Device test · Audio
Headphone Test
Six randomised trials, your answers, one verdict on how your headphones are wired. The test does not tell you which channel it is using until after you have answered — which is the only way swapped wiring can be caught, because a swapped pair sounds perfectly normal to someone who already knows which button they pressed.
How this test works
Each trial generates a 440 Hz sine tone in your browser and routes it to one channel, the other, or both, using the Web Audio API's stereo panner. Nothing is downloaded and no audio file is involved — the tone is synthesized on your device at the moment you press the button, at a fixed modest level, with a short fade at each end so that starting and stopping cannot themselves produce the click this page is meant to help you diagnose.
The order is drawn from a printed integer seed, and it is balanced rather than random: every run contains exactly two left trials, two right and two both. That balance is what the verdicts depend on. A sequence that happened never to test the left channel could not distinguish a silent left channel from one that was simply never played, and a test that cannot tell those apart should not be printing verdicts about either.
When six answers are in, the pattern is resolved against a small table. All answers matching is a correctly wired pair. Every single-channel trial reported from the opposite side, with the both-channel trials still centred, is a swapped pair. One side reported as silent while the both-channel trials come only from the other is a dead channel — though this page cannot tell you whether the fault is in the headphones, the cable, the jack or the source, and it says so rather than guessing. Everything reported as centred is a mono output. Anything else is named inconclusive, which is a real result and not an error.
Frequently asked questions
What does this headphone test actually check?
How your headphones are wired, and whether both channels reach your ears. It plays six randomised trials — some to the left channel only, some to the right only, some to both — takes what you report hearing, and resolves the pattern into one verdict: correctly wired, channels swapped, one channel silent, or both channels carrying the same signal. It is a test of the equipment and the signal path, and it is not a measurement of anything about you.
Is this a hearing test?
No. This page tests headphones, not hearing. Nothing here measures a threshold, an audible range, a frequency response or any other property of a person, and no result it produces means anything about your ears. Every verdict is a statement about wiring and channels. If you have a concern about your hearing, a browser page is the wrong instrument and a qualified professional is the right one.
Why does the test hide which channel is playing?
Because knowing ruins the result. If you press a button marked Left and then ask yourself which side you heard, expectation supplies a large part of the answer — and that is exactly why swapped headphones are so easy to live with for years. A swapped pair sounds completely normal until something plays a tone without telling you which side it used. Randomising the order and withholding the channel until after you answer is what makes swapped wiring detectable at all.
The verdict says my channels are swapped. What should I do?
Check the obvious causes in order: headphones worn the wrong way round, an extension or adapter that reverses the pair, a balance or channel-swap setting in the operating system, or a mislabelled cable. Re-run this test after changing one thing at a time — the seed is printed, so you can run the same sequence again or take a fresh one. This page can tell you the channels are reversed; it cannot tell you where in the chain the reversal happens.
What is the seed for?
It fixes the trial order. The same seed always produces the same sequence of left, right and both trials, so a result can be re-run rather than wondered about — useful when you change a cable and want the identical test again. Press New seed for a different order. Each run is balanced regardless of the seed: two left trials, two right and two both, because a sequence that never tested one side could not tell a silent channel apart from an untested one.
Device Bench