How to sync audio and video recorded separately on a Mac

Why are my audio and video not matching?

Usually for one of two reasons, and they need different fixes:

An offset. The camera and the recorder were started at different moments, so the sound is early or late by the same amount all the way through.
Drift. No two devices' clocks run at exactly the same speed. A difference of 100 parts per million sounds tiny, but over an hour it adds up to 0.36 seconds, eleven frames, so a take that starts in sync ends with lips and words visibly apart.

An offset is fixed by sliding the sound. Drift is fixed only by playing one recording very slightly faster or slower so the two clocks agree. If the gap is the same at the start and the end of a take, it is an offset; if it grows, it is drift.

How do you record audio and video separately?

Record the voice on a recorder or a microphone close to the speaker, keep the camera's own microphone on as a guide, and give both a shared sound to line up on. In practice:

Leave the camera's sound on. Its scratch track is what you sync external audio against later.
Match the settings. Record at 48 kHz, the rate video uses, on both devices.
Mark the start. Clap once, sharply, in view of the camera and in range of both microphones. A slate does the same with a clapper.
Jam timecode if you can. If both the camera and the recorder take timecode, set them to the same time before you roll.
Record a little room tone. Thirty seconds of the room with nobody speaking fills the gaps your cuts leave later.
Keep takes sensible. On very long recordings, a second clap at the end makes drift easy to see.

A man in a red shirt brings his hands together for a single clap in bright sunlight, his face out of frame. — One sharp clap gives the camera and the recorder the same moment to line up on: a spike in both waveforms and a frame where the hands meet.Photo by Lisa Fotios on Pexels

Should you sync by clap, by waveform or by timecode?

Method	How it works	Good for	Watch out for
Clap or slate	Line up the spike in both waveforms with the frame where the hands meet	Short takes, any gear	Fixes the offset only; drift remains on long takes
Waveform	Software compares the recorder's sound with the camera's scratch track	Interviews, podcasts, a second camera	Needs the camera's own sound, and enough of it in common
Timecode	Each file is stamped with its start time and placed by it	Shoots with jammed recorders and cameras	Phones and many cameras write no timecode; a recorder jammed wrongly lands in the wrong place

In practice, timecode gets you close and sound gets you exact.

How do you sync audio and video in Sharpy?

Put the camera clip and the recorder's file in the edit, then:

Choose Synchronize by Sound. Sharpy lines up the recorder's sound with the camera's by what they both heard, the way you would by eye on a clap, but across the whole recording. Or ask the agent to sync the two clips (edit_clip).
Let it take out the drift. Sharpy measures how far the two clocks differ and, where that would move the end of the clip noticeably, plays the recorder at its own clock's speed, with pitch following. At a typical drift of 300 parts per million, that shifts pitch by half a cent, a tenth of what an ear can tell apart.
Check what moved. The clip moves with everything linked to it, to the sample, and is linked to the camera clip so they stay together when you cut.
Decide about the scratch track. The camera's own sound stays on until you ask for it to be switched off, and Sharpy tells you it is still on.

If two recordings do not share enough sound to be sure they are the same moment, Sharpy refuses them as "not one sound" rather than guessing.

Tested on a meeting, a headset recording against a microphone across the room, every true pair was found within 0.07 to 1.2 ms beyond the far microphone's own acoustic delay, drift was measured within 0.3 to 2.2 parts per million, and every unrelated pair was refused. On the base Mac mini M4, two 21-minute recordings sync in 7.5 seconds. It handles drift of up to 500 parts per million.

The same finder lines up a second camera angle by what both microphones heard, so a two-camera interview or a video podcast cuts between angles with the mouth still moving with the words.

How does sync by timecode work in Sharpy?

Choose Clip ▸ Synchronize by Timecode. Sharpy reads the camera file's start timecode and the time stamp a Broadcast Wave recorder writes into its file, and places the clip where the two say, across midnight the shorter way. Where both files carry sound, sound then refines the position to the sample, with the drift taken out, as long as it lands within a frame of the timecode. If sound would move it further, the timecode position is kept and the gap is reported, so a recorder jammed to the wrong time is seen instead of silently moved.

Phone recordings carry no timecode, and Sharpy says so rather than pretending; sync those by sound.

How to fix audio that is out of sync with video?

Find out whether it is an offset or drift, then slide or retime the sound. Play the start and the end of the clip and watch a hard consonant, a p or a b, against the lips:

The same gap at both ends: an offset. Slide the sound by that amount. In Sharpy, Synchronize by Sound measures the offset for you to the sample.
A gap that grows: drift. The sound has to play very slightly faster or slower. Synchronize by Sound measures it and corrects it.
Fine in the editor, wrong after export: check the exported file in a second player before you blame the edit. Sharpy writes every cleaned voice file in time with its recording, and its window plays the same mix the render writes.

Once the sound is in place, removing background noise and audio mixing for video take it the rest of the way.

Frequently asked questions

Do I need a clapperboard to sync audio?

No. A single sharp hand clap does the job, and syncing by waveform needs no clap at all, only the camera's own sound.

What is audio drift?

The slow slide out of sync caused by two devices' clocks running at slightly different speeds. At 100 parts per million, an hour drifts by 0.36 seconds, eleven frames.

Can I sync two cameras as well as a recorder?

Yes. Sharpy lines up a second camera by what both microphones heard, the same way it lines up a recorder.

Does syncing change the pitch of the voice?

Only by the drift's tiny amount. At 300 parts per million, the change is half a cent, far below what anyone can hear.

Does Sharpy need timecode to sync?

No. Synchronize by Sound works on any two recordings of the same moment; timecode is used when your files carry it.

Sources

Checked on 18 October 2026.

EBU Tech 3285: BWF, a format for audio data files in broadcasting (Broadcast Wave Format)