How Sound Travels in a Room – Physics in Plain English

Why you should understand how sound moves

If you speak in a room and hear an echo, you're experiencing physics in real time. Sound propagation follows measurable laws, and these very laws determine whether your home office echoes or whether you can relax and listen to music in your living room. Over the past seven years I've analysed more than 1,200 rooms, and in 80 per cent of cases the problems could be explained by three fundamental physical principles: reflection, absorption and diffraction. Sound propagates at 343 metres per second in room air, which means in a 5-metre-wide room a sound wave bounces between the walls roughly 34 times per second before being noticeably damped. This fact alone explains why small rooms are often more acoustically demanding than large ones. If you understand the mechanisms behind it, you'll avoid expensive mistakes and deploy absorbers or diffusers precisely where they must work physically.

Direct sound, reflections and reverberation: the three phases

Every sound in a room passes through three phases. First, the **direct sound** reaches you: the shortest connection from the source to your ear, usually a straight line. This direct sound carries most of the energy and determines how loud you perceive something. Milliseconds later the **early reflections** arrive—sound waves that bounce once off a wall, ceiling or floor before reaching your ear. In a typical 4 by 5 metre office, early reflections arrive about 10 to 30 milliseconds after the direct sound; the human brain merges them with the original signal, which is why speech often sounds coloured or more reverberant. The third phase is the **diffuse reverberation**: hundreds of overlapping reflections that fill the room evenly and decay slowly. The reverberation time, measured as RT60 in seconds, indicates how long the sound level takes to drop by 60 decibels. An empty living room with parquet flooring and plasterboard walls often reaches RT60 values around 1.2 seconds, whereas 0.4 to 0.6 seconds would be ideal for speech intelligibility.

In a consultation last year, a psychotherapist faced the problem that patients could hear snatches of conversation through the corridor. The solution lay not in thicker walls but in understanding the early reflections: sound travelled across the ceiling of the waiting room, was barely absorbed there, and passed through the thin corridor wall to the outside. Two acoustic sails on the ceiling (NRC 0.85) reduced the sound pressure in the waiting area by 8 decibels—enough to restore confidentiality. This shows that if you know where sound travels, you save material and money.

Frequency determines behaviour: low tones are more stubborn

Sound waves are pressure fluctuations that propagate in waves. The **wavelength** λ (lambda) is calculated from the speed of sound c (343 m/s) divided by the frequency f in hertz: λ = c / f. A 100-hertz tone has a wavelength of 3.43 metres, a 1,000-hertz tone only 0.343 metres. This difference explains why low frequencies behave physically differently from high ones.

Low frequencies **diffract** around obstacles; a 50-hertz hum travels effortlessly around a desk. High frequencies behave more like light: they are reflected sharply by edges or scattered by rough surfaces. In practice, this means that if in a room with a 2.50-metre ceiling height you only hang 4 cm thick acoustic panels on the wall, you improve intelligibility (because mids and highs are absorbed), but the bass continues to boom. Bass frequencies need either very thick absorbers (at least λ/4, so for 100 Hz around 85 cm) or specially tuned panel resonators.

Reflection, absorption and scattering: the three tools

**Reflection** occurs when sound hits a surface whose dimensions are large compared to the wavelength. Smooth, hard walls (concrete, glass, plasterboard) reflect over 95 per cent of the incident energy in the mid and high frequency range. The angle of reflection equals the angle of incidence—you know this from billiards. In rectangular rooms this creates **flutter echoes**: sound bounces back and forth between parallel walls, you hear a metallic "zing-zing-zing" when you clap your hands. I advise you to do this test in your room: clap once and listen for flutter echo. If present, you need absorbers or diffusers on at least one of the two opposite walls.

**Absorption** converts sound energy into heat. Porous absorbers (mineral wool, foam, acoustic felt) work through friction: air molecules move in the fine pores, rub against each other and lose energy. The **absorption coefficient α** (alpha) ranges between 0 (no absorption) and 1 (complete absorption). A 5 cm thick acoustic foam typically achieves α = 0.60 at 500 Hz, α = 0.95 at 2,000 Hz, but only α = 0.10 at 125 Hz. Porous absorbers are therefore high-pass filters; they largely let lows through. If you buy "acoustic foam" indiscriminately, without knowing the thickness and density, you'll probably only solve high-frequency problems.

