I am using dubbing studio, and often have overlapping dialogue. Would creating a new dubbing projec | ElevenLabs | Page 1

desert irisBOT Nov 9, 2025, 1:16 PM

#

AI-generated Answer for soundofmusso

Short answer: No. You’ll get better results by keeping all characters in a single Dubbing Studio project and managing them on separate speaker tracks.

Why:

Dubbing automatically separates speakers and can split overlapping speech into distinct tracks, so you don’t need one project per character. (elevenlabs.io)
In Dubbing Studio you can create multiple speaker tracks, reassign clips between speakers, and split/merge clips to resolve overlaps directly on the timeline. (elevenlabs.io)
You can set per‑track voice behavior (e.g., Track clone for a consistent voice across that speaker’s clips), which is designed for multi‑speaker projects. (elevenlabs.io)

Note: ElevenLabs recommends dubbing content with a maximum of nine unique speakers at a time for best quality. Keep this in mind if your project involves very large casts. (elevenlabs.io)

Sources:
Dubbing | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation
Dubbing | ElevenLabs Documentation

mighty sky Nov 10, 2025, 1:36 AM

#

For my manual dubbing, I've uploaded my separate .csv file with the correct timecodes. However, upon opening my project, the timecodes are incorrect, and I have to manually correct them.

desert irisBOT Nov 10, 2025, 1:36 AM

#

AI-generated Answer for soundofmusso

Likely cause: the CSV’s timecode format isn’t being parsed the way you expect.

What to check and how to fix

Confirm your CSV follows the required schema and uses one of the three supported timecode formats exactly: seconds, hours:minutes:seconds:frame, or hours:minutes:seconds,milliseconds. The CSV must strictly follow the predefined format. (elevenlabs.io)
If you use hours:minutes:seconds:frame, set the correct frame rate when creating the dub. In the API you can pass csv_fps to control how frame-based timecodes are interpreted; if you don’t set it, FPS is inferred from the timecodes (which can lead to mismatches). Create the project in mode=manual with your CSV and csv_fps set, then open it in Dubbing Studio. (elevenlabs.io)
ElevenLabs added csv_fps for “more precise timing control,” so using it for frame-based CSVs is the recommended path to avoid off-by-frame offsets. (elevenlabs.io)
Reference examples of correctly formatted CSVs (including seconds and milliseconds forms) here to validate your file before upload. (elevenlabs.io)

Docs for Dubbing Studio and Manual Dub are here, including CSV examples and exports. (elevenlabs.io)

If you still see drift after these steps, please try creating the project via the API with mode=manual and an explicit csv_fps that matches your source video’s frame rate, then open the result in Dubbing Studio for editing. (elevenlabs.io)

mighty sky Nov 12, 2025, 4:24 AM

#

mighty sky For my manual dubbing, I've uploaded my separate .csv file with the correct time...

When I merge two timespaces in the dubbing studio, the translations I have made are replaced with automatic translations. How can I prevent htis?

desert irisBOT Nov 12, 2025, 4:24 AM

#

AI-generated Answer for soundofmusso

Short answer: There’s no documented setting in Dubbing Studio to “lock” manual translations against changes when you merge timeline clips. The product guide notes that when you split or merge clips, the Speaker Cards (where you edit transcripts/translations) update to reflect those timeline changes, and it does not describe a way to prevent that. (elevenlabs.io)

What you can do (per official docs):

Avoid structural edits after you’ve translated: finalize your splits/merges first, then edit the transcript/translation in the Speaker Cards. The guide explicitly shows where to edit translations and that Speaker Cards update on merge. (elevenlabs.io)
If you need fixed segments that won’t be changed by merges, create the project as a Manual Dub. Manual Dub makes Studio “stick to your exact clips and speaker assignment” based on a CSV you supply (speaker, start_time, end_time, transcription, translation). (elevenlabs.io)
Before making large timeline changes, export a CSV (includes speaker, start_time, end_time, transcription, translation) so you have a backup you can re-use if needed. (elevenlabs.io)

References:

Dubbing Studio product guide: Speaker Cards; Split and Merge clips; Exports; Manual Dub. (elevenlabs.io)

mighty sky Nov 12, 2025, 6:34 AM

#

Dubbing studio: the handle bars to lengthen or shorten a speaker card are too thick. And for some very short speaker cards, the handle bars are not available at all.

desert irisBOT Nov 12, 2025, 6:34 AM

#

AI-generated Answer for soundofmusso

I can’t find any official documentation about changing the thickness of the resize handles, or about handles being hidden on very short clips. The Dubbing Studio guide only documents these timing workflows:

Trim without grabbing handles: place the playhead exactly using the timecode field on the right, click Split at that position, then delete the unwanted piece and (optionally) Merge adjacent clips. (elevenlabs.io)
Let the system adjust duration: right‑click the clip and switch to Dynamic Generation so the clip length follows the text length. Note this can affect sync if there isn’t space before/after the clip. (elevenlabs.io)
After changing lengths or settings, regenerate stale audio (individually or via Generate Stale Audio) so the audio matches the new timing. (elevenlabs.io)

If the resize handles are hard to use on your device, please report it to Support with a short screen recording and your browser/OS details so the team can investigate. (help.elevenlabs.io)

Sources:
Dubbing Studio | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation
How can I contact support? – ElevenLabs

mighty sky Nov 12, 2025, 7:24 AM

#

manual dubs: when exporting as a zip audio track, the tracks are named from Track_0 to Track_<number of tracks>. They should be just given their orioginal track names... For example, 'background', 'foreground', '<speaker name> and soforth

desert irisBOT Nov 12, 2025, 7:24 AM

#

AI-generated Answer for soundofmusso

I can’t find any official ElevenLabs documentation that describes how WAV “separate tracks” ZIP exports are named or any setting to preserve original track names. The Help Center only notes that Dubbing can export WAV as “Audio – separate tracks for each speaker, downloaded as zip file,” without detailing filename conventions or customization options. (help.elevenlabs.io)

If you need timeline metadata rather than per-file WAVs, AAF (timeline data) export is also available, which may suit certain workflows—but the docs don’t state anything about track-name preservation. (help.elevenlabs.io)

For a definitive answer or to request support for original track naming in ZIP exports, please contact ElevenLabs Support via the Help Center contact form. (help.elevenlabs.io)

Sources:
Which file formats are supported by Dubbing? – ElevenLabs
Which file formats are supported by Dubbing? – ElevenLabs
How can I contact support? – ElevenLabs

mighty