You’re about to export a tutorial when you hear the presenter say “Open the Projects tab.” The screen recording shows the right menu, but it’s now called “My Projects.” It’s a two-second fix if all you hear is the voice-over. On camera, you can still see the presenter say the old line.
Lipsync Studio takes an existing clip and a replacement recording and adjusts the speaker’s mouth to follow the new audio. Start with one line so you can compare the result with the take you already have.
Lipsync Studio and this blog share the same owner.
Open the Lipsync Studio lip-sync workspace.
Hear the original and the replacement
This seven-second test uses an animated dragon and synthesized dialogue. The player starts with the source clip and its original sound, then plays the result from the video-to-video lip-sync model used by Lipsync Studio. The source asks, “Can one picture hold a conversation? Let us try it with two voices.” The replacement begins, “Yes. Keep both voices on the same timeline, and leave the silence in place.”
Watch the mouth as each line starts and notice what happens during the pauses. The two versions use different speech tracks, so you can hear the change the model had to follow.
Record the pickup before you touch the picture
If someone else is on camera, get their consent before changing what they appear to say. Listen to the lines on either side of the mistake, then record a pickup at the same pace and distance from the mic. Try “Open My Projects” before rewriting the whole sentence. If the new line lands in roughly the same amount of time, you can usually keep the cuts around it.
Trim a test clip around the line and record the replacement as a separate voice track. Leave a little room around the first and last words for the edit. Keep the original audio on a muted track in your editor so you can compare timing and recover the room tone at the seams.
Choose a lip-sync mode for the shot
For a single speaker facing the camera, try Lip Sync Video (Only Lip Region). It needs visible mouth movement in the original clip.
Now picture a hardware review with two hosts in the same shot. The host on the left says the recorder is set to 16-bit when it’s actually set to 24-bit. In Lip Sync Video (Speaker Control), upload the video first, then use the Control Who Speaks mask to mark the host on the left. Add the pickup and check that the other host’s mouth stays still where it should.
In either mode, start by rendering just the line and a little of the footage on either side.
If the speaker turns away during the corrected word or covers their mouth, consider cutting to the screen recording at that point. The audience can hear the corrected line while seeing exactly where to click. Save the lip-sync pass for the parts of the take where the face adds something.
Watch the result, then listen through the cut
Start with the clip muted. Look for a mouth that keeps moving through a pause, an obvious jump at the first word, or movement from the other person in frame. Turn the sound back on and check the consonants, then play the correction with the clips on either side. A line can look right in isolation and still stick out because the mic level or room sound changed.
Add your updated captions in the final edit. If the original captions are burned into the picture, use a clean source clip when you can so the old line stays out of the revised video.
If the original shot includes keyboard clicks or a door closing, bring those back on separate tracks where they belong. Duck music under the pickup if it obscures the new words.
For a first test, put the correction back in the timeline and play straight through the cut. If you find yourself listening to the edit instead of following the tutorial, go back to the pickup and check its level and room tone.