Audio

Audio Latency Test

Browser audio is never instant — there's a delay between scheduling a sound and hearing it. This tool reports the latency the browser admits to (base and output, in milliseconds), then runs a tap-along test that includes the full chain: scheduling, output buffering, the audio device itself, and the speaker. Useful for music apps, games, video calls, and rhythm games.

Browser-reported latency

Values exposed by the AudioContext. baseLatency is the minimum the browser claims; outputLatency includes the device pipeline.

Base

—

Output

—

Sample rate

—

Tap-along test

Five clicks will play, 1.5 seconds apart. Tap the button or press the spacebar at the exact moment you hear each one.

Why audio latency matters

Every step between your code and your ears adds delay: the OS audio buffer, the browser's internal scheduling, the audio driver, the DAC in your sound card, and any wireless or processing layer (Bluetooth, AirPods, surround receivers). For most listening this is invisible — a few extra milliseconds make no difference to a podcast. But for anything that requires synchronisation with action — playing music software, gaming, video calls, rhythm games, live performance — latency dictates whether the experience feels responsive or sluggish. Anything under 30 ms feels instant; 30–80 ms is acceptable for casual use; over 100 ms starts to feel disconnected; over 200 ms is genuinely problematic for live interaction. This tool gives you two views of your system: what the browser says, and what your ears actually receive.

How to run a latency test

1
Read the reported latencies
The first panel shows what the browser exposes. baseLatency is the minimum, set by the browser's internal buffer. outputLatency adds the OS and device pipeline. Bluetooth headphones and AirPods will report (or imply) much higher numbers than wired headphones.
2
Press start
Five evenly-spaced clicks will play, 1.5 seconds apart. Get ready — the first one fires half a second after you click.
3
Tap on the beat
When you hear a click, tap the big button — or press the spacebar — as fast as you can. Aim for the click itself, not where you think it should be. The tool times your tap against the moment the click was scheduled.
4
Read your results
Each row shows your delta (how late your tap was, in milliseconds). The summary shows median, mean, and standard deviation across the 5 clicks.
5
Subtract reaction time
Auditory reaction time is typically 200–250 ms. The portion of your delta beyond that is approximately your audio output latency — the delay between the speaker producing the sound and your awareness of it.

How latency adds up

Total perceived latency = 
    audio buffer (browser)  +  output device pipeline
  + speaker / driver  +  air travel  +  ear → brain reaction
 
Wired headphones, USB DACs:    typically  10–30 ms
Built-in laptop speakers:      typically  20–50 ms
Bluetooth audio (modern):      typically 100–200 ms
Bluetooth (low-latency codec): typically  40– 80 ms
 
Tap delta = scheduledTime - tapTime
          ≈ audioPipeline + airTravel + reactionTime
 
Reaction time alone is roughly:
    200–250 ms (auditory, normal alertness)

AudioContext.baseLatency is what the browser controls — usually a single audio buffer worth of samples (about 5–25 ms depending on the platform). AudioContext.outputLatency adds the OS audio pipeline. Neither of those values capture what happens between the digital-to-analog converter and your eardrum, which is where the biggest variable — your headphones or speakers — lives. The tap test measures the entire chain at once but mixes in your reaction time. By comparing the two, you can isolate whether latency lives in the browser or in your hardware.

Reference: MDN — AudioContext.outputLatency

Interpreting your results

Median tap delta	What it suggests
200–250 ms	Pure reaction time; near-zero output delay. Wired or USB audio with a fast browser pipeline.
260–320 ms	30–70 ms of audio latency on top of reaction. Typical for a modern laptop with built-in speakers.
330–450 ms	80–200 ms of audio latency — Bluetooth or buffered output. Most Bluetooth headphones; fine for music, slow for live use.
Above 500 ms	Substantial audio delay or you missed a click. Try the test in a different browser or with wired audio.

Common use cases

Decide if your headphones are responsive enough for music software
Compare wired vs. Bluetooth audio latency on the same device
Diagnose feeling 'out of sync' on video calls or livestreams
Test a rhythm or music game's calibration before playing
Verify a USB DAC's claimed low-latency mode is actually low-latency
Spot-check that a browser update hasn't regressed audio scheduling
Demonstrate the perceptual difference between 30 ms and 200 ms
Calibrate audio offset settings in a digital audio workstation

Related tools

Hearing Range Test

Find your high-frequency hearing cutoff with a 20 Hz - 22 kHz sine sweep.

Tone Generator

Play sine, square, sawtooth, or triangle waves at any frequency from 20 Hz to 20 kHz.

Stereo Channel Test

Verify left, right, and both channels are wired and balanced correctly.

Microphone Test

Live input meter with peak, RMS, and clip detection — no recording.

Frequently asked questions

Why does my Bluetooth headphone show much higher latency?

