Headphone boxes advertise frequency response to 40 kHz. Your ears stopped going that high a long time ago. Enter your age for a rough idea of where your own ceiling sits — and how much of the spec sheet is being sold to a part of you that has already retired.
The number above is a population median read off a table. This plays real tones through whatever you are wearing and asks you, one frequency at a time, whether you can hear them. It takes about a minute.
Set your volume before you start. Turn it down to a level where a quiet voice would be comfortable, then use the 1 kHz calibration tone to settle it. The test never plays louder than that calibration tone. If anything is uncomfortable, press Stop.
Headphones or earphones, not speakers — and both sides on. Nothing is recorded and nothing is sent anywhere; the tones are generated in this page.
It does not mean high-resolution audio is pointless, and it does not mean an expensive headphone is wasted on you. Almost everything that makes headphones sound different from each other — tonal balance, bass extension, how the midrange is voiced, distortion, how the drivers behave under load — happens far below 10 kHz, comfortably inside anyone's range. A 45-year-old with a 14 kHz ceiling hears essentially all of what separates a good headphone from a bad one.
What it does mean is that ultrasonic frequency-response claims are marketing, not a feature. When a spec sheet leads with "5 Hz – 40 kHz", it is advertising territory nobody's ears occupy, and it tells you nothing about how the headphone actually sounds in the range you live in. Judge by tonal balance and measurements in the audible band instead.
An audiogram is measured on calibrated equipment in a quiet room, with the level at each frequency known in decibels. None of those three things is true here. The level depends on your volume knob, the room depends on where you are sitting, and the equipment is whatever headphone you happen to be wearing.
That last one matters more than people expect. Above about 10 kHz, headphones are neither flat nor consistent — the gap between two models up there, and between one fitting and the next on the same model, is larger than anything you would tolerate further down. This site does not publish measurements, so take no figure from us on it; it is the one thing every set of high-frequency measurements agrees on. So a tone you cannot hear at 16 kHz may be a tone your headphone barely produced. The honest reading of a result here is “the highest tone this headphone, at this volume, in this room, got into my ear”, and the way to separate ear from headphone is to run it again on a different pair.
Bluetooth adds a second reason for caution: several codecs discard the top of the band before it reaches the driver at all. If you are testing wirelessly, a ceiling around 15 or 16 kHz may be the codec rather than you.
The two silent trials are there for one reason. Listening hard for a tone at the edge of hearing, people hear tones that are not playing — this is ordinary, not a character flaw, and it is why a test without catch trials reliably flatters whoever takes it. If you answered yes to a silent trial, the result above is read as a ceiling you did not quite reach.
The table at the top of this page is population medians, not a hearing test. The tone test below it is a self-test, not a hearing test either. If something has changed recently in one ear, or you have ringing that will not settle, see an audiologist rather than a calculator.