Hearing Frequency Test

Read the number as a property of the whole chain. That is your headphones, your sound card, your volume setting, the noise in the room, and then your ears. It stops at whichever of those runs out first, and on most setups that is not the ears.

Small speakers and cheap earbuds fall away above 15 kHz. Laptop speakers are usually worse. Some audio drivers quietly filter the top of the range. A Bluetooth codec throws the highest frequencies away on purpose, because almost nothing is heard up there.

A real measurement of hearing is done with calibrated headphones in a quiet booth, and the level arriving at the eardrum is known in decibels. Nothing in a browser knows any of that. So this page reports where your chain stopped and does not translate it into anything about you.

If sounds seem muffled, if one ear is quieter than the other, or if there is ringing, that is a question for a doctor. It is not a question a web page can answer, and this one will not try.

Every other version of this test asks whether you can still hear the tone. That question cannot be answered wrongly. Nobody can check you, so the score measures how willing you are to say yes.

It is worse than it sounds. Play no tone at all and a good share of people still report hearing something. A quiet room is full of small sounds, and expectation fills the rest in.

So the tone hides in one of two windows here and you say which one. Guessing is right half the time and no more. Getting three in a row right by luck happens about 12 per cent of the time. A step is only cleared when every try at it is right.

That is why this run takes longer than the usual sliding tone. The extra minute buys the one thing that makes the number worth reading.

The headline is the highest step where every try was right. That is the conservative figure and it is the one to quote.

The second figure is the highest step where you got anything right at all. It is usually one rung higher, and the gap between the two is the interesting part. A tone at the edge of hearing is not simply absent. It fades into something you catch about half the time, and the two numbers put a bracket around that band.

Steps cleared is the count behind the headline. It moves when you change the starting point, so two runs are only comparable when they started in the same place.

Hearing at the top of the range declines with age in almost everyone, gradually and from early adulthood. It is the most reliable pattern in the whole of audiology. It is why a teenager and their parent get different numbers on the same speakers.

The inside of the ear works a little like a piano laid out along a coiled tube. High notes are picked up at the near end and low notes deep inside. The cells that handle the top of the range are the ones every loud sound reaches first.

None of that makes the figure here a measure of damage or of age. Age moves a great many measured numbers, and reaction time by age shows what that looks like when the measurement is done properly. Nothing on this page is done properly enough for that comparison.

There is one genuinely useful thing to do with it. Run it twice on the same headphones at the same volume, once now and once in a year. Even that is a rough comparison of the whole chain rather than of your ears.

The frequency of the tone is exact. Everything about how loud it arrives is not, and loudness is what decides whether a tone at the edge is heard.

There is no way to ask your operating system how loud its output is. There is no way to know what your headphones do above 16 kHz. There is no way to hear the room you are sitting in. The volume setting on this page moves the number, which is the clearest possible demonstration of the problem.

Turning it up to chase a higher figure is missing the point twice over. The number is not a score to beat, and a louder tone at the top of the range is not a kindness to your ears.

One thing here is clean, and it is worth saying. The pitch of a tone survives any playback chain intact, which is why the relative pitch test can be trusted where this one cannot. Loudness is the part that does not survive.

It is not a hearing test and it is not a screening tool. It cannot detect hearing loss, it cannot rule it out, and a number you dislike is not a finding.

Most of everyday hearing happens far below the frequencies this page explores. Speech lives between roughly 250 Hz and 6 kHz, so someone who clears nothing above 12 kHz here may hear conversation perfectly well.

Practice will not move this number, and there is nothing to train. If you want the version of this argument that applies to the tests that can be trained, the piece on reaction time makes it.