Start a project →

Free Audio Sync Offset Finder: line up your mic audio with your video

Choose your camera video and the separate mic or recorder audio. The tool finds how far apart they are and tells you exactly how much to move the mic audio, in milliseconds and in frames, with a confidence score. It also measures the drift between the two clocks and gives you the mic audio already lined up as a WAV you can drop at the start of your timeline. Everything runs in your browser: your files never leave your device.

Updated · Built by Kevin Tabares, long-form YouTube editor at Umbrella Creators.

Find the offset between a video and a mic recording

Your files stay on your device. Nothing is uploaded: your browser decodes and compares both recordings right here, in this tab.

How to use it

  1. Choose the video: the camera file with its own sound, even if that sound is poor. Any file with audio works as the reference.
  2. Choose the mic or recorder audio: WAV, MP3, M4A or even a second video. Up to 60 minutes each.
  3. Press Find the offset. Read how far to move the mic audio, check it by ear in the preview and download the lined-up WAV.

Both recordings only need to hear the same sound for a few seconds: speech works best, and a clap at the start helps but is not needed. If the camera file has no sound at all, there is nothing to match, so record with the camera mic on even when you use a separate mic.

Long files: from WAV files, and from MP4, MOV and MKV files with AAC or PCM sound (what cameras and recorders write), the tool reads only the audio, in small pieces straight from your disk. Comparing two 60 minute files that way used about 0.6 to 0.7 GB of memory in my tests in Chrome, even when one of them was a 60 minute video. Other formats (MP3, FLAC, OGG, WebM) are decoded whole by the browser first, and a 60 minute MP3 needed about 3.3 GB at that moment. Saving a 60 minute WAV needs about 1.6 GB more while it is written. If a long file fails on a laptop with little memory, export its audio as a WAV and use that.

Step-by-step guide

For the whole process around this tool, read How to sync audio from a separate mic with your video: clap and waveform sync in Premiere Pro and DaVinci Resolve, and how to fix drift on long recordings.

FAQ

How does this audio sync finder work?
Your browser decodes both files, mixes each one to mono and compares how their loudness rises and falls in three frequency bands, at every possible offset, to find where they match. Then it checks up to 15 short windows spread across the recording and lines up the actual waveforms with a phase correlation (GCC-PHAT), which compares timing rather than tone. That is why a bright lav against a dull camera mic, different levels, background noise or a 44.1 kHz file against a 48 kHz one still line up to a fraction of a millisecond.
What is audio drift and why does my mic go out of sync over time?
Every camera and recorder counts time with its own clock, and no two clocks run at exactly the same speed. The difference is tiny but it adds up: a mic file 0.01% fast is 6 ms ahead of the video after one minute and 0.36 s ahead after an hour. The tool measures the offset at the start, middle and end of the overlap, gives you the drift in milliseconds per minute and the clip speed that cancels it, and the download can correct it for you by resampling the mic to the video's timing.
How accurate is it?
On my test files the offset came out within 0.005 ms of the true value on recordings of 2 to 13 minutes, and within 0.05 ms on 60 minute pairs: real narration recorded through a different EQ, pink noise 10 dB under the voice, a 44.1 kHz mic against a 48 kHz video and offsets that fall between samples. The camera files were MP4, MOV, MKV and WebM, some of them with the audio track starting later than the picture, which the tool reads from the file itself. A mic made 0.01% fast measured 5.999 ms per minute of drift (true value 5.999). The tool also tells you when the mic is in opposite polarity, and when the offset jumps inside the file because a recorder lost a few milliseconds of sound, instead of hiding it. Keep three limits in mind. The tool lines up the sound as each device heard it, and sound travels about 34 cm per millisecond, so a camera 3 m away hears you about 9 ms after a mic on your collar. In a very echoey room the match can lock onto an echo: the checks then agree less, the confidence drops to medium and the preview lets you check it by ear. And HE-AAC audio, rare in recordings, is left to the browser's own decoder, which placed it about 20 ms late in my Chrome test.
What if the two files do not share any sound?
Then there is nothing to match, and the tool tells you so instead of guessing: the confidence stays under 55 and no offset is shown. The usual causes are a video exported without its camera sound, two takes that were not recorded at the same time, or the wrong file. The camera's own mic is enough, even if it sounds bad.
Does it upload my files?
No. Your browser decodes both files and compares them on your own device, and the lined-up WAV is written right here in this tab. Nothing leaves your computer.
How do I use the lined-up WAV in my editor?
Put the video at the start of your timeline and the WAV at exactly the same start, then mute or delete the camera audio. The WAV is as long as the video, with silence where the mic was not recording. It keeps the sample rate of your mic file if that is a WAV, and it is 48 kHz for other formats.

Who made this

I'm Kevin Tabares. I've edited long-form YouTube videos since 2015 for channels like Rex & Alexa, Puff and Mud. When a creator sends a camera file and a separate mic recording, lining them up is the first job, so I built this to give the exact offset and drift in seconds, without uploading anything. If you want an editor for your channel, get in touch.

More free tools

A loudness normalizer to bring the voice to -14 LUFS, a background noise checker for hiss and hum and an audio clipping detector, all free and without signup.

See all tools →