Dubbing used to cost more than the original production. Today, with the right model and the right settings, you can translate a 5-minute video into 30+ languages in an afternoon — keeping your voice across all of them.
Upload the source
Quality of dub follows quality of source. The cleaner your audio, the better Lumen can transcribe, translate and re-synthesise. If your source has heavy background music, run Voice Isolator on it first to extract just the dialogue, then re-mix the music underneath the dub.
Source & target language
Pick what's spoken in the video and what you want it translated to. Lumen auto-detects but trusting the picker is faster.
Voice cloning
Leave Voice cloning on. This is what keeps the dub feeling like you. Lumen samples your original voice and re-renders the translated audio in the same timbre, pacing and emotion.
If your video has multiple speakers, bump Speakers up — Lumen will diarise (separate them), clone each one, and keep each one's voice consistent in the target language.
Preserve background
This separates your dialogue from your music and ambience, dubs only the dialogue, then re-layers everything underneath. Leave it on. Without it, the dub flattens your soundscape.
Lip sync
The most expensive setting and the most impressive. Lip sync re-renders the mouth shapes of every face in the video so they match the new language perfectly. Use it for hero shots, B2B videos, and anything where viewers will be watching faces. Skip it for fast-cut social shorts where it's overkill.
What to expect
A 1-minute dub with cloning + preserve background + lip sync takes roughly 90 seconds and costs ~18 credits. Without lip sync, costs drop to ~10 credits and runtime to ~50 seconds.
Iterate by language
If the first dub feels stiff, the issue is almost always the source script being translated too literally. Lumen translates idioms culturally where it can, but rare phrases or product names can throw it. Try generating with a slightly cleaned-up version of your transcript as the source.