Skip to the sound test

Speaker & Headphone Test

100% private

Check your left and right channels, balance, polarity and bass — then get a verdict, not just a beep. No microphone, and nothing to upload.

Turn your volume down to about a third first. Test tones are much louder than music at the same setting, and you can always turn it up once the first sound plays.

Channel Left Idle
Channel Right Idle

Four short sounds, with one question after each. It takes about half a minute, and at the end you get a plain-English verdict — a dead channel, sides swapped, mono instead of stereo, or a speaker wired backwards — along with what to do about it. Put both earpieces on, or sit between both speakers, and press Start.

Which side did you hear that from?

1 What are you testing?

This changes the advice at the end — swapped headphones are worn round the wrong way, swapped speakers are a cable at the back.

2 Test sound

Keep Noise unless you have a reason not to. A pure tone gives your ears almost no clues about direction, which is exactly what this test needs them to have.

Working…
Advanced options

A share of the tool’s own ceiling, not of your volume. Even at 100% these tones stay well below full scale — use your device’s volume control for real loudness.

Checking usage…

More tests

The individual tones, for when you already know what you are listening for.

A single pure tone at any frequency from 20 Hz to 20 kHz, in either channel or both. Useful for finding the exact pitch at which something starts to buzz, and for checking one earpiece against the other.

The slider is spaced by pitch rather than by number, so the bass end is as reachable as the treble. Drag it while the tone plays and it retunes without stopping.

A slow rise from 20 Hz to 20 kHz over about 22 seconds. It should get steadily higher in pitch and stay roughly even in loudness — a sudden hole, a rattle, or the sound vanishing early tells you where your speakers give up. When you stop hearing it near the end, press the button and we’ll record the frequency.

Most adults stop hearing somewhere between 15 and 17 kHz, and that ceiling falls with age — it is normal, not a fault. Your speakers may also simply stop before your ears do. This is a rough listening check, not a hearing test, and it is no substitute for an audiologist.

Steps down through the bass in third-octaves, from 200 Hz to 20 Hz, holding each for a couple of seconds. Press the button at the first step you cannot hear and we’ll record the one before it as your low limit.

For scale: a phone speaker usually gives up around 300–500 Hz, a laptop around 150–200 Hz, a decent bookshelf speaker reaches 50–60 Hz, and below about 30 Hz you feel it more than hear it. If a step buzzes or clatters rather than humming cleanly, that is not bass — that is the next test.

! Rattle check

A steady 80 Hz tone, held for six seconds at a healthy level. That is where a driver moves furthest, so a torn surround, a loose grille or a resonating panel will announce itself as a buzz or a clatter over the top of the note.

What your browser reports about the audio path to your speakers. All of it comes from the browser’s own audio engine — no microphone, no permission, and nothing that identifies you or your machine.

Step by step

How to test your speakers or headphones

About half a minute, no account, no app, and an answer at the end rather than a row of buttons.

Say what you are using

Headphones or speakers. It changes nothing about the sounds — it changes the advice, because swapped headphones are worn the wrong way round and swapped speakers are a cable at the back.

Turn the volume down

To about a third. Test tones carry far more energy than music at the same setting. You can always turn it up after the first sound; starting loud is how people hurt their ears.

Answer what you heard

Each sound is followed by one question with four answers. Answer for the side the sound actually came from, not the side you expected — the whole test hinges on that distinction.

Read the verdict

You get a named fault and what to do about it, not a shrug. Copy the whole thing to the clipboard if you need to paste it into a support chat or a returns request.

Private by design

Nothing here to upload, and no microphone in sight

Most sound-test pages stream audio files from a server, wrap the controls in advertising, and ask for your microphone to “check the other end”. This one asks for nothing, because it needs nothing — the tones are made on your device, out of arithmetic, the moment you press play.

No microphone

Your browser never asks for microphone access here, because this tool never uses it. Nothing in your room is listened to, recorded or analysed — your own ears do the listening, and your answers stay in this tab.

Nothing to upload

There is no file to choose, no account to make and nothing stored between visits. The verdict is worked out here, shown here, and gone the moment you close the tab.

Works with the wifi off

Every sound is synthesised rather than downloaded, so once this page has loaded it needs no connection at all. Turn off your wifi and run the whole test — no site that streams audio files can do that.

To be precise about what does leave your device — because “100% private” is easy to say and worth checking. Two things, and neither is anything you did here:

1. When a test finishes, this page sends one anonymous message to our counter saying “sound-test was used”. Not your answers, not the verdict, not which frequencies you could hear. It exists so the usage number in the panel is a real one rather than something we invented.
2. Google Analytics records the page view, that a test completed, and the verdict’s short code — dead-left, swapped, ok. That is a fact about a piece of hardware, and it is how we tell whether this page is diagnosing anything useful. You can block it and the tool works exactly the same.