Bluetooth audio has to encode, transmit, decode, buffer, and reassemble the signal — each step costs time. Standard SBC and AAC profiles add 100–250 ms; aptX and aptX Low Latency reduce this significantly but require both the source device and the headphones to support the codec. AirPods are typically in the 150–200 ms range over standard Bluetooth, with lower latency only when paired with Apple devices that negotiate proprietary modes.

Why does the browser-reported latency seem too low?

AudioContext only knows about the parts under its control: its own buffer (baseLatency) and what the OS reports about the output device (outputLatency). It doesn't know how long the signal will spend travelling over Bluetooth, sitting in your AV receiver's processing chain, or being buffered by your TV. The tap test exposes all of that — which is why the numbers usually disagree.

How do I subtract my reaction time?

Average auditory reaction time for an alert person is 200–250 ms. If your tap delta median is, say, 280 ms, the rough math is 280 − 230 ≈ 50 ms of audio output latency. For a more accurate baseline, run a visual reaction-time test separately and use your own personal number (it varies between people and across the day).

Why do my taps vary by 50+ ms?

That's reaction-time variance, which is normal. Even within a single session, individual tap reactions can be 50–80 ms different. The median across 5 clicks is more stable than any single tap. For a tighter result, run the test 2–3 times and average the medians, or take more clicks at a time.

Does the test browser-shop for the lowest latency?

No — it uses the audio context the page creates. To compare browsers, run the tool in each one. Chrome and Firefox tend to be similar on Windows; Safari on Mac is sometimes lower for Apple devices. Mobile browsers usually have higher latency than desktop, especially over Bluetooth.

Is anything sent to a server?

No. All scheduling, timing, and statistics happen entirely in your browser using AudioContext and performance.now(). No microphone access is requested, no audio is recorded, and nothing is uploaded.

Can this measure microphone (input) latency too?

Not on its own. Measuring input latency requires playing a known tone, recording it through the microphone, and timing the round-trip — which needs microphone access and a way to physically connect output to input (like sitting next to a speaker). That's a separate test and not currently part of this tool.

Is 50 ms of latency a problem?

Depends on the use case. For passive listening or video, you'd never notice. For real-time interaction — playing a virtual instrument, a rhythm game with tight timing windows, monitoring your own voice while singing — 50 ms is at the edge of perceptible. Above 100 ms, most people notice. Above 200 ms, almost everyone does, and live performance becomes very difficult.

Last updated 2026-04-25

Audio Latency Test

Why audio latency matters

How to run a latency test

Read the reported latencies

The first panel shows what the browser exposes. baseLatency is the minimum, set by the browser's internal buffer. outputLatency adds the OS and device pipeline. Bluetooth headphones and AirPods will report (or imply) much higher numbers than wired headphones.

Press start

Five evenly-spaced clicks will play, 1.5 seconds apart. Get ready — the first one fires half a second after you click.

Tap on the beat

When you hear a click, tap the big button — or press the spacebar — as fast as you can. Aim for the click itself, not where you think it should be. The tool times your tap against the moment the click was scheduled.

Read your results

Each row shows your delta (how late your tap was, in milliseconds). The summary shows median, mean, and standard deviation across the 5 clicks.

Subtract reaction time

Auditory reaction time is typically 200–250 ms. The portion of your delta beyond that is approximately your audio output latency — the delay between the speaker producing the sound and your awareness of it.

How latency adds up

Total perceived latency = 
    audio buffer (browser)  +  output device pipeline
  + speaker / driver  +  air travel  +  ear → brain reaction
 
Wired headphones, USB DACs:    typically  10–30 ms
Built-in laptop speakers:      typically  20–50 ms
Bluetooth audio (modern):      typically 100–200 ms
Bluetooth (low-latency codec): typically  40– 80 ms
 
Tap delta = scheduledTime - tapTime
          ≈ audioPipeline + airTravel + reactionTime
 
Reaction time alone is roughly:
    200–250 ms (auditory, normal alertness)

Interpreting your results

Median tap delta	What it suggests
200–250 ms	Pure reaction time; near-zero output delay. Wired or USB audio with a fast browser pipeline.
260–320 ms	30–70 ms of audio latency on top of reaction. Typical for a modern laptop with built-in speakers.
330–450 ms	80–200 ms of audio latency — Bluetooth or buffered output. Most Bluetooth headphones; fine for music, slow for live use.
Above 500 ms	Substantial audio delay or you missed a click. Try the test in a different browser or with wired audio.

Common use cases

Decide if your headphones are responsive enough for music software

Compare wired vs. Bluetooth audio latency on the same device

Diagnose feeling 'out of sync' on video calls or livestreams

Test a rhythm or music game's calibration before playing

Verify a USB DAC's claimed low-latency mode is actually low-latency

Spot-check that a browser update hasn't regressed audio scheduling

Demonstrate the perceptual difference between 30 ms and 200 ms

Calibrate audio offset settings in a digital audio workstation