Audio Latency Test

Clicks out, microphone listens in: an estimated round-trip delay for your output + input chain.

Ready

Ready

Turn your system volume down first

Start at a low volume, especially with headphones and low bass tones, then raise it slowly.

For a better estimate: use speakers + the built-in mic in a quiet room with other audio apps closed. Wired chains usually measure tens of ms; Bluetooth often 100–300 ms. This is a rough whole-chain estimate affected by browser and OS buffers — not lab data, and it cannot isolate a single device.

This tool runs locally in your browser — no audio is uploaded.

How to use

  1. 1

    Set up the loop

    Use the built-in microphone and play through speakers if possible — that is the most reliable loop. With headphones, let the click leak to the mic or place it near the earcup; Bluetooth adds its own delay.

  2. 2

    Run the measurement

    Press Measure. The page plays several short clicks and timestamps when each one comes back through the microphone; the median of the detected round-trips is shown.

  3. 3

    Read it as an estimate

    Expect 30–150 ms with Bluetooth headphones, and tens of milliseconds on wired chains. Compare setups relatively; the number is not a lab measurement of any single device.

How it works & limits

Each measurement schedules a sharp 4 kHz click at a known AudioContext time and watches the incoming microphone stream for the level spike it causes. The gap is the round trip: output buffer + DAC + (wireless link, if any) + air + ADC + input buffer. Repeating clicks and taking the median softens occasional misses.

Honest limits: browser timers, buffer sizes, and OS mixing all shift the result by tens of milliseconds, and a missed detection (quiet room, quiet speakers, muted mic) is reported as a failure rather than a fake number. Bluetooth codecs buffer heavily and vary with firmware — treat repeated runs and relative comparisons as the meaningful signal, not the absolute value.

FAQ

Why is Bluetooth latency so high?

Bluetooth codecs buffer audio in frames for robustness; typical round trips land between 100 and 300 ms depending on codec (SBC/AAC/aptX) and firmware. That is physics and product design, not something a website can fix.

Why did the test report "no click detected"?

The microphone never heard the click loudly enough: mic muted, volume too low, or headphones sealing the sound in. Raise the volume, use speakers, or move the mic closer, then retry.

What is a good latency number?

Wired chains usually measure somewhere in the tens of milliseconds; under about 50 ms feels immediate for music. For rhythm games or instruments, lower is better — Bluetooth is rarely suitable.

Why do results change between runs?

OS buffers, CPU load, and Bluetooth conditions vary run to run. The median over several clicks helps, but only big differences (say, 40 ms vs 180 ms) are meaningful.

Does this measure my sound card exactly?

No. It measures the whole chain in both directions, including the browser. Use it to compare devices and setups, not as a hardware specification.

Related tools

Mic Test

Live level meter, device picker, and local record & playback. Nothing uploaded.

Open tool

Spectrum Analyzer

Live FFT bars, waveform, and spectrogram from your mic — all computed locally.

Open tool

Headphone Test

Check left/right channels, smooth panning, and balance on any headphones or earphones.

Open tool