves a good deal of somebody else’s money, which over a practice’s lifetime has turned out to be the better trade.
