To sync audio and video recorded on separate devices, keep the camera’s own sound, put both recordings on the timeline, and line them up by their waveforms. Either let the editor match them (Synchronize in Adobe Premiere, Auto Sync Audio in DaVinci Resolve, Synchronize Clips in Final Cut Pro, AI multicam sync in ChatCut) or slide the recorder until a clap lines up. Then check a moment near the end. If it still matches, mute the camera’s sound and carry on. If the gap has grown, the two devices drift apart, and the recorder audio needs a small speed change.
This guide is about recordings of the same take: a camera and a separate microphone recorder, or two cameras, filmed at the same time. It covers why sync fails, how to tell the three kinds of problem apart, exact steps in five editors, and how to fix drift. It doesn’t cover generating lip-sync for new audio or cutting to a beat. Menu paths were checked against each vendor’s documentation on November 4, 2026.
How to sync audio and video
- Keep the camera’s scratch audio. It’s the reference the recorder is matched against, even if it sounds bad.
- Put the camera clip and the recorder file on separate tracks.
- Align them with your editor’s automatic sync, or by lining up a clap or another sharp sound.
- Check the end. Play a word or a clap near the end of the take. If it’s out, you have drift; fix it before going further.
- Mute the camera’s sound, but keep it on the timeline until the edit is finished.
- Export and watch the beginning, middle and end of the file in a different player.
Why recordings fall out of sync
Knowing the cause tells you which fix to use.
- They started at different moments. A camera and a recorder never start at exactly the same instant, so every pair has an offset. This is the normal case, and it’s the easy one: move the audio once.
- Each device keeps its own time. A recorder set to 48 kHz and a camera set to 48 kHz are each counting with their own clock, and those clocks differ very slightly. Audacity’s manual describes the result: a track recorded at a different speed progressively drifts apart, so moving it can never line it up along its whole length (Audacity). A tiny difference is invisible in a 30-second clip and obvious in a 20-minute interview.
- Phones record variable frame rate video. To save space, phones and screen recorders often let the frame rate change during a recording. Some editors play this footage with sound slowly or suddenly out of step. Adobe’s own community FAQ on variable frame rate recommends converting such files to a constant frame rate when its workaround doesn’t hold (Adobe community FAQ).
- A recording stopped and restarted, or dropped frames. Then the sync is right until one moment and wrong after it.
- The player adds a delay. A TV, soundbar or Bluetooth headphones can delay the sound on their own, while the file itself is fine.
One cause that’s often blamed is not real on its own: a 30 fps video and 48 kHz audio don’t conflict. Frame rate and sample rate measure different things, and editors play them together all the time. What matters is whether the two devices’ clocks agree.
Tell offset, drift and playback delay apart
Before you fix anything, measure the gap twice: once at a clap or sharp sound near the start and once near the end.

How close is close enough? The broadcast standard ITU-R BT.1359 reports that viewers on average start to notice when sound runs about 45 milliseconds ahead of the picture or about 125 milliseconds behind it (ITU). At 30 frames per second, one frame lasts 33 milliseconds, so aim to be within a frame and be stricter about sound arriving early than late.
Line up the clap by hand
This works in any editor with a multi-track timeline, and it’s how you check what automatic sync did.
- Stack the recordings. Camera clip on the top tracks, recorder file on an audio track below.
- Find the clap in both waveforms. A hand clap or a slate is a tall, narrow spike that stands out from speech. Zoom the timeline in until you can see its shape.
- Slide the recorder until the two spikes line up. Match where the spike starts, not its loudest point.
- Fine-tune by ear. Play a sentence with both tracks audible. Close but not exact sounds hollow or echoey; aligned sounds like one voice. Nudge the recorder a frame at a time until the echo disappears.
- Check the picture. Step frame by frame to the moment the hands meet. The spike should start on that frame or the next one.

No clap? Use any short, sharp sound both devices heard: a door, a cough, a hard “p” or “t” at the start of a word. If the camera recorded no sound at all, match a visible action instead, such as hands meeting, a cup put down or the first frame of a slate app’s flash.
Sync automatically in your editor
Automatic sync compares the two waveforms and finds where they match. It needs the camera’s own audio, and it works best when both recordings contain the same speech or a clear clap.
