Sound attenuation is the reduction in sound energy as it travels from one point to another. Every sound you hear is quieter than it was at its source, whether that loss happens naturally over distance or because something (a wall, insulation, earplugs) absorbed or blocked part of the energy along the way. Understanding how and why sound loses strength matters in contexts ranging from building design to hearing protection to industrial noise control.
How Sound Loses Energy
Sound is a pressure wave that moves through a medium, whether that’s air, water, wood, or concrete. As it travels, several physical processes steal energy from the wave and reduce its intensity.
Absorption is the most intuitive. Every medium has some degree of viscosity and thermal conductivity. When sound waves cause particles in a material to vibrate, friction between those particles converts acoustic energy into heat. The wave’s amplitude shrinks with every bit of energy lost to friction. Soft, porous materials like fiberglass or acoustic foam are particularly good at this conversion, which is why they’re used to quiet rooms.
Scattering occurs when sound waves hit irregularities or particles within a medium. Instead of continuing in a straight line, the wave’s energy disperses in multiple directions. This doesn’t destroy the energy, but it spreads it out so that less of it reaches any single point.
Distance alone reduces sound intensity even in open air with nothing to absorb or scatter it. Sound radiating outward from a point source spreads over an ever-larger area. This follows the inverse square law: every time you double your distance from the source, the sound intensity drops by about 6 decibels. Move ten times farther away and the drop is roughly 20 dB. That’s why a lawnmower sounds dramatically quieter from across a field than from the sidewalk next to it.
Measuring Attenuation in Decibels
Sound attenuation is expressed in decibels (dB), a logarithmic scale that reflects how humans actually perceive loudness. The core formula compares two sound pressure measurements:
Change in intensity (dB) = 20 × log(P₂ / P₁)
Here, P₁ and P₂ are the sound pressures at two different points. The logarithmic relationship means that halving the sound pressure doesn’t halve the decibel reading; it reduces it by about 6 dB. A 10 dB reduction sounds roughly half as loud to the human ear, even though the actual intensity has dropped by 90%. This scaling is important because it means small-sounding dB reductions can represent large real-world differences in sound energy.
Absorption vs. Transmission Loss
When evaluating acoustic materials, two properties describe different jobs. The absorption coefficient measures how well a surface soaks up sound energy within a room, preventing echoes and reverberation. A thick carpet absorbs a lot; a tile floor absorbs very little. Sound transmission loss (STL), on the other hand, measures how much sound a wall, floor, or ceiling blocks from passing through to the other side.
These two qualities rarely go hand in hand. Most materials are good at one or the other. A heavy concrete wall has excellent transmission loss (sound has a hard time getting through it) but poor absorption (sound bounces right off its surface). A panel of acoustic foam absorbs sound effectively within a room but does almost nothing to stop sound from traveling through to the next room. Effective noise control in buildings often requires layering materials that address both properties.
Common Ratings: STC and NRC
Two standardized ratings help consumers and builders compare acoustic performance without needing to interpret raw lab data.
STC (Sound Transmission Class) rates how well a wall or floor-ceiling assembly blocks airborne sound. Testing involves mounting an assembly between two rooms, playing sounds across 16 frequencies between 125 Hz and 4,000 Hz (the range of human speech), and measuring how many decibels are lost on the other side. The resulting transmission loss values are plotted on a curve and compared against a standard reference. A higher STC number means better sound blocking. A standard interior wall might rate around STC 33 to 35. An STC of 50 or above means loud speech on the other side is barely audible.
NRC (Noise Reduction Coefficient) rates how well a surface absorbs sound within a room, on a scale from 0.0 to 1.0. An NRC of 0.0 means the surface reflects all sound (like a smooth concrete wall), while 1.0 means it absorbs everything. Soft materials like thick curtains, upholstered furniture, and acoustic ceiling tiles score higher. Hard surfaces like glass and polished stone score lower. NRC matters most in spaces where echo and reverberation are problems, like open offices, restaurants, or recording studios.
Attenuation in Building Construction
Architects and builders use four main strategies to attenuate sound between rooms or floors, often combining several at once.
- Adding mass: Dense, heavy materials let less sound pass through. A double layer of drywall blocks more sound than a single layer, and adding a layer of mass-loaded vinyl between them improves performance further.
- Decoupling: Breaking the physical connection between two surfaces interrupts the vibration path. A resilient channel between drywall and studs, or a wire-suspended ceiling below floor joists, prevents sound vibrations from traveling directly through the structure. This is the most effective method for controlling low-frequency sound, which is the hardest to block.
- Absorption within cavities: Filling wall or ceiling cavities with fiberglass or mineral wool insulation absorbs sound energy that would otherwise bounce around inside the cavity and eventually transmit through.
- Isolation of impact sound: Wood joist floors transmit a lot of impact noise (footsteps, dropped objects). Resilient underlayments beneath floating floors isolate the finished surface from the structural slab, dramatically reducing the thud that reaches the room below.
No single technique solves every noise problem. A massive concrete floor blocks airborne sound well but still transmits impact noise unless a resilient underlayment or floating floor is added on top. Layering strategies is what separates a quiet building from one where you hear every conversation next door.
Attenuation in Hearing Protection
Earplugs and earmuffs attenuate sound before it reaches your eardrums. Their performance is rated using the Noise Reduction Rating (NRR), a number in decibels printed on the packaging.
The NRR doesn’t translate directly to real-world protection, though. OSHA’s method for estimating actual noise reduction requires subtracting 7 dB from the NRR, then subtracting the result from the measured noise level. So if you’re working in a 100 dB environment and your earplugs have an NRR of 29, the estimated noise reaching your ears is 100 minus (29 minus 7), or about 78 dB. That 7 dB correction accounts for the difference between laboratory testing conditions and how the devices perform when real people wear them in real workplaces.
Fit matters enormously. Foam earplugs that aren’t fully inserted, or earmuffs that don’t seal around glasses frames, can lose a significant portion of their rated attenuation. The best-rated protector is only as good as its seal against your skin.
Why Frequency Matters
Sound attenuation is not uniform across all pitches. Low-frequency sounds (bass, rumbling, traffic noise) are harder to attenuate than high-frequency sounds (speech consonants, bird calls, alarms). Low-frequency waves are longer and carry more energy, which lets them bend around obstacles and pass through materials that easily block higher pitches.
This is why you can hear the bass from a neighbor’s music through a wall but not the vocals. It’s also why decoupling (physically breaking the vibration path) is so important in construction: adding mass alone may handle mid and high frequencies, but low-frequency control requires interrupting the structural connection that lets those long waves travel through solid materials. STC testing captures this frequency dependence by measuring performance across the full 125 to 4,000 Hz range rather than relying on a single frequency.