Motion Control + Lip Sync: Copy a Performance Onto a Face

Motion control copies how; lip sync copies what is said

AI motion control transfers movement from a reference video onto a character image. You keep the face you care about and copy the body timing: a dance, a hand gesture, a presenter stance. After that, if the mouth needs to match a new line, you run lip sync. Two tools, one character.

This is how you steal a performance legally from footage you own. Wireflow- and Dzine-adjacent motion tools, and Magic Hour-style image-to-video, live in the same category. LipSyncing AI keeps motion control next to talking photo so you do not pretend a dance generate is also a dialogue generate. It copies the reference mouth only if that mouth is in the reference. A new script is a second pass.

Motion control vs image to video, talking photo, and ad clone

Use Motion Control when you have a reference performance and a character still. Use Image to Video when you want prompted camera motion without a performance donor. Use Talking Photo when there is no body move to copy, only speech. Use Singing Photo if the point is a lyric on a still.

Use AI Ad Clone when the brief is an ad's layout, not a body performance. Use Video Lip Sync when the person in the tape should stay the person in the tape. Do not use a competitor's actor as a motion reference you plan to publish.

How to copy a performance, then talk

Full body if the motion is a dance. Bust if it is a talk. The still has to contain the parts that will move.

Match the still's framing to the reference. A bust still plus a full-body dance forces the model to invent legs. A full-body still plus a talking webcam reference wastes the legs. Pair like with like.

• Upload the character still you have rights to.
• Upload a 3–30 second reference clip you own or licensed. Subject clearly visible.
• Generate motion. Preview locally. Check identity lock on turns.
• If the job is dance or gesture only, export. Keep the reference soundtrack only if you have the rights.
• If the job is dialogue, run Video Lip Sync or Talking Photo on the result with the new line.
• Do not prompt a new script inside the motion job and expect visemes for free — that is a second generate.

Quality checklist for identity under motion

The point is identity lock while the body follows the donor. If the face becomes a cousin mid-turn, the still or the reference is too hard.

Pause on the hardest turn. If the face collapses, shorten the reference or pick a donor with fewer profiles. Lip sync cannot restore a face the motion pass already lost.

• Character stays the character on turns the model allows.
• Hands are acceptable on social gestures; close-up fingers may need a higher-quality pass or a different crop.
• Dance timing matches the reference.
• Mouth in the motion pass is from the reference, not from your new script — verify before you post a “talking” dance.
• After lip sync, the new line hits plosives without destroying the body move.
• Duration of the reference is the duration you actually wanted.

Credits for two passes

Each generate uses the meter. One credit = $0.01 on a plan. First 30 seconds = 150 credits ($1.50). Extra second = 5 credits ($0.05). 45s = 225 credits. 60s = 300 credits. A 12-second dance is still 150 credits. A 12-second talking pass on top is another 150. A 30-second reference that you then talk over is 300 credits for the pair.

Do not upload a 60-second routine “to be safe” — 300 credits of motion you will cut. Loop or cut longer dances. Preview locally. No free plan. Packs are $0.08 per credit and never expire. Yearly is 30% off, no auto-renew. Subscription credits spend first, then packs. Unused plan credits roll over while the plan is active and cancel if you cancel.

Rights: own the move and the face

You can technically run any dance file. You may not have the right to the performance or the music. Use references you own or have licensed. The character still has its own likeness and IP rules. TikTok dances as references are a licensing minefield if you publish commercially.

Filming a friend dancing in your kitchen is a clean donor if they agree to ads. Downloading a trending performance is not. The music is a third problem even when the friend agreed.

• Character image you own or generated as original.
• Reference performance you filmed or licensed.
• Music licensed separately if it ships in the export.
• No competitor talent as the donor for a published ad.
• Lip-sync audio is a third rights object: script and voice you own.

Common motion-control plus lip-sync failures

Using a headshot for a dance, using a stolen TikTok as the donor, and expecting one generate to do motion and a new line.

Posting the motion pass with the reference song still attached is how you get a Content ID hit and a rights complaint in one file. Strip audio you do not own before you add a new line.

• Headshot + full-body dance: upload a full-body still.
• New script inside motion-only: run lip sync second.
• Image to Video used when you had a donor clip: this page.
• Talking Photo used when you needed a copied stance: motion first.
• 60-second donor: cut to 30 or pay 300 credits.
• Stolen choreography and a stolen song: two rights failures, however clean the identity lock.