Headphone lab · audio delay

Audio Latency Test

Measures how late sound reaches you: by tapping to a beat and a flash, or by looping a click through a microphone.

Your sound output Not started Use the headphones or speakers you want to measure, at a comfortable volume. Wireless headphones: connect them first, then press Start sound.
Browser’s own estimate
—
Sample rate
—

Twenty clicks at a steady 100 per minute. The first four are a count-in; tap on each of the next sixteen — on the pad, or with the space bar — as you would in a rhythm game. Watch nothing but your finger: the pad stays dark on purpose.

Your taps around the beat

What we measured

Sound, tapping to the beat
—
Picture, tapping to the flash
—
Sound after the picture
—
Round trip, microphone loopback
—
  • Run a test; its reading appears here.

Why tapping twice

Nobody taps exactly on a beat: most people land a little early on a steady rhythm, some late, and each person is consistent. A single tap test mixes that habit into the result. Tapping to a flash carries the same habit with no sound in it, so subtracting the two leaves only how much later the sound arrives than the picture — the number that matters for lip sync and for hearing a shot when you see it.

Next checks

The same headphones, other questions.

Words on this page
Audio latency

The time from a program making a sound to it reaching your ears: the audio buffer, the connection (a few milliseconds by cable, 100 to 300 ms over ordinary Bluetooth) and the headphones.

Quick answers before you test

How much delay do Bluetooth headphones add?

Usually 100 to 300 ms with the standard codecs, 40 to 80 ms with low-latency ones; wired and USB headphones a few to 30 ms. The loopback measures yours.

Why does a video stay in sync but a game does not?

Video players know the delay and hold the picture back to match. A game cannot wait: you hear the shot after you see it.

What should I set as my rhythm game offset?

The beat test’s median is what a calibration screen measures. Enter it in the game’s audio offset; some games count it as negative.

Why tap to a flash as well?

Everyone taps a little early or late on a steady beat. The flash test carries the same habit without sound, so the difference is how much later the sound arrives than the picture.

What the result actually measures

The beat test schedules twenty clicks on the audio clock; after a four-click count-in you tap on each, and every tap is carried onto the same clock, so (tap − click) holds the whole path from rendering the click to your ear, plus your timing habit and the input delay. The flash test does the same with flashes drawn on animation frames, which holds the display path plus the same habit. Their difference is how much later sound reaches you than the picture, judged against the broadcast thresholds (ITU-R BT.1359: noticed beyond 45 ms early or 125 ms late). The loopback plays five clicks while an AudioWorklet records the microphone from the exact frame each click starts, and cross-correlation finds the click to the sample: output plus input delay, with no human in it. The browser’s own estimate of its output delay is shown alongside.

Measurement boundary: Tap tests measure you as well as the hardware; sixteen taps give a median within a few milliseconds for most people, but a distracted run spreads wide — repeat it. The loopback includes the microphone’s own delay, usually 5 to 20 ms on built-in microphones. Windows often reports only its buffer as the browser’s estimate, not the Bluetooth link, which is why the measurements exist.

How to interpret the output

Sound, tapping to the beat

Your sound path plus your timing habit: the number a rhythm game calibrates. With wired headphones most people land within 50 ms of the click.

Sound after the picture

Beat minus flash. Up to 125 ms late and 45 ms early goes unnoticed; past 185 ms late it is clearly out of step.

Round trip, microphone loopback

Output plus input delay, measured to the sample. Under 45 ms is wired; 45 to 90 ms a low-latency wireless link; 100 to 250 ms is ordinary Bluetooth.

Troubleshooting unexpected results

The loopback says not heard

Turn the volume up, hold the microphone right against the speaker or put the earcup over it, and turn off noise suppression in your headset’s software, which can swallow short clicks.

My taps spread wide

Tap with one finger on a hard surface or the space bar, look away from the screen in the beat test, and repeat; the median of a calm run is the number to keep.

Bluetooth delay is very high

Make sure the headphones are not using the hands-free profile (a microphone in use switches to it), keep the computer within a few metres, and update the Bluetooth driver; for games, a cable or 2.4 GHz dongle removes the problem.

Frequently asked questions

How do I test audio latency?

Press Start sound, tap along to twenty clicks, then to twenty flashes. The page shows your sound delay, your picture delay and the gap between them; the microphone loopback gives a number with no human timing in it.

How do I test Bluetooth headphone latency?

Connect them, run the microphone loopback with an earcup held over the microphone, and compare with a wired pair: the difference is what the Bluetooth link adds.

Is 200 ms of audio delay bad?

For music and video, no — players hide it. In games, yes: you hear footsteps and shots a fifth of a second after they happen. Use a cable or a low-latency dongle for competitive play.

What is a good audio latency for gaming?

Under 45 ms from click to ear goes unnoticed; wired and USB headsets sit there. Low-latency wireless reaches 40 to 80 ms.

Is this the same as input lag?

No. Input lag is from your hand to the screen; this is from the computer to your ears. The input latency test measures the other one.