The science of soundproofing and acoustics

Sound is energy. It is made, it travels, and it arrives. Every noise problem sits somewhere on that path, and the two ways of dealing with noise work at different points on it. Acoustic treatment absorbs sound inside a room. Soundproofing reduces the sound that passes between rooms. This page sets out the physics of both, with the sources, so the difference is clear and can be checked. The aim here is to help you work out which problem you actually have, so you can fix the right one.

What is the difference between soundproofing and acoustic treatment?

Acoustic treatment absorbs sound inside a room to reduce echo and reverberation. Soundproofing, properly called sound insulation, reduces the sound energy passing from one space to another. They act on different stages of the same process and they are not interchangeable. Absorbing sound in a room does nothing to stop it reaching the room next door.

All sound moves through three stages. Generation, where a source sets air or a structure vibrating. Propagation, where that energy travels as a wave through air or as vibration through a solid. Reception, where it arrives at a listener or sets a surface vibrating so that surface becomes a new source. Energy is only ever reduced by being reflected away or turned into heat, and the process ends when all of it has become heat. Absorption acts on propagation inside the room, converting energy to heat before it re-radiates. Insulation acts on the transfer of energy into and through the structure. This distinction runs through everything below.

Sources: Hopkins, Sound Insulation, 2007. Cox and D'Antonio, Acoustic Absorbers and Diffusers, 3rd edition, 2017.

How is sound generated and described?

Sound is generated when a vibrating surface accelerates the air next to it, producing a travelling wave of alternating compression and rarefaction. The energy moves; the air itself does not travel with it. Four quantities describe the result: sound pressure, the fluctuation about normal air pressure, in pascals; particle velocity, the small back and forth velocity of the air; sound intensity, the flow of power per unit area, in watts per square metre; and sound power, the total energy rate of the source itself, which does not change with the room around it.

Level is measured logarithmically in decibels. Sound pressure level is twenty times the logarithm of the pressure against a fixed reference of twenty micropascals, the quietest sound a healthy ear detects. The reference values are set in ISO 1683.

Sources: ISO 1683:2015. Kinsler, Frey, Coppens and Sanders, Fundamentals of Acoustics, 4th edition, 2000. Fahy and Gardonio, Sound and Structural Vibration, 2nd edition, 2007.

How does sound travel through the air?

In open air a point source radiates a spreading sphere of sound. Because the same power is spread over a larger and larger surface, intensity falls with the square of the distance, and the level drops by about six decibels each time the distance doubles. This holds only in a free field, meaning no reflections, as outdoors or in an anechoic chamber. Inside a room it holds only close to the source. Further away the reflected sound, the reverberant field, builds up and the level stops falling with distance. It is then set by how absorbent the room is, which is why adding absorption lowers the general noise level in a room.

Sources: Kuttruff, Room Acoustics. Kinsler et al., Fundamentals of Acoustics, 4th edition, 2000.

How does sound pass through a wall?

When an airborne sound wave meets a wall, the changing pressure pushes and pulls the surface into bending vibration. That vibration travels through the solid, and the far face, now vibrating, pushes the air on its side and radiates a new sound wave into the next room. The wall does not let sound through as if through a hole. It receives energy on one side and re-emits it on the other, and the far surface acts as a fresh source.

The sound that comes out the other side is quieter than the original for four reasons that compound. Most of the energy is reflected at the first surface, because air and a solid are so different that little energy crosses between them, an effect called impedance mismatch. Some of the vibration is turned into heat inside the wall by internal damping. Only a small fraction is passed through by an effective wall in the first place. And the far surface radiates less than perfectly efficiently across most of the range.

Sources: Fahy and Gardonio, Sound and Structural Vibration, 2nd edition, 2007. Cremer, Heckl and Petersson, Structure-Borne Sound, 3rd edition, 2005.

What is the difference between airborne and impact sound?

Airborne sound and impact sound are not separate phenomena. They are two ways energy enters the same chain. Airborne sound, a voice or a television, first travels through the air and then has to cross into the wall, losing most of its energy at that boundary. Impact sound, a footstep or a dropped object, is delivered straight into the structure by contact, skipping that lossy first step, so far more energy enters the structure and it is harder to control. After that first difference both behave the same way: they travel through the structure and re-radiate as airborne sound elsewhere.

