How to measure millisecond audio-video delay on a webcam

Try It Yourself

Measure the offset without manual frame-stepping

Produces a synchronised reference and reports the gap in milliseconds directly, which saves scrubbing through a recording by hand.

Run the Audio/Video Sync Test

Audio and video take separate paths through your hardware and software before reaching the screen — different USB controllers, different driver stacks, different buffering.

There is no way to reliably estimate the gap without measuring it directly, and an offset applied by guesswork usually overshoots in the opposite direction.

Record yourself clapping once, clearly, in frame. Then step through the recording to find the exact moment your hands visually meet, and check the audio waveform for the moment the clap peaks.

Skip the manual frame-stepping with the Audio/Video Sync Test, which reports the gap directly.

1. Clap once, sharply, well inside the frame and close to the microphone.
2. Step frame by frame to the frame where your hands first touch, and note the timestamp.
3. Find the audio peak in the waveform and note that timestamp.
4. Subtract one from the other — at 30fps each frame is roughly 33 milliseconds.

Broadly, an offset under roughly 30 milliseconds is imperceptible to most viewers. Between 30 and 100 milliseconds it is noticeable to some but not distracting to most.

Beyond about 100 milliseconds it reads as clearly wrong and becomes distracting, which is the point at which correcting it is worth the effort.

Do not assume the direction. Depending on your camera and microphone processing pipelines, either audio can arrive late or, less commonly, video can be the side lagging behind.

Getting the direction wrong means applying a correction that doubles the problem rather than removing it, so establish it before adjusting anything.

If a capture card or DSLR is in the chain, measure that stage with the Capture Card & DSLR Latency Auditor.

If you normally use a USB microphone with a webcam, or a Bluetooth headset, measure with that exact combination.

Swapping to a different microphone for testing tells you nothing about the setup you actually use, because the latency belongs to the specific devices in the chain.

A driver update, a different USB port, or a new microphone can all change this number.

It is worth re-checking periodically rather than assuming a value measured once still holds, particularly after a system update.

To check whether the microphone is adding processing delay, run the Advanced Microphone & Room Acoustics Test.

Audio Mixer → gear → Advanced Audio Properties → Sync Offset (ms)

Enter the measured figure directly against the relevant audio source; the correction applies at capture.

Timeline → unlink audio, then nudge the track

Most editors let you enter an offset numerically rather than dragging, which is more precise and repeatable.

No manual sync offset

The only levers are removing Bluetooth audio and reducing video encode load.

No manual sync offset

Switching to a wired headset is the practical fix, since no correction control exists.

1. Set up with the exact camera, microphone and connection you normally use.
2. Record a single sharp clap well inside the frame and close to the microphone.
3. Step frame by frame to the frame where your hands first touch and note it.
4. Find the audio peak in the waveform and note that timestamp.
5. Subtract the two and convert frames to milliseconds using your recording frame rate.
6. Note the direction — whether audio or video is the late one.
7. Repeat the measurement two or three times and take the median.
8. Re-measure after any driver update, port change or new microphone.