Sound test
Four checks that cover what actually goes wrong with speakers and headphones: a channel that has gone quiet, wiring that cancels the bass, a range that has narrowed, and sound that arrives after the picture. Each one gives a result you can act on rather than a number to admire. Runs in the browser, on whatever you listen through.
Start the sound testRuns entirely in your browser
The run, in order
- Speaker and headphone testPlay sound through the left speaker, the right, or both, and find out whether a channel is dead, swapped or quiet. Free, in the browser.
- Polarity testFind out whether one of your speakers is wired backwards, which drains the bass and hollows out the middle. Free, runs in the browser.
- Frequency range testSweep from 20 Hz to 20 kHz and find where your headphones, speakers or hearing stop. Free, runs in the browser.
- Audio and video sync testA flash and a click on one beat. Slide the click until they land together and read the offset between picture and sound in ms. Finds Bluetooth and TV delay.
Why this order
Channels first, because a channel that has gone quiet is the commonest fault in this category and the quickest to find, and because two of the later tests are meaningless if one side is silent. Left, right, both, and a note of which side is which, so a swapped cable is caught before it is mistaken for anything else.
Then polarity. A fault nothing measures: both channels play, both meters read normal, and the system sounds hollow because the two speakers are cancelling each other in the middle. It comes second because it needs both channels working to show, and because it is completely free to fix once found.
Then range. A sweep from the lowest to the highest frequency, and the point at which you stop hearing it is a property of the equipment and of your hearing together. It is placed here because it takes a minute of attention and because its result is information rather than a fault — most of the time the number is simply what it is.
Sync last. The only test here that is about time rather than sound, and the one that matters most on anything wireless. It asks you to sit and judge for a minute, which is why it is at the end rather than the start.
What a browser sound test cannot see
A web page can make a sound. It cannot hear one. Everything after the sound card — the amplifier, the driver, the room, the seal of an earcup against a head — is on the far side of the only sensor it has, which is you.
So there is no loudness figure here, no distortion measurement, no frequency response curve. Any page that shows you those from a browser is showing you what it sent, not what arrived. What a browser can do honestly is put a known sound on each path and ask what you heard, and every test here is built that way. The one measurement it can make by machine — the round trip through a microphone — has its own page in the microphone run.
When to run it
Something sounds wrong and you want to know whether it is the equipment, the wiring, a setting or the source. Channels and polarity separate those in two minutes.
Buying used headphones or speakers. A silent side, a broken cable at the plug and a driver that has been overdriven are the three faults to expect, and the first two tests find them before money changes hands.
New wireless headphones, to know their delay before it surprises you in a game. Films will usually be fine; the Bluetooth delay checklist covers what to do about the rest.
After changing anything in the chain — a new cable, an adapter, a receiver, a soundbar — because a single reversed connection somewhere in it produces a fault that every level meter will miss.
Running it properly
Use the equipment the question is about, and sit or wear it as you normally would. For speakers, the polarity test depends on sitting between them.
Set the volume to normal listening level before starting, and do not turn it up to chase the frequency sweep: the top of the range is quiet by nature and driving it hard damages tweeters and ears in that order.
Close anything else that is playing sound or holding the microphone — a call left open in another tab will switch a Bluetooth headset to its telephone profile and change every result on this page.
Common questions
What can a browser sound test actually check?
Anything that comes down to what reaches your ears. Whether each channel plays, and on the correct side. Whether the two are wired in phase, which is what gives sound a centre and a bottom end. How high and how low you can hear through this equipment, which is a property of the equipment and of you together. And how far the sound lags the picture, which is the question on any wireless headphones. What it cannot do is measure loudness, distortion or frequency response in numbers — a browser can play a sound and cannot listen to what came out, so those need a microphone and an analyser. Nothing here claims them.
Should I test with speakers or headphones?
Whatever you have the question about, and for the channel and polarity checks, the thing you actually use. Headphones will pass the polarity test even when the speakers are wired backwards, because the cancellation happens in the air between two speakers and there is no air between two earcups. If the question is about a soundbar or a pair of monitors, sit where you sit and play through them.
Why is there no volume or loudness test?
Because a web page cannot hear. It can ask the sound card for a level and cannot know what the amplifier, the speaker and the room did to it. Any page showing you a decibel figure for your speakers is showing you the number it sent, not the number that came out. The tests here are all ones where your ear is the right instrument.
The sound is fine here but bad in a game or on a call.
Then the equipment is fine and the application is the variable. On Windows, a Bluetooth headset switches to a telephone-grade profile the moment an app opens the microphone, which is why music sounds good and calls sound thin on the same headphones. Games and calls also apply their own processing. If every test here passes, the equipment has been ruled out, and that is worth knowing before buying different equipment.
Which test matters most for buying used headphones?
The channel test, then the frequency range. A driver that has failed or a cable that has broken at the plug shows as one side quiet or silent, and that is the common fault. A range that stops well short of what the headphones should reach — under about 14 kHz on a pair sold as full range — is a driver that has been damaged, usually by being driven too hard. Both take a minute, and both are things a seller cannot show in a photograph.
Checks that use these tests
- Bluetooth headphones delayedSound behind the picture on Bluetooth: measure the delay in ms, see how much is the link, and find which fix — codec, profile, player offset or cable — applies.
- Sound in one ear onlyWork out whether it is the plug, the pairing, the balance, wax in the mesh, the cable, or the headphones. Checks in the order that costs nothing first.