Adobe Premiere
In Adobe Premiere (called Premiere Pro until early 2026), sync on the timeline (Adobe):
- Put the camera clip and the recorder file on the timeline on separate tracks.
- Select both and choose Clip → Synchronize.
- Choose audio as the sync point and pick the audio track channel that holds the camera’s reference sound.
- Select OK, then play the start and the end to check.
Clip → Merge Clips in the Project panel combines one video clip with up to 16 audio channels into a new merged clip without changing the source files (Adobe). Adobe’s current help lists In points, Out points, timecode and clip markers as its sync points, so it suits slated shoots (Adobe). If you shot on a phone and the sync wanders, see the variable frame rate note under fixing drift.
DaVinci Resolve
Resolve syncs clips in the Media Pool before you edit. These steps follow Blackmagic’s DaVinci Resolve 21.1 reference manual (chapter 20, “Syncing Audio and Video”) (Blackmagic):
- Put the camera clips and recorder files that belong together in one bin, or select both bins.
- Right-click the bin and choose Auto Sync Audio.
- Set Synchronize Using to Waveform.
- Leave Use channel number on Auto, or pick a cleaner channel if one is noisy.
- Turn on Retain embedded audio to keep the camera’s sound on its own track. Without it, the synced recorder audio replaces the camera’s audio channels.
- Click Sync.
If Resolve reports that some clips couldn’t be synced, select just that camera clip and its recorder file and try again. Without camera sound, mark the clap with an In point, Out point or marker on both clips and choose that option under Synchronize Using instead.
Final Cut Pro
Final Cut Pro for Mac creates a new synchronized clip and leaves the originals untouched (Apple):
- In the browser, select the camera clip and the recorder clip.
- Choose Clip → Synchronize Clips (Option-Command-G).
- Select Use audio for synchronization.
- Select Disable audio components on AV clips to mute the camera’s scratch track in the new clip.
- Click OK.
To nudge the result, select the synced clip, choose Clip → Open Clip, and drag the recorder clip left or right. Timecode sync is faster and frame-accurate if both devices recorded matching timecode.
CapCut
CapCut’s desktop app has a one-step sync, described in CapCut’s own guide (CapCut): import both files, drag them to the timeline, select both with Ctrl-click (Windows) or Cmd-click (Mac), right-click and choose Sync video and audio. CapCut doesn’t publish which versions or regions include it, so if the command is missing, line up the clap by hand. To fix a clip whose own sound is out, right-click it and choose Extract audio, then move the audio (CapCut).
ChatCut
ChatCut’s sync command is called AI multicam sync, and it works for one camera plus one recorder as well as for several cameras (Timeline docs). In ChatCut Desktop, import with My Assets → Upload → Files… or by dragging files in from Finder or File Explorer; the desktop app reads them from your disk, which matters for long recordings.
- In My Assets, select the camera clip and the recorder file (Cmd-click on macOS, Ctrl-click on Windows).
- Right-click one of them and choose AI multicam sync.
- ChatCut places both at the playhead on separate tracks and shows Preparing AI multicam sync with a percentage while it aligns them.
- Play the clap and a sentence near the end, and look at the waveforms.
- Mute the camera’s sound with the speaker control in its track header (track controls).
If the clips are already on the timeline, select both, make sure each plays at 1× speed, right-click and choose AI multicam sync. ChatCut moves the clips into place; it doesn’t change their speed or cut between angles. If it can’t find a confident match, it leaves them where they were, and if aligning would overlap two clips on the same track it makes no change, so move one to another track and try again. To fine-tune by hand, select the recorder clip and press E or R to nudge it one frame (add Shift for five) (shortcuts).
You can also ask the agent. Type @ in the chat box and pick your files in place of @camera and @recorder:
The built-in agent uses ChatCut credits for its work even when it only moves your own clips (Plans and credits).
For a whole shoot (several cameras, separate microphones, a camera that stopped and restarted), the AI Multicam Editor Skill does more. You upload 2 to 16 recordings from one session. It works out from their audio and transcripts which files are separate angles and which are pieces of one camera, places every file at a single measured offset on a synced master timeline, and by default builds a second timeline that cuts to whoever is speaking. It doesn’t time-stretch to hide drift; it reports a changing offset for you to review. Ask for sync only if you don’t want camera changes.