This is why a footstep can be heard through a wall as well as through a ceiling. Once energy is in the structure it spreads through whatever the structure connects to. What separates airborne from impact is where the energy got in, not whether a given element can carry it.

Sources: ISO 10140 series. Hopkins, Sound Insulation, 2007.

What is sound absorption and how does it work?

Sound absorption is the conversion of sound energy into heat inside a room. A porous material such as mineral wool or acoustic felt works because sound forces air to move through its fine, connected structure, and friction and heat exchange in those tiny channels turn the sound energy into a small amount of heat. Porous absorbers work best at middle and high frequencies and lose effect at low frequencies unless they are made very thick or spaced off the wall. Two other types exist: panel absorbers, a limp sheet over a sealed cavity, which work at low frequencies, and Helmholtz resonators, a cavity vented through a neck, which absorb a narrow band of low frequencies.

Absorption is rated by the absorption coefficient, a figure between zero and one at each frequency, and summarised by single numbers such as NRC and the weighted coefficient defined in ISO 11654, measured to ISO 354. PET felt is a porous absorber. It reduces echo and reverberation within the room it is placed in. It does not add meaningful mass and it does not stop sound passing through a wall to another room. Absorbing sound in a room and blocking sound between rooms are two different jobs, and this is the physical reason a panel cannot do the second.

This is the problem absorption does solve, and it is a real one. A hard, echoey room makes your own television, music and conversation harder to make out, and it makes a space more tiring to work and rest in. Softening those reflections with a porous material makes a room easier to listen in and calmer to be in. There is a knock-on worth knowing too. When a room is hard to hear in, people turn the volume up, and that louder sound is exactly what pushes more noise through the wall to a neighbour. Treating the room so you can hear clearly at a lower volume can quiet a wall problem indirectly, by removing the reason to turn things up in the first place. Our acoustic felt range is made for this side of the problem.

Sources: Cox and D'Antonio, Acoustic Absorbers and Diffusers, 3rd edition, 2017. ISO 354:2003. ISO 11654:1997. Kuczmarski and Johnston, NASA, Acoustic Absorption in Porous Materials, 2011.

What is reverberation time?

Reverberation time is how long sound takes to fade in a room, defined as the time for the level to fall by sixty decibels after the source stops. It is longer in large rooms and shorter in absorbent ones. The Sabine equation gives it as a constant times the room volume divided by the total absorption. The constant is often quoted as 0.161, but it is not truly fixed: it depends on the speed of sound and therefore on temperature, sitting near 0.161 at twenty degrees and drifting either side with temperature. For very absorbent rooms the Eyring form is used instead.

Sources: Kuttruff, Room Acoustics. ISO 3382-2:2008.

What is the mass law?

Across the middle of the frequency range, a single solid wall blocks sound in proportion to its mass. The heavier the wall, and the higher the frequency, the more it blocks. As a rule, the sound reduction rises by about six decibels each time the mass per unit area is doubled and by about six decibels each time the frequency is doubled. This is the mass law, and it is the reason heavy walls block better than light ones.

The mass law carries a constant, and there is no single settled value for it, which is a real feature of the physics rather than a gap. The value depends on the angle at which sound strikes the wall. For the diffuse, all angles case used in building work it sits between about minus 47 and minus 48; for sound arriving straight on it is nearer minus 42. Anyone looking for one definitive number will not find one, because the number is conditional by nature. The six decibels per doubling, however, is not disputed.

Sources: Cremer, original 1942 derivation. Fahy and Gardonio, Sound and Structural Vibration, 2nd edition, 2007. Hopkins, Sound Insulation, 2007.

Why does decoupling improve soundproofing?

Mass alone is a slow way to gain sound insulation, because each doubling of weight buys only about six decibels. Splitting a wall into two separate leaves with a gap between them does much better. Above a certain low resonant frequency the two leaves act largely independently, and the reduction climbs far faster with frequency than a single wall of the same total weight. Filling the cavity with a porous material damps resonances inside the gap and improves the middle and upper range further.

