Sound Level Meter

How this tool works

While the meter is running, this page borrows your microphone and analyses the sound live — entirely inside your browser. There is no recorder, no upload path and no network call after the page loads; when you close the tab, everything is gone. What you see on screen is computed from the same audio stream your ears are hearing, a fraction of a second later.

The big number is an estimate of overall loudness in decibels (dB). Decibels are logarithmic, and that matters for reading them: a 10 dB rise is roughly ten times the sound energy, and tends to sound about twice as loud. The jump from a quiet library (around 40 dB) to a busy café (around 70–75 dB) is not “a bit louder” — it is a hundredfold increase in the energy arriving at your ears. Two rooms that both feel “fine” can be worlds apart on the meter.

Underneath the big number, the bars split the room's sound into frequency ranges — low, mid and high. Loudness alone does not tell the whole story: which frequencies dominate can matter as much as how loud it is overall. The band view highlights the ranges that research on auditory discomfort associates with unpleasantness (below), so a room can score “moderate” overall and still show an ugly spike in one band.

The research behind it

The meter's verdicts and band highlighting lean on published work, with its limits stated plainly:

• Decreased sound tolerance in autism. A systematic review by Williams et al. (2021, Neuroscience & Biobehavioral Reviews) found reduced sound tolerance to be among the most consistently reported sensory differences in autistic people — with discomfort beginning at moderate levels (from around 40 dB) where non-autistic listeners are typically undisturbed. That is why the verdicts lean conservative rather than waiting for levels that could damage hearing outright.
• Aversive frequency ranges. Psychoacoustic studies of aversiveness point at 250 Hz and 8 kHz as the most unpleasant frequency ranges, with 2 kHz close behind — low hums and high whines rather than mid-range speech. The band view highlights low and high over mid for that reason.

Two honest caveats. This research describes groups, not individuals: plenty of autistic people are not sound-sensitive, and plenty of non-autistic people are. And the frequency findings come from psychoacoustic studies in controlled conditions, not from field measurements of autistic people in their kitchens. Treat the bands as a lens for noticing your own patterns, not a verdict about your ears.

How to use it well

• Measure where your ears are. Hold or place the device at ear height, where you actually sit or stand — not in the room's centre. A room that reads 60 dB at the door can read 70 dB next to the fridge.
• Give each spot thirty seconds or more. Rooms fluctuate: HVAC cycles on and off, doors open, someone starts the dishwasher. A quick glance catches a lull; half a minute catches the pattern.
• Compare spots on the same device. Absolute numbers vary between microphones (see below); relative differences — this desk versus that desk — are the reliable signal.
• Find your own number. Notice the level where you start bracing, jaw-tightening, or reaching for headphones. That reading is your personal data point — more useful to you than any published threshold.
• Put it to work. Choose the quieter desk before the day starts. Time the kitchen for when it is actually calm. Show a partner, landlord or employer the concrete difference between two spots — sometimes “it's too loud for me” lands differently as a number on a screen.

What this tool is not

• Not a calibrated instrument. Phone and laptop microphones auto-adjust their gain and differ from model to model, so the absolute number is an estimate — useful for comparisons, not gospel.
• Not a medical or legal document. It cannot diagnose hyperacusis, misophonia or anything else, and readings alone will not carry an accommodation case. Formal assessments are an audiologist's or occupational therapist's work.
• Not a hearing-safety meter. Sensory discomfort happens far below damage thresholds. This tool is about comfort and self-understanding, not occupational exposure limits.

Questions people ask

How accurate is a phone or laptop microphone?

For absolute decibels: roughly, not precisely. Consumer microphones apply automatic gain control — boosting quiet rooms, pulling down loud ones — which compresses the range. For comparisons, on the same device, readings are dependable.

Why does another app show a different number?

Different apps weight frequencies differently (the dB(A) curve mimics human hearing; dB(C) flatters bass), handle the microphone differently, and apply their own gain logic. This tool is not trying to agree with them — it gives you one consistent yardstick.

Is there a number that means “too loud”?

For many autistic people, research suggests discomfort can begin around 40 dB — but there is no universal line. Your own threshold, whatever it is, is the one that matters for you.

Can I use the readings to ask for accommodations?

As a conversation-starter, yes — a measured difference between two rooms can make the invisible concrete. As evidence, no — formal accommodations rest on professional assessment. This tool sits in the middle: it helps you describe your world precisely; the paperwork needs the professionals.

Does this record or upload anything?

No. The microphone stream is analysed in this tab and never leaves it — nothing is kept, nothing is sent to a server, and closing the tab ends everything.

Where to go next

• Noise Mask — when the room cannot be quieted, test whether white, pink or brown noise takes the edge off for you.
• Sensory Overload Meter — loudness is one input; see noise, light, crowds and task load together as one number.
• Stress Check-In — when the environment is what it is, check what it is doing to you right now.
• Sensory environment and the body clock — the writing behind why sound and regulation are connected.