Fix audio that drifts
Drift means the recorder runs at a very slightly different speed from the camera. Moving it can’t fix that, so you change its speed by the amount you measure.
- Sync the start with any method above.
- Measure the gap at the end. Find a sharp sound near the end, ideally a tail clap, and note how far the recorder’s version is from the camera’s.
- Work out the ratio. Divide the time between your two sync points in the recorder by the same span in the camera. For example, if two claps are 600.0 seconds apart in the video and 600.3 seconds apart in the recorder, the recorder runs 0.05% long, so it needs to play at 600.3 ÷ 600.0 = 100.05% speed.
- Apply it to the audio, not the picture, keeping the pitch unchanged:
- Adobe Premiere: select the recorder clip, choose Clip → Speed/Duration, enter the speed as a percentage, and select Maintain audio pitch (Adobe).
- DaVinci Resolve: select the clip, choose Clip → Change Clip Speed, set the speed by percentage, frame rate or duration, and turn on Pitch Correction (21.1 manual, chapter 58).
- Audacity 4: drag the clip’s lower stretch handle slightly, then click the speed badge on the clip and type the exact percentage (Audacity). Export a WAV and use it in your editor.
- Re-sync the start and check the middle and end. Small changes like this sound natural; listen to the voice anyway.
ChatCut’s Speed menu offers common playback rates rather than fine percentages, and AI multicam sync expects clips at 1×. Correct a drifting recorder in an audio editor first, then import the corrected file and sync it.
The arithmetic also explains why drift sneaks up on you. At 0.05%, the gap grows by 30 milliseconds every minute. That’s invisible in a one-minute clip, but past the 125-millisecond threshold after about four minutes.
If the gap doesn’t grow steadily, it isn’t clock drift:
- It jumps at one point. Split the recorder clip at the jump and sync each part on its own.
- It wanders on phone footage. Check whether the video is variable frame rate. In the free MediaInfo app, choose View → Tree and look for Frame rate mode: Variable. Convert the file to a constant frame rate with a tool such as HandBrake or Shutter Encoder, then sync the converted copy. Adobe’s FAQ suggests both tools for files its own workaround can’t fix.
Fix an offset inside one file
Sometimes there’s no second recorder: a single video file, such as a screen recording or a downloaded clip, plays with its sound a little late. Separate the sound from the picture and move it.
In ChatCut, right-click the clip and choose Detach Audio, then select the audio clip and nudge it with E or R until a hard consonant matches the lips. Most editors have the same command under a name like Unlink or Detach Audio; in CapCut it’s Extract audio. Then export a new file.
VLC’s audio delay (Tools → Track Synchronization) only changes how VLC plays the file. It’s useful for watching a bad file, or for measuring how far out it is, but the saved video stays out of sync.
Check the exported file
Watch the export, not the timeline. Problems such as a variable frame rate clip or a scratch track left switched on are easiest to catch in the finished file.
- Beginning, middle and end. Look at a plosive (“p”, “b”) or a clap in each part of the video.
- Only the sound you want. If you hear a faint echo, the camera’s scratch track is still on.
- A different player and device. Play it on a phone as well as the computer. If it’s only wrong on one TV or Bluetooth headphones, the delay is in the device.
- The last minute of long files. Drift is worst at the end.
Once the master is in sync, the edit can start: for interviews, our guide to removing silences and filler words covers the next pass, and if the recorder picked up hum or traffic, see how to remove background noise from video.
Record the next take so it syncs easily
- Always record sound on the camera, even when it’s useless for the final mix.
- Clap in view at the start, with both devices already recording. On takes longer than a few minutes, clap again at the end so you can measure drift.
- Don’t pause one device and not the other. Every restart creates a new offset.
- Set both devices to 48 kHz, the usual rate for video. It won’t prevent drift, but it avoids conversions.
- Use timecode if your camera and recorder support it; it’s the fastest and most accurate way to sync long shoots.
- Choose a constant frame rate on phones and screen recorders when the app offers the option.
If you’re still choosing software for interviews and podcasts, our comparison of editors for podcasts and talking-head video looks at that question.