There is one thing that undoes it. Any rigid connection across the cavity, a fixing or a tie, carries vibration straight from one leaf to the other and short circuits the gap, which is why decoupled constructions are built to break that path. So effective soundproofing of a wall combines two moves: adding mass, and decoupling a second layer from the first. This is the principle behind Shell, which adds mass and a separated inner layer to an existing wall on one side.

Sources: Hopkins, Sound Insulation, 2007. Fahy and Gardonio, Sound and Structural Vibration, 2nd edition, 2007.

What is the coincidence effect?

At high frequencies a single wall has a weak point called the coincidence dip, where its sound reduction falls below what mass alone would predict. It happens when the wavelength of the bending vibration in the wall lines up with the incoming sound, letting the sound drive the wall very efficiently. The lowest frequency at which this occurs is the critical frequency, and it depends on the wall's thickness and stiffness: thick heavy panels have a low critical frequency, thin ones a high one. Damping in the wall blunts the depth of the dip.

Sources: Cremer, Heckl and Petersson, Structure-Borne Sound, 3rd edition, 2005. Vigran, Building Acoustics, 2008.

What is flanking transmission?

Flanking is sound that gets from one room to another by paths going around the separating wall rather than through it: along connected floors and ceilings, through junctions, up shared structure, or through gaps and service holes. Because the total sound getting through is the sum of every path, the loudest path sets the limit. Once flanking carries more than the wall itself, improving the wall achieves almost nothing. This is the main reason real buildings perform worse than laboratory tests of the same wall, and it is why sound insulation is a whole construction problem, not just a question of the wall.

Sources: ISO 12354-1:2017. Hopkins, Sound Insulation, 2007.

Do acoustic panels stop noise from neighbours?

No. Acoustic panels and felt absorb sound inside the room they are in, which softens echo and makes a room sound less harsh. They do not add the mass needed to stop a neighbour being heard through a wall. This follows directly from the physics above: absorption acts on sound bouncing around inside a room, while blocking sound between rooms is about mass and decoupling in the structure. They are different jobs. A panel sold as a fix for a neighbour heard through a party wall is solving a problem the buyer does not have. Stopping sound coming through from next door is a soundproofing job, not an absorption one.

What about noise through floors and ceilings?

Noise through a floor or ceiling is usually impact sound, footsteps and knocks delivered straight into the structure from above, and it behaves differently from voices through a wall. A wall system treats the wall and does not address a floor or ceiling. Reducing impact sound is a separate job, worked on at the floor with resilient layers and floating constructions, or at the ceiling. Felted handles floors and ceilings as an enquiry rather than an off the shelf figure, and you can send the details over for a straight answer on which kind of noise you have and what can be done.

How is sound insulation measured and rated?

Sound insulation is measured two ways, and the difference matters. In a laboratory, with side paths suppressed, a wall's own performance is measured as the sound reduction index, summarised as a single weighted number. In a real building the whole separation between two rooms is measured, including flanking and workmanship, and written as a standardised level difference. Field quantities carry a prime mark to show they include everything the real building adds. Spectrum adaptation terms, written as C and Ctr, adjust the single number for different kinds of noise, with Ctr weighting for low frequency sources such as traffic and bass. Field figures are systematically lower than laboratory ones, because the laboratory removes the flanking, gaps and imperfect workmanship that a real building has.

Sources: ISO 717-1:2020. ISO 16283:2014. ISO 10140-2:2021.

What are the UK building regulations for sound between homes?

In England the governing document is Approved Document E, Resistance to the passage of sound, in its 2003 edition with later amendments, in force from April 2015. Between separate homes the airborne requirement is written as the standardised level difference with the low frequency correction, DnT,w + Ctr, and higher is better. For purpose built dwellings the minimum is 45 decibels for both separating walls and separating floors. For homes formed by converting an existing building the minimum is 43 decibels. For impact sound through separating floors, measured as the standardised impact sound level and where lower is better, the limit is 62 decibels for new dwellings and 64 decibels for conversions. Internal walls and floors within a single home have a separate laboratory minimum of 40 decibels. Scotland, Wales and Northern Ireland set their own figures under separate guidance.

Source: HM Government, The Building Regulations 2010, Approved Document E, Resistance to the passage of sound, 2003 edition incorporating 2004, 2010, 2013 and 2015 amendments. Tables 0.1a and 0.2.