And one thing we want to be straight about: we are not measuring your speakers. There is no microphone involved, so nothing here can hear what actually came out. We generate the tones and keep score of your answers — the judgement is entirely your own ears. Any page claiming to measure your audio quality without listening to it is claiming something impossible.

At a glance

Tool specifications

What it checksLeft and right channels, stereo balance, polarity, bass reach, high-frequency reach, buzz and rattle
Faults it namesSilent channel, swapped channels, mono instead of stereo, balance offset, reversed polarity, no output at all
Sound sourceGenerated in your browser — pink noise and sine tones. No audio files are downloaded
Frequency range20 Hz to 20 kHz
Maximum levelCapped at about −11 dBFS, well below full scale, at every setting
MicrophoneNot used, and never requested
Where processing happensIn your browser, on your device — nothing is uploaded
Works offlineYes, once the page has loaded, including as an installed app
Sign-upNot required, not offered
PriceFree

Questions

Sound testing, answered

Is anything recorded or uploaded?
No, and there is nothing here that could be. The tool never asks for your microphone, never reads a file and never writes your answers anywhere. Every tone is generated by your own browser from a formula. The simplest proof: disconnect from the internet and run the whole test — it works exactly the same, which no site that streams audio files can manage.
One side is silent. What do I do?
Work out whether it is the headphones or the computer before you buy anything. Plug them into a different device and run this test again there: if the same side is silent, the headphones are at fault; if both sides work, the problem is in your computer. The most common cause by far is the left/right balance slider in the sound settings, which is easy to knock by accident and which nothing else on your machine will warn you about.
How do I know if my channels are swapped?
That is exactly what the first two steps establish. The tool plays a sound to the left channel only and asks which side you heard it from; if you say your right, and then the right-channel sound comes out of your left, they are crossed. Swapped channels are surprisingly easy to live with unnoticed — most music sounds fine either way — but they ruin films, games and anything mixed for direction.
What is the polarity test actually doing?
It plays the same low tone twice: once normally, and once with the right channel inverted, so that its speaker pushes air out at the moment the left one pulls it in. When both speakers are wired correctly the first version sounds fuller and more solidly centred, and the second sounds thin and hard to place. If the second one sounds fuller, one of your speakers is wired with its + and − terminals the wrong way round. The effect is much weaker on headphones, where each driver has its own ear and no shared air to cancel in — so treat a headphone result as a hint rather than a verdict.
Why can’t I hear the 20 Hz tone, or the 18 kHz one?
Almost certainly nothing is wrong. Very few speakers reproduce 20 Hz at all — a phone gives up somewhere around 300 Hz, a laptop around 150 Hz — and human hearing at the top end falls steadily with age, from around 20 kHz in a child to 15 kHz or lower by middle age. Both limits are normal. What is worth investigating is a tone that appears, disappears and reappears as the frequency rises, or one that buzzes rather than hums.
Can these tones damage my speakers or my hearing?
Not at sensible volumes. Every sound here is capped well below the maximum the browser could produce, and each one fades in and out rather than starting abruptly, which is what actually stresses a driver. The real risk is your own volume control: a steady test tone carries much more energy than music at the same setting, and it will feel far louder. Start at about a third and turn up from there.
Does this tell me how good my speakers sound?
It tells you two honest things and refuses to pretend to a third. It tells you how low and how high your speakers actually reach, because you report where the tone disappears, and it tells you whether they buzz under load. What it cannot tell you is how they measure — that needs a calibrated microphone in a treated room, and any web page claiming to do it without listening to your speakers is claiming something impossible. Our tones are known; your ears are the instrument.
Why do I have to answer questions? Can’t it just detect the problem?
A browser can tell you what it sent to the sound card. It cannot tell you what came out of the speaker, whether the cable is crossed, or which ear you are wearing which cup on — the interesting faults all live past the point where software can see. Answering could be avoided by asking for your microphone, but that would mean listening to your room to diagnose your headphones, and the result would still be confounded by echo and background noise. Four taps is a better trade.

If it turns out to be the hardware

When the fault is the headphones, not the computer

If the same side stayed silent on a second device, the test has done its job and the pair is finished. These are current picks from our catalogue, priced across the range — a cheap replacement pair through to a set worth keeping.

Prices and availability as of 11 August 2026, and they change often — Amazon’s page is the one that counts. As an Amazon Associate we earn from qualifying purchases, at no extra cost to you. Browse every audio product we track

Keep it handy

Save it, share it, or tell us what’s missing

Come back to it

Worth keeping for the next pair of headphones, or the next time a laptop starts sounding lopsided. Bookmark it, or pin it to the top of the tools list on this device.

Something not right?

If the verdict was wrong, if a sound never played, or if you have a fault this test doesn’t catch, tell us. The form opens with the tool already filled in, so you only have to describe what happened.