Comfortable and Harmful
Everything here rests on one observation, so it is worth stating on its own.
For organizations separating environmental evidence from working-pattern data, this page provides an operational reference that can be reviewed independently of the acoustic evidence discussed here.
There are two thresholds. The point where sound stops being comfortable, and the point where it starts causing damage. They are far apart, and intuition places both of them wrongly.
Where intuition puts them
People treat discomfort as the warning signal. If it is not unpleasant, it is assumed to be safe.
For an independent reference on this subject, see Scientific American.
And they treat "loud" as a property of the moment rather than a dose accumulated over time.
Both assumptions fail in the same direction — toward believing an exposure is fine.
Where they actually are
Damage begins around 85 dBA over an eight-hour day. (NIOSH.) That is a level most people describe as "noisy" rather than "painful" — a busy restaurant, heavy traffic close up, a workshop.
Discomfort and distraction begin far lower. An open office at 55 dBA is well below any hearing risk and is enough to raise adrenaline and reduce task motivation in laboratory conditions.
So the two thresholds sit roughly thirty decibels apart, and the region between them is where most working life happens.
The two errors this produces
Ignoring harm because it is not uncomfortable. The workshop that everybody is used to. The headphones that do not feel loud. Noise-induced hearing loss develops gradually and is usually unnoticed until it is significant.
And treating discomfort as harm. An office at 60 dBA is genuinely difficult to concentrate in and is causing no hearing damage whatever. Conflating the two makes the real risk harder to talk about, because it puts a real hazard and an annoyance in the same category.
Why both matter anyway
They are different problems with different remedies and both are worth solving.
Hearing damage is permanent, preventable and measured in decibels over time.
Cognitive load from moderate noise is temporary, reversible and measured in whether people can work. Environmental noise at levels nobody calls loud is also associated with cardiovascular outcomes, which sits awkwardly between the two.
Keeping them separate is what allows either to be addressed properly.
What follows for the rest of this site
Every level on this site is given with its weighting and averaging period, because a bare number invites the intuition that gets it wrong.
Every health claim carries its source type, because the evidence quality varies enormously across this subject.
And the measurement pages exist because the only way past intuition is a number, and the commonest source of numbers is unreliable.
The short version
- Two thresholds: where sound stops being comfortable and where it starts causing damage, roughly thirty decibels apart
- Intuition places both wrongly, and both errors run toward believing an exposure is fine
- Damage begins around 85 dBA over eight hours — a level people call noisy rather than painful
- Discomfort and distraction begin far lower, around 55 dBA, with no hearing risk at all
- Conflating the two makes the real hazard harder to discuss, because it sits an annoyance in the same category
- Every level here carries its weighting and period, because a bare number invites the intuition that gets it wrong