The effect also depends on **distance from the wall**. Porous absorbers work best where the **particle velocity** of the sound wave is maximal, which is λ/4 in front of the reflecting wall. For 125 Hz (λ = 2.74 m) that would be 68.5 cm distance. In practice you therefore often mount absorbers with a 10 to 20 cm air gap behind the panel; this shifts the effect to lower frequencies. A 5 cm thick panel with 15 cm spacing works acoustically like a 20 cm thick panel directly on the wall—an elegant solution when room depth is limited.

**Scattering** (diffusion) distributes sound evenly in many directions without absorbing it. Diffusers are usually structured surfaces: bookshelves, corrugated wood panels or QRD diffusers (Quadratic Residue Diffuser) with mathematically calculated recesses. A well-placed diffuser behind the listening position in a home recording room prevents rear-wall reflections from arriving at the ear out of phase with the direct sound, without making the room "dead". Diffusion is sensible when you want to preserve liveliness, for instance in music rooms or large living rooms where too much absorption stifles the atmosphere.

Standing waves and room modes: when the room resonates

In enclosed rooms **standing waves** form when the room dimensions are integer multiples of half the wavelength. The lowest resonance (first axial mode) occurs at f = c / (2·L), where L is the room length. In a 5-metre-long room the first longitudinal mode lies at 343 / (2·5) = 34.3 Hz. Such modes amplify certain frequencies by 10 to 20 decibels, while other frequencies are cancelled out. You hear this as booming, uneven bass: at one spot in the room the subwoofer thumps, two metres further it sounds thin.

Axial modes (between two parallel surfaces) are strongest, tangential modes (across four surfaces) weaker, oblique modes (across all six surfaces) weakest. In practice the first three to five axial modes dominate the sound picture below 200 Hz. I recommend entering the room dimensions into a mode calculator (available free online); if you see three modes at 48 Hz, 49 Hz and 51 Hz, you know the room has a massive problem in this range that can only be mitigated with thick bass absorbers or clever furniture placement.

Avoid **square or cubic rooms** if you have influence over the building design. A 4 by 4 metre room has identical modes in length and width that overlap and amplify the problem. The ratio 1.00:1.28:1.54 (about 4 by 5.12 by 6.16 metres) is considered acoustically favourable because modes are more evenly distributed. In existing rooms, asymmetric furnishing sometimes helps: a sofa not centred but 1.2 metres from the rear wall; a shelf diagonally in the corner—this breaks the symmetry and dampens individual modes.

Sound propagation in open-plan layouts: without walls it becomes harder

In a project for an architectural office (open work area, 60 square metres, 8 staff) we installed a total of 12 square metres of ceiling sails and lowered the reverberation time from 1.3 seconds to 0.6 seconds. The perceived volume dropped, concentration improved, measurable through fewer sick days and higher productivity, according to internal feedback. This shows that even without walls, effective acoustics are achievable if you understand the propagation paths.

• **Ceiling sails** above the dining area (at least 3 to 4 square metres, NRC 0.80 or higher)—they catch sound before it floods the whole room.
• **Room dividers with absorbent backing** (e.g. acoustic partitions, 180 cm high)—they act as a sound screen for direct sound, even if they don't reach the ceiling.
• **Textiles and carpets**—a 3 by 4 metre carpet with fleece backing absorbs a surprising amount in the mid and high-frequency range, making voices softer.
• **Furniture as diffusers**—a half-height bookshelf across the room scatters sound without optically partitioning.

Why thin curtains and carpets alone aren't enough

A widespread misconception: "I'll hang heavy curtains, then it'll be quieter." Heavy curtains (e.g. blackout curtains, 400 g/m²) do indeed have an absorption coefficient α of about 0.30 to 0.50 at mid frequencies, but only if they hang **spaced from the wall** and **in folds**. Pulled flat against the wall they do almost nothing, because the particle velocity directly at the wall is near zero. A typical 2 by 2.5 metre curtain offers 5 square metres of area; at α = 0.40 it effectively absorbs 2 square metres. In a 20 square metre room with 50 square metres of wall surface that's a drop in the ocean.

