How to read a noise specification

How to read a noise specification

Noise is treated as a secondary specification and is frequently the one that decides whether a product gets used. If you share a home, thin walls make an otherwise excellent object useless.

Most listings handle this with adjectives. "Whisper quiet" is not a measurement. A decibel figure is better — but it is easy to misread, and there are noise problems it cannot describe at all.

The scale is not linear

Decibels are logarithmic. This is the part that trips people up, because the numbers look like an ordinary scale and behave nothing like one.

Roughly:

- +3 dB is about double the sound energy — perceptible, but not dramatic.
- +10 dB is about ten times the energy, and is generally perceived as roughly twice as loud.

So the gap between a 42 dB device and a 48 dB device is not "a bit over ten percent noisier." Six decibels is a substantial, clearly audible difference. Conversely, 45 dB versus 46 dB is not a difference you would reliably identify in a blind comparison.

The practical reading: differences under about 3 dB are close to noise in the specification itself. Differences of 5 dB and up are real.

What a number roughly corresponds to

These are approximate reference points, useful for calibration rather than precision:

| Level | Rough equivalent |
|---|---|
| ~30 dB | A quiet bedroom at night |
| ~40 dB | A refrigerator running in the next room |
| ~50 dB | Quiet conversation, light rainfall |
| ~60 dB | Ordinary conversation at close range |

Most of the relevant range for these products sits between about 40 and 55 dB. Below roughly 40 dB, ambient household noise starts to cover the sound anyway.

What the number leaves out

Here is where a single figure stops being sufficient.

Measurement conditions are rarely stated. Distance from the microphone, whether it was measured in an anechoic chamber or an office, which intensity setting was running, and whether the device was in open air or against a surface — all of these move the result substantially. Two manufacturers can measure honestly and publish different numbers for identical hardware. A figure without conditions attached is a rough indication, not a comparable measurement.

Peak versus typical. A figure is usually taken at one setting. A device that is quiet at low intensity and loud at high intensity has one number on the page and two experiences in use.

Pitch matters more than level. This is the largest omission. A low-frequency hum is absorbed by walls, bedding and furnishings far more readily than a high-frequency whine. Two devices at an identical 45 dB can carry very differently through a wall depending on where their energy sits in the frequency range. Low, deep and rumbly travels less than thin and high — and no consumer specification reports this.

Contact noise is separate from airborne noise. A device resting on a hard surface transmits vibration into that surface, which then radiates sound. A hard headboard or a wooden floor can be louder than the device itself. This is not captured by any measurement of the device in isolation.

Load changes everything. Motors are measured unloaded. Under pressure against a surface, a motor's tone and level both shift, usually upward.

What actually helps, in practice

The specification is one input. These do more:

- Soft surfaces absorb. Bedding, cushions and rugs all reduce transmitted noise substantially. A hard floor or a wall-mounted headboard does the opposite.
- Lower intensity settings are quieter — non-linearly. Motor noise tends to rise faster than intensity does. A device with a usable low end is a quieter device in practice, whatever its headline figure.
- Continuous is less noticeable than pulsing. Steady sound fades into the background of attention; a pattern that starts and stops keeps drawing it back. A pattern mode can be more noticeable than a steadier setting at the same measured level.
- Ambient noise covers a lot. A fan, ventilation or ordinary street noise raises the floor above which anything needs to be heard.

What we publish

Where we have a measured figure for a product, we state it on the product page along with the intensity setting it was taken at. Where we do not have one, we leave the field empty rather than substituting an adjective.

An empty field is more useful than "whisper quiet," because it tells you something true: we do not have this number. An adjective tells you nothing and implies something.

FAQ

Is a lower decibel number always better?

For your purposes, generally yes — but the difference has to be large enough to matter. Under about 3 dB, treat two products as equivalent on noise and decide on something else.

Why do reviews disagree with the specification?

Because reviewers measure in real rooms, at unstated distances, under load, often at higher settings. Both figures can be honest and still differ by a lot. Reviews are more useful for the qualitative description — "high-pitched," "deep hum" — than for the number.

Is an air-pulse or suction device quieter than a vibrating one?

Not necessarily, and the character differs. Air-pulse mechanisms often produce a lower-pitched sound that carries less through walls, but they also make a distinct sound when the seal against the skin breaks. See how suction toys actually work.

Does a quieter device mean a weaker motor?

Not reliably. Motor mounting, housing design and damping affect emitted noise independently of power. It is fair to say that, all else equal, more power tends to produce more noise — but all else is rarely equal between two different designs.

---

Specifications on our product pages are stated with their measurement conditions where we have them, and left blank where we do not.