Lip Sync or Re-perform: Choosing How to Redub a Clip

Quick answer: Lip sync is the default: the picture stays frame for frame, every line is voiced again and the mouths are moved to match; the original ambience is dropped. It costs 2 credits per second of clip and works on clips of any length. Re-perform plays the clip again with the new lines, so the delivery and timing fit the words and the sound is a full new performance, but the picture may shift a little. It costs 2× the per-second price of the clip’s tier and is offered only for clips of 15 seconds or less. In SceneMixer, Lip sync is priced per second of clip and works at any length, while Re-perform is limited to clips of 15 seconds or less.
Still from “Wings of the Shattered Spire”, an AI short drama generated end-to-end by SceneMixer (5 characters, 1 episode)
SampleWings of the Shattered SpireShot on Film1 episodeProduced end-to-end by SceneMixer from the scriptWatch the episodes

Same 10-second clip on Wan 3.0 · 480P: Lip sync costs 10 × 2; Re-perform costs 10 × 20. Try Lip sync first.

How Lip sync works

The site voices each line of the segment in the voice shown on its row — by default Original voice, a clone of that speaker’s voice from this clip, when the clip allows it — lays them out one after another from the start of the clip with short pauses in between, and uses that as the clip’s new soundtrack. The picture is kept, and the speakers’ mouths are moved to fit the new audio. Two consequences follow. The pauses of the original are not kept: lines that were spread across the clip now come one after another. And the soundtrack holds only the voices: the take’s own music and ambient sound are gone (the episode’s score is a separate layer, mixed back under the new version automatically).

It looks best when faces are turned toward the camera. On faces in profile, or when a speaker is small in the frame, the result can look or sound less natural; listen to the voice before you run it.

How Re-perform works

Re-perform treats the new lines as a new take: the clip is made again from itself with the lines replaced and the chosen library voices as the speakers’ voice samples (Original voice is a Lip sync option). Because the performance is made fresh, the timing and delivery follow the new words and the soundtrack is a complete take, ambience included. The framing and positions follow your clip, but faces and small movements can differ a little. It uses the same machinery as AI visuals, which is why it has the same price and the same 15-second limit.

Choosing

  • A word or two changed, lines already back to back: Lip sync.
  • A clip longer than 15 seconds: Lip sync (Re-perform is greyed out).
  • The picture must not change at all: Lip sync.
  • The take’s own ambience matters, or the new line needs a different rhythm: Re-perform. (The episode’s score comes back by itself either way.)
  • The line is much longer than the old one: neither may fit; edit the storyboard, add seconds to the shot and regenerate.

Price side by side

In SceneMixer, Lip sync costs 2 credits × the length of the whole clip in seconds. Re-perform: 2× the per-second price of the tier the clip was made on × the length. On any tier, Re-perform is many times the price of Lip sync; both prices are printed on the two mode buttons before you choose. Both are held when you press and refunded if the redub does not go through. Every fix price.

Either way

Both write changed words back into the storyboard, both save a Redubbed version and keep the generated clip, and a segment keeps one redubbed version. Change a line after the video is made.

Lip syncRe-perform
PictureEvery frame keptMade again; small changes possible
VoicesRe-voiced, one after another from the startA new performance
The take’s own soundDropped; the episode’s score returns on its ownNew soundtrack, ambience included
Clip lengthAnyUp to 15 seconds
Price2 per second2× tier price per second
Ad projectsYes; in a newly scored ad, re-compositing mixes the ad’s music back under it. In an ad made before the new scoring, the clip’s own music goes with the replaced soundtrack — re-perform or regenerate that clip insteadYes

FAQ

Why is Re-perform greyed out?

The clip is longer than 15 seconds, or re-rendering is unavailable for this clip at the moment. Use Lip sync.

Why does the lip-synced clip sound empty?

At first it holds only the voices: the take’s own ambience is dropped, and the episode’s score is mixed back under it automatically, usually within minutes. Re-perform keeps a full soundtrack from the start, on clips up to 15 seconds.

Can I keep the original pauses with Lip sync?

No. The lines are placed one after another from the start. Re-perform follows the new delivery instead.

The redubbed clip looks or sounds wrong. What now?

Switch back to the clip you started from in SceneMixer's Version History; it is always kept. A redub that looks off on a side-on or small speaker usually does better as Re-perform on a clip of up to 15 seconds.

Which looks more natural?

For faces toward the camera, Lip sync is usually convincing and much cheaper. Side-on or very small speakers are harder for it; if the result looks off, try Re-perform on a clip up to 15 seconds, or regenerate the segment with the new line.

Start with Lip sync

It is the cheap one; switch to Re-perform when the sound needs a new take.

Open your project

Updated November 15, 2026