Carpets help mainly against impact sound and high frequencies. A wool carpet (10 mm thick) reaches α ≈ 0.50 at 2,000 Hz, but only α ≈ 0.05 at 125 Hz. If your room booms (typical sign of bass problems), a carpet changes nothing about the booming but makes voices clearer because highs are damped. I therefore advise you to use curtains and carpets as a **supplement**, not as the main measure. The main load should be carried by porous absorbers at least 5 cm thick (preferably 8 to 10 cm) or panel absorbers for bass.

Practical rules of thumb for your room acoustics

A calculation example: you have a 4 by 5 metre office (20 square metres). Target absorber area 20 · 0.20 = 4 square metres at NRC 0.80. You buy 8 acoustic panels of 60 by 60 cm each (0.36 square metres each, 2.88 square metres total) with NRC 0.85. Effective absorber area: 2.88 · 0.85 = 2.45 square metres—you're still 1.55 square metres short of the target. Solution: two additional panels or a ceiling sail with 2 square metres and NRC 0.90 (effectively 1.8 square metres). This is how you calculate concretely, rather than buying "by feel".

• **Absorber area**: plan 15 to 25 per cent of the floor area as effective absorber area (NRC-weighted). In a 20 square metre room, therefore, 3 to 5 square metres of absorber with NRC 0.80 or higher.
• **Early reflection positioning**: clap your hands and listen for where it echoes. Place absorbers where the sound reflects once before reaching you—typically side walls at ear height, ceiling above the seating position.
• **Bass absorption**: for noticeable effect below 150 Hz you need either 15 cm thick mineral wool with 10 cm air gap or panel resonators (chipboard on battens, cavity filled with insulation wool).
• **Diffusion instead of absorption**: if the reverberation time is already short (RT60 < 0.4 seconds) but the room sounds "dead", replace individual absorbers with diffusers or structured surfaces.
• **Break symmetry**: avoid identical treatment of opposite walls; this prevents flutter echoes and standing waves.

Software and measurement: when you want to know precisely

Anyone who wants to go deeper should try a measurement with **REW (Room EQ Wizard)**, free software for Windows and Mac. You need a measurement microphone (e.g. Dayton EMM-6 for £50) and an audio interface. REW shows you the frequency response at each listening position, the reverberation time frequency-dependent (RT60 curve) and waterfall diagrams that visualise standing waves. In a measurement you immediately see whether there's a 15-decibel peak at 80 Hz (typical for a room mode) or whether the reverberation time above 1,000 Hz is too long.

Alternatively there are apps like **AudioTools** (iOS, about £20 for the RT60 module) or **SignalScope**—less precise than REW, but sufficient for everyday purposes. I myself use REW in every consultation to document objectively before and after the measure. In one case the mid-range reverberation time fell from 1.1 seconds to 0.5 seconds, and the peak at 63 Hz was reduced by 8 decibels. The client heard the difference immediately, but the measurement gave him the assurance that the £890 investment (6 bass absorbers, 8 broadband panels) actually worked.

If that's too technical for you, commission a consultation. An on-site analysis with measurement costs between £300 and £600 in the UK—often well invested given the costs you save through mis-purchases.

What you should do now

You now know that sound propagates at 343 metres per second, that low frequencies have long waves and are therefore difficult to damp, and that reverberation arises from hundreds of overlapping reflections. You understand why 5 cm foam swallows highs but lets bass through, and why square rooms are acoustically problematic. This knowledge enables you to read product descriptions critically: if a supplier advertises "universal acoustic panels" without stating thickness, density and NRC value, you know the physics are missing and you'll steer clear.

My advice: start with a stocktake. Clap your hands, listen for flutter echoes. Measure the room dimensions and calculate the first three room modes. Identify which frequency ranges cause problems (reverberation on speech = mids/highs; booming = bass). Then choose absorbers that physically suit the task: 5 to 8 cm thick for mids and highs, 15+ cm or panel resonators for bass, diffusers for liveliness. Position them where sound hits obstacles—walls, ceiling, corners. And if you're uncertain, ask: a consultation will save you the third mis-purchase.