How to get the best flow
A proven workflow for getting clean transcripts, natural timing, and token-efficient renders is to use CHAMELAION's Dubbing Studio. Follow the steps below to fix text first, then lock everything in with a final render 💯.
Step-by-step, the efficient way
Translate one target first, Lip-Sync off.
Upload your file, confirm the detected source, pick a single target source, leave Lip-Sync off for now, then run the first translation. This saves tokens while you fix text and timing 👍.
Open the result in the Dubbing Studio.
From your file, select the target language and click Edit in Studio. The Studio shows your Source and Target as blocks and timeline clips, with buttons per block: Transcribe, Translate via the arrow, and Generate Audio.
Verify and correct the source transcript.
Play Source and Target, fix obvious misreads first, for example, classic “Haus” versus “Maus,” since any error in the source will propagate into all targets.For each changed block, re-run the steps in order.
Click the Translate arrow on that block to update the target text, then click Generate Audio to synthesise a fresh dub in the correct cloned voice. If speaker labels are off, set the correct Speaker number before generating. For more information on how to fix Speaker Diarization, take a look at the article on that. It explains everything in detail, so you have a perfectly translated video 😎.
Tighten timing and pacing.
Use Selection mode to nudge clips and trim edges, and use the Zoom slider for precision. Split, fuse, or add clips as needed to make delivery feel natural.If you want to learn more about all the tools CHAMELAION's Help Centre has to offer, take a look at our Studio Tour article. This will get you up to speed 🚗.
Render, then add Lip-Sync if you want it.
Click Generate in the lower right. In the dialogue, choose whether to keep background sounds and whether to apply Lip-Sync for this render. For the most efficient run, select Lip-Sync only after you have made all changes to the transcripts.
Add more languages, then download
With a clean source, add additional targets back in the file view, or export SRTs for review and reupload if you prefer text-heavy edits.Now you can enable Lip-Sync directly without checking for transcript errors. CHAMELAION handles everything from now on 🤝 you just wait and look pretty 😉.
Why this workflow
Getting the source transcript right is everything; every target language builds on it. Fix the source once, then retranslate the changed lines and regenerate audio so your target stays consistent. New languages you add later will benefit from the corrected source automatically when you translate them.
When to use background sounds and language style
Choose Keep background sounds in the Generate dialogue if you want music and ambience mixed back in. Set Language Style to casual or formal before translating to match your audience's tone. You can always regenerate with updated choices. Any regenerated audio will mirror the changes you made to the source transcript 👍.
Quick checklist
Translate one target, Lip-Sync off.
Fix source transcript, then per changed block: Translate arrow, Generate Audio.
Adjust timing, assign speakers.
Generate to bake in edits.
Enable Lip-Sync, generate the final pass.
Add other languages, download what you need
Done!
Follow this loop: fix text first, then timing, then render, and only then add Lip-Sync for a natural, consistent result that is easy on your tokens and ready to share.