What is the sound absorption coefficient (α)?
The sound absorption coefficient — written α ("alpha") — is the fraction of sound energy a surface absorbs rather than reflects. It runs from 0 to 1: α = 0.00 reflects everything, α = 1.00 absorbs everything. It is dimensionless — it has no unit — and it is always quoted per frequency band, because nearly every real material absorbs treble far better than bass. Values here come from our sourced absorption database. We don't sell panels.
The short version
- α = absorbed energy ÷ incident energy. A ratio, so it has no unit.
- Range 0 to 1. Our 60-material dataset spans α 0.01 to 1.00.
- One number per octave band, not one number per material — the frequency dependence is the whole point.
- The unit belongs to the surface, not the material: absorption area A = S × α, measured in sabins.
- Published values above 1.00 are normal, and not a mistake — see the edge-effect section below.
What α actually measures
When sound strikes a surface, the energy splits: some reflects back into the room, some is absorbed (converted to heat by friction in the material, or lost through it). α is simply the absorbed share of that arriving energy:
α = Eabsorbed ÷ Eincident
Because both sides of the division are energies, the units cancel and α comes out as a bare number. That is the answer to the most common question about it: the sound absorption coefficient has no unit at all.
Here is what that split looks like for one real row from the database — carpet, heavy, on foam pad at 500 Hz, where α = 0.57:
Why there is never just one α
A material does not have "an" absorption coefficient — it has one for every frequency band. This is the single most important thing to understand about α, and the reason a single-number summary such as NRC can mislead. The same carpet, heavy, on foam pad, across all six standard octave bands:
| Band | 125 Hz | 250 Hz | 500 Hz | 1 kHz | 2 kHz | 4 kHz |
|---|---|---|---|---|---|---|
| α | 0.08 | 0.24 | 0.57 | 0.69 | 0.71 | 0.73 |
It absorbs 73% of the energy at 4 kHz and only 8% at 125 Hz — roughly 9× more treble than bass. Quoting one number for this material would hide exactly the information you need to fix a boomy room.
What the numbers mean in practice
Four real rows from the database, read at 500 Hz:
| Material | α at 500 Hz | Energy absorbed | What that means |
|---|---|---|---|
| Marble or glazed tile | 0.01 | 1% | Effectively a mirror for sound. Reflects almost everything. |
| Gypsum board (12 mm on studs) | 0.05 | 5% | A hard wall. Slight bass absorption from the cavity behind it, nothing in the mids. |
| Carpet, heavy, on foam pad | 0.57 | 57% | Genuinely absorptive in the mids and highs; nearly transparent to bass. |
| Mineral wool, 100 mm, on wall | 1.00 | 100% | A full-range absorber — the reason panels are built from it. |
The unit that does exist: absorption area (sabins)
α is unitless, but the quantity you actually design with is not. Multiply a coefficient by the surface area carrying it and you get absorption area:
A = S × α — measured in m² sabins (or ft² sabins)
A room's total absorption is the sum over every surface, A = Σ (Sᵢ × αᵢ), and that total is what drives reverberation time. The same 10 m² of wall gives very different results depending on what covers it, at 500 Hz:
| 10 m² covered with… | α | Absorption area |
|---|---|---|
| Marble or glazed tile | 0.01 | 0.1 m² sabins |
| Gypsum board (12 mm on studs) | 0.05 | 0.5 m² sabins |
| Carpet, heavy, on foam pad | 0.57 | 5.7 m² sabins |
| Mineral wool, 100 mm, on wall | 1.00 | 10.0 m² sabins |
That is the whole reason coefficients matter: they convert an area of material into an amount of absorption you can add up. Our RT60 calculator does this sum for a whole room, and the porous absorber calculator predicts α itself from a material's thickness, airflow resistivity and air gap.
How α is measured — and why values above 1.00 appear
Published coefficients are not calculated from theory; they are measured. The standard method (ISO 354 / ASTM C423) puts a sample of known area into a reverberation chamber and derives α from how much faster the room's sound decays with the sample present. This is random-incidence absorption — sound arriving from all directions at once — which is why chamber figures differ from a normal-incidence impedance-tube measurement.
