AI Video Dubbing
Dub any video into another language. Voice Creator Pro separates each speaker and clones their voice from the source audio automatically, so everyone keeps their own voice in the dub, with lip sync and styled subtitles.
Video Dubbing takes a finished video and re-voices it into another language. Drop in a video, pick a target language, and Voice Creator Pro transcribes the speech, translates it, works out who is speaking, clones each voice from the source audio, and dubs every line in that speaker's own voice. You can then lip-sync the video to the new audio and add styled subtitles.
You can dub into 21 languages, including Spanish, French, German, Japanese, Korean, Chinese, Hindi, Arabic, and more. Dubbing is on the desktop app, where it runs on your own computer, and on VCP Cloud. Open it from Dubbing in the Video section of the sidebar.
How It Works
1. Add Your Video
Click New project and drop in a video, or click to choose one. Dubbing takes MP4, MOV, MKV and WebM files.
2. Choose Your Settings
- Source language - The language spoken in the video. Leave it on Auto-detect to have it picked for you.
- Target language - The language to dub into.
- Number of speakers - Leave it on Auto-detect, or set it if you know how many people speak. It makes speaker detection much more accurate.
- Speaker detection - Choose pyannote or Nemotron. Nemotron is better at quick back-and-forth and short replies, and handles up to 8 speakers. On the desktop app, Nemotron needs an NVIDIA graphics card.
- Clone voices from the video - On by default, so every speaker keeps their own voice. Untick it to pick a Default voice instead. You can set a different voice for each speaker once they are detected.
Open Advanced settings for more control:
- Voice model - The model that speaks the dub. Only models that support your target language are listed.
- Transcription model - The model that transcribes the original speech.
- Timing - How dubbed lines fit the original timing:
- Balanced (recommended) - Keeps natural speed, speeding lines up only when needed to stay in sync.
- Match original timing - Speeds up long lines so each stays in its original slot. Tightest sync, can sound rushed.
- Keep natural speed - Never speeds up audio. Lines can run long and drift later than the original.
- Keep background music / SFX - Keeps the original music and sound effects under the new dialogue.
- Lip sync after dubbing - Lip-syncs the video as soon as the dub is done, with a Lip sync quality (steps) slider.
On VCP Cloud you see the estimated token cost before you start.
3. Start Dubbing
Click Start dubbing. Voice Creator Pro reads the video, separates the voices from the background, transcribes, detects speakers, translates, generates the new speech, then mixes and exports it. You can switch to another part of the app while it runs, and the editor opens when dubbing finishes.
Along the way:
- Each voice is cloned from the video. A speaker who only says short lines still gets their own cloned voice.
- Talking over each other is kept. When two people talk at once, the quieter person's line and quick replies like "okay" are kept, and dubbed lines can overlap the way the original did.
- Pauses stay where they were. The dub pauses where the speaker paused, in translated dubs too.
- Laughs and other sounds are kept. Laughter, sighs and coughs from the original are kept out of the cloned voices and played back in the dub.
4. Review, Fix and Export
Play the dub in the editor, fix any lines that need it, add subtitles and lip sync, then click Export video.
The Editor
The video is on one side and the list of segments, one per line, is on the other. Above the video, switch between Original, Dubbed and Lip-synced to compare. Click a segment to jump to it in the video.
Fix a Line
Hover a segment to see its tools:
- Play - Hear the dubbed line on its own.
- Edit - Click the pencil icon to change the translation.
- Edit the original text - Click the original line above the translation to fix the transcript, then click Re-translate.
- Regenerate - Generate the line again in the same voice.
- Generation settings for this segment - Click the sliders icon to change the voice model's settings for just this line.
- Start and end times - Type a new Start time or End time.
- Reassign speaker - Click the speaker name to move the line to another speaker, or choose New speaker… and pick a voice for them.
- Merge next - Combine the line with the next segment.
- Split - Split a line where a second speaker starts. Click the word where the next speaker starts.
- Delete - Remove the line from the dub.
Segments also show badges. Pending means the line was edited and needs regenerating. No voice means its speaker has no voice yet, so pick one in Speakers and regenerate it. Edited means the original text changed and the line needs re-translating. stretched shows how much a line was sped up to fit.
