How to Soundproof a Music Room

Building research published by the National Research Council of Canada sets out the whole method in one place, and it has not changed since. Mass comes first: layers of solid material such as plasterboard, timber or concrete resist being moved by a pressure wave. Decoupling comes second, and in lightweight construction it does more work than mass, because building a wall or floor as two mechanically independent structures with an air space between them greatly reduces what crosses it.

Absorption in the cavity is third, filling the space between the two structures so the air gap does not simply act as a drum. The research is usefully unglamorous about the detail: the cavity insulation needs to be only about 75mm, and its exact position within the cavity does not matter much, which means more wool than that buys very little.

Airtightness is fourth and is the one most often lost during construction. The same research puts it plainly: containing sound is in many ways like containing water, and unless the room is close to airtight the sound leaks into the adjoining spaces. It also names the honest tension in that, since complete airtightness is undesirable because the room still needs ventilating, so the goal is eliminating unwanted leaks rather than sealing the room shut.

The rule quoted most often in this field is that sound insulation improves by up to 6 dB for each doubling of the mass of a single leaf partition, and by a similar amount for each doubling of frequency. It is a real relationship and it is the reason a second layer of plasterboard is worth adding.

It is also heavily qualified, and the qualifications rarely travel with the rule. The mass law assumes an infinite panel with negligible bending and shear stiffness, no damping forces, and plane waves arriving at any angle, and it holds only below the critical frequency at which coincidence effects take over. Real walls are finite, stiff and damped, so real walls deviate from it.

There is also a point at which adding mass stops being the answer. Approved Document E notes that where an adjoining masonry wall already has a mass of more than 375 kg per square metre, a wall lining may not be necessary because it may not give a significant improvement. Past a certain weight of structure, the next gain has to come from decoupling instead.

Approved Document E defines flanking transmission as sound transmitted between rooms by any path other than the direct one through the separating element, and a flanking element as any part of the building that contributes to that transmission without being the separating wall or floor. In a house, those elements are the ordinary structure: joists bearing into a wall, a floor screed or boarding running continuously beneath a partition, ceiling voids, wall cavities, ductwork, pipework and the junctions between all of them.

The construction rules in that document are largely rules about breaking those paths. Floor joists are to be supported on hangers rather than built into a separating wall. Spaces between joists running at right angles to a wall are to be sealed with full depth timber blocking. Junctions between a ceiling and a wall lining are to be taped or caulked, air paths into cavities blocked with solid blocking, and the mass of a floating screed left out of the calculation entirely because it is not doing that job.

The document is direct about what happens when those paths stay continuous through a conversion, warning that additional treatments may be required and that specialist advice may be needed, and noting elsewhere that an improved ceiling helps only where there is no significant flanking transmission. This is the single most common reason a carefully specified room underperforms: the wall was built correctly and the sound went around it.

Two different problems run in parallel and need different treatments. Airborne sound propagates through the air and is addressed with mass and isolation. Impact sound results from direct contact with a building element, travels onward as structure-borne sound, and is addressed with resilient layers and floating floors. The regulations treat them as separate quantities and even in opposite directions, setting a minimum figure for airborne insulation and a maximum figure for impact sound transmission.

For a music room this distinction is not academic. A drum kit is an impact source as much as an airborne one, a piano transmits through its own frame into the floor, and an amplifier cabinet couples directly into whatever it stands on. Adding mass to the walls of a room whose main problem is energy going into the floor addresses the wrong half of the problem, which is why floor isolation is usually where a music room build starts rather than where it finishes.

A room performs at the level of its weakest element, and the weakest element is almost always a door. The building research is blunt about the numbers: a typical hollow core internal door rates around 10 to 15, and even a solid core timber door in a gasketed frame reaches only about 28 on the same scale. Where there is no lobby or buffer space, the door alone controls the level of sound entering or leaving the room, whatever the walls are made of.

The reason a small weak element dominates is arithmetic rather than intuition. Composite performance is averaged across the area using transmission coefficients rather than decibels, and an unsealed gap transmits everything that reaches it, so a small proportion of open area drags the whole result down. Published acoustics research goes further, showing that a hole can draw energy from a region considerably larger than its own physical size at its resonant frequency, so a gap is not merely a hole of its own size.

Ventilation is the same problem with a duty attached, because a sealed room has to be ventilated mechanically. Approved Document F sets a minimum whole dwelling rate of 0.3 litres per second per square metre of floor area, and also sets an acoustic expectation, that a ventilator should not exceed 30 dB LAeq in a noise sensitive room when running at its minimum continuous rate, with resilient mountings where structure-borne noise could be a problem. The isolation side is handled with lined ductwork, where the building research suggests a minimum of three metres of acoustic liner in each duct entering the room, with bends rather than a straight path, and notes that lining reduces the free area so the duct may need to be larger.

Every double leaf construction has a frequency at which it performs worse than the parts would suggest. The air in the cavity acts as a spring between the two panels, and at the mass air mass resonance the panels move out of phase and the insulation dips. Published measurements put that dip typically between 60 and 100 Hz depending on the depth of the cavity, which is precisely where a kick drum and the low strings of a bass guitar live.

Cavity absorption does not rescue it. The same research states that fibrous absorption in the cavity improves transmission loss above the resonance region but gives no improvement at the resonance itself, and lists the trend towards bass driven popular music as one of the reasons the problem has become more noticeable. Deepening the cavity moves the dip lower, which is one of the reasons a serious music room is built thicker than a room built for speech.

The measurement conventions hide this. European sound insulation regulations, this country's included, apply a default frequency range starting at 100 Hz, and only a couple of countries incorporate low frequency descriptors into their requirements between dwellings. A low E on a bass guitar sits at about 41 Hz. A compliant 45 dB rating is therefore silent about the exact band in which a music room is most likely to fail.

Putting those constraints together gives an honest picture. A room inside a terraced or semi-detached house is bounded by four things at once: flanking paths through structure shared with a neighbour, the door, the ventilation penetrations, and a resonance dip sitting in the bass. Three of the four can be improved with money. The first cannot, because the details that break flanking paths sit inside a structure you only half own.

A detached building in a garden starts from a different position, and the difference is definitional rather than promotional. If a flanking element is a part of the building that carries sound between rooms, a separate structure across the garden shares none: no continuous joist, no party wall, no screed, no ceiling void, no shared duct. The paths that Approved Document E spends whole sections controlling do not exist to be controlled, which is why the design can address mass, decoupling, the door and the ventilation independently instead of negotiating with an existing building.

Before spending on either, it is worth knowing what does not work. Building research states directly that applying absorbing materials such as acoustic tile, cork or carpet to the surface of a wall, floor or door does not substantially reduce transmission between rooms, and only reduces reflections within the room. Closed cell foams such as polystyrene are poor absorbers as well, so they do neither job. Lightweight decorative panels are not heavy enough to reduce transmission on their own. Sockets should be offset by around half a metre rather than sitting back to back, and recessed downlights need boxing so the ceiling layer stays unbroken. The difference between isolation and treatment is where most wasted money in this field goes.

If the room in the house is the wrong room, that is worth finding out early rather than after the work. The Institute of Acoustics is the professional body to approach for independent advice on an existing building, and telling us what you play and where you live will get you an honest view of whether a separate structure in the garden answers the problem better than treating a room indoors.