That method also explains the coefficients above 1.00 you will find in manufacturer datasheets. Sound diffracts around the exposed edges of a finite sample and is absorbed there as well, so the sample behaves as if it were larger than its nominal area — but the coefficient is computed against the nominal area, so the ratio can land above 1.00. It is a measurement artefact, not a physical impossibility: no surface absorbs more energy than reaches it. Our dataset caps values at 1.00 because planning estimates should not inherit that artefact.
α versus NRC
NRC is a summary of α, not an alternative to it: the average of the 250, 500, 1000 and 2000 Hz coefficients, rounded to the nearest 0.05. It deliberately excludes 125 Hz, which is why a material can carry a strong NRC and still do nothing for bass. Use α to design; use NRC to compare quickly. The full argument, with data, is in our guide to NRC.
Where to get α values
Coefficients are material- and construction-specific, so they have to be looked up rather than derived. Our absorption coefficient database lists all 60 materials across all six octave bands with a cited source for every row, downloadable as CSV under CC BY 4.0. For a specific commercial product, always check that manufacturer's current datasheet — and note whether they quote chamber (random-incidence) or tube (normal-incidence) figures, because the two are not interchangeable.
Frequently asked questions
What is the sound absorption coefficient?
The sound absorption coefficient (α, "alpha") is the fraction of sound energy that a surface absorbs instead of reflecting. It runs from 0 to 1: α = 0.00 means the surface reflects all the energy that reaches it, and α = 1.00 means it absorbs all of it. A value of 0.57 means about 57% of the incident energy is absorbed and roughly 43% is reflected back into the room. It is always quoted per frequency band, because almost every real material absorbs high frequencies far better than low ones.
What is the unit of the sound absorption coefficient?
It has no unit. α is a dimensionless ratio — absorbed energy divided by incident energy — so the units cancel out, which is why it is written as a bare number between 0 and 1 rather than as so many decibels or sabins. The related quantity that *does* carry a unit is absorption area, A = S × α, where S is the surface area: in metric that is measured in m² sabins, and in imperial in ft² sabins. So a material has a coefficient (no unit) and a surface has an absorption area (sabins).
What is the formula for the sound absorption coefficient?
α = E_absorbed / E_incident — the sound energy not reflected, divided by the energy that arrived. In practice α is not calculated from that definition but measured: a sample is placed in a reverberation chamber and the change in the room's decay time is used to derive it, per ISO 354 / ASTM C423. To go from coefficients to a room's total absorption you sum each surface separately: A = Σ (Sᵢ × αᵢ), which is the quantity that then drives the Sabine reverberation-time formula.
Can the absorption coefficient be greater than 1?
In published chamber measurements, yes — values like 1.05 or 1.10 appear regularly, and they are not errors. A reverberation-chamber test (ISO 354 / ASTM C423) measures a finite sample, and sound diffracts around its exposed edges and is absorbed there too, so the sample behaves as though it were slightly larger than its measured area. The coefficient is computed against the nominal area, so the ratio can exceed 1.00. Physically no surface absorbs more energy than reaches it. Our dataset caps values at 1.00 for planning use, and spans 0.01 to 1.00.
What is a good sound absorption coefficient?
It depends entirely on the frequency you need to fix, which is why a single "good" number is misleading. Above about α 0.70 in a band, a material is doing real work in that band. But carpet, heavy, on foam pad reaches α 0.57 at 500 Hz while managing only α 0.08 at 125 Hz — so it is excellent for echo and useless for boomy bass. Read the band that matches your problem, not the average.
What is the difference between the absorption coefficient and NRC?
α is the per-band measurement — one value for each octave band, showing how absorption changes with frequency. NRC compresses four of those bands (250, 500, 1000 and 2000 Hz) into a single averaged number, rounded to the nearest 0.05, and ignores 125 Hz entirely. Use α when you are designing treatment, especially for bass; use NRC only as a fast rough comparison between products.
Sources
Coefficients on this page are drawn from the same sourced dataset as the rest of the site — every row cites its origin. Measurement method: ISO 354 / ASTM C423 reverberation-chamber testing. See the database for per-row citations and the methodology page for how values are compiled and normalised.