When you have edited several lines, click Regenerate all in the notice above the list, or Run again at the top of the editor. Your manual edits are kept, and only edited, pending, or outdated segments are regenerated.
Use Search segments or press Ctrl+F to find a word, and Copy text to copy the original or the translation.
Speakers
The Speakers panel above the segments lists everyone in the video. For each speaker you can:
- Select voice - Keep their Original voice or pick another voice, like one you designed or one from your library.
- Play cloned reference - Hear the clip their voice was cloned from.
- Trim bad audio off the reference - Cut noise or another voice off the start or end of the clip.
- Use a different clip as the voice reference - Pick a cleaner line from this speaker.
- Rename this speaker and Remove speaker - When you remove a speaker, you choose who gets their lines.
- Add speaker - Pick a line spoken by someone who was missed. That line moves to the new speaker and becomes their voice reference.
Voice changes apply the next time you regenerate.
Undo and Keyboard Shortcuts
Use Undo and Redo at the top of the editor, or press Ctrl+Z and Ctrl+Y. Space plays and pauses the video.
Glossary
Click Glossary to list terms the translator must keep as they are, like product names, or always translate the same way. Lines that don't follow the glossary yet are marked Outdated and are redone the next time you click Run again.
Lip Sync
Click Lip sync to sync the mouth movements to the dubbed audio. When it's done, the Lip-synced tab shows the result. Lip sync works on videos up to 60 seconds long.
On the desktop app, lip sync needs an NVIDIA graphics card with at least 8 GB of memory, and the settings icon next to Lip sync sets the quality and sync strength. On VCP Cloud you see the token cost before it starts, and you can stop it with Cancel lip sync.
Subtitles
Click Subtitles on the video to add captions made from the translation. Pick one of about 60 built-in styles, customize the font, outline, shadow, background box, word highlighting and motion, and save your own styles to reuse. You can also restyle a single caption, or click a word to give it its own colour, size or animation.
When lines overlap, both captions stay on screen, one above the other.
See Subtitle Styles for every styling option.
Export
Click Export video and choose:
- Download video - The dubbed video. When subtitles are on, they are burned in.
- Download audio only (.wav) - The dubbed audio on its own, for example to upload as an extra language track on YouTube.
- Export to After Effects (.jsx) - A script that builds the subtitles as layers in After Effects.
- Download subtitles - An SRT, WebVTT or JSON file, or the styled subtitles as a Transparent video (Animation .mov, ProRes .mov or WebM) at 1×, 2× or 4× the video size. The JSON file has word timings, timecode and speaker names, ready for Premiere and After Effects.
Click the gear icon for Project settings to choose the container, codec, quality, resolution, frame rate and audio format of the exported video.
Exports keep going if you leave the screen, and the result is waiting when you come back. On VCP Cloud, burning in subtitles uses tokens, and free plans export watermarked video at 720p.
Supported Languages
You can dub into 21 languages, including Spanish, French, German, Japanese, Korean, Chinese, Hindi, and Arabic, among others. The full list is in the Target language menu when you start a dub.
Use Cases
Localize Video Content
Reach a global audience by dubbing tutorials, courses, and marketing videos into multiple languages, with the presenter keeping their own voice in each one.
Podcasts and Interviews
Dub multi-speaker conversations so each host and guest keeps their own voice in the new language.
Social Media
Repurpose a single video across regions by exporting a dubbed version per language, complete with styled captions for sound-off viewing.
Film and Education
Translate scenes, lectures, and training material with matching lip movement so the result feels natural.
Next Steps
- How to Dub a Video - Step-by-step walkthrough from import to export
- Subtitle Styles - Every style and setting for your captions
- Voice Cloning - Clone a voice yourself to assign to a speaker instead of the automatic clone
- Voice Changer - Re-voice a recording into a single target voice
Project Settings
Configure generation, segmentation, output, and model settings for your audiobook.
Subtitle Generator
Generate accurate, perfectly timed subtitles for any video or audio, with word-level timestamps, translation into 21 languages, styled burned-in captions, and SRT, WebVTT, JSON or transparent video export.