What this is, in short
The Dubbing Studio is where you fine-tune your translation. Fix, add or delete transcript lines, adjust timing and pacing, assign speakers, and preview changes, then click Generate to make those edits part of the actual media. From your file, open a language and select Edit in Studio to get here. Edits only become part of the final video after you generate the video.
Where things are
Left pane, blocks
Your source is split into speaking blocks. Each block is processed independently and appears as a clip in the timeline. Use the buttons below each source block to run the pipeline in order: Transcribe Audio or edit the text manually, translate via the arrow, then Generate Audio for the target voice. Behind the scenes, CHAMELAION transcribes, translates, and synthesises your audio, and if Lip-Sync is on, it aligns mouth movements to the target speech.
Keep in mind: To save your tokens, you will never see aligned lip movement in the Dubbing Studio. However, if you select it, it will still make it into the final version of your Video 😉.
Centre, preview
The video preview always shows the most recent state of your edits. This is a preview only; it becomes permanent when you press Generate.
Bottom, timeline
You will see two tracks, Source and Target, with clips that correspond to the blocks. Use the Zoom slider for frame-accurate work or to zoom out for a wide view.
Track switch
Click Source or Target above the tracks, or click a track's name directly, to listen to that audio track while you work.
Edit basics, per block.
Transcribe → Translate → Generate
Run steps in order on any block that needs an update. The target clip updates when you generate audio for that block.Assign speakers
Change the Speaker number in the top left of a block to tell CHAMELAION which voice should speak that line. The target audio is synthesised in the corresponding cloned voice for that speaker when you generate the block 🗣️.Delete
Remove a block with the trash icon.Add a new line
Switch from Selection mode to Add Clip, then click in the timeline where the new line should live. Paste or type text, then generate audio.
Split at the playhead
Click the scissors at the current position to split one clip into two for finer control ✂️.Fuse
Click the fuse icon to merge two blocks in the timeline for a cleaner editing experience.
Auto Speed alignment
Click the speed alignment button to automatically align your blocks to the correct part of the video. This only shows up when a block is longer or shorter than usual 👍.
Timeline tools you will use a lot
In Selection mode, move and trim; drag clips to nudge timing, or drag edges to tighten starts and ends.
Zoom for precision, the slider in the lower right lets you zoom in for detailed edits or zoom out for structure 🔍.
Source vs. Target monitoring: use the toggle above the tracks, or click a track to compare the original and dub quickly.
Preview and generate
When you are happy with the text and timing, click Generate in the lower right. In this dialogue, you can also decide whether to include Background Sounds and whether Lip-Sync should be applied for this render. Generate locks your edits into the media so downloads reflect your latest choices.
Good to know
Downloading Video, Audio, or SRT later will mirror the exact render you generated. If you adjust anything, generate again before downloading.
If you work with SRT, you can export, edit, and reupload, then generate to bake those updates into the video. Updating the source SRT realigns all target languages; to change only that language, upload a target SRT. For more information, check out our SRT article 😄.
Best-practice flow
Translate one target language first, with Lip-Sync off.
Open the result in the Studio, listen to Source and Target, correct transcript lines and timing.
Generate.
Turn on Lip-Sync in the Generate dialogue and generate again for the final pass. This saves tokens and time by avoiding syncing lines you would change anyway.
For more information, check out our Best Practice collection 💪.
Quick reference to buttons and modes
Transcribe Audio runs speech-to-text on a source block.
Translate arrow creates the target text for that block.
Generate Audio: synthesises target audio in the cloned voice for the assigned speaker.
Speaker selector, assigns which speaker’s cloned voice says the line.
Trash deletes a block.
Selection mode, move and trim clips.
Add a Clip, insert a new block.
Scissors, split a clip at the playhead for detailed edits.
Source / Target monitor, switch what you hear while editing.
Zoom slider, adjust the timeline view.
Generate and render the latest edits, choose whether you want Background Sounds and Lip-Sync on or off, then proceed to download.
Fuse button: merge two blocks into a single block.
Auto speed alignment appears when one of your blocks is longer or shorter than usual, and it automatically aligns the block with the video.
Done!
That is it. Use blocks for text fixes, the timeline for timing, monitor Source or Target as you go, then generate to lock everything in. Your downloads will match exactly what you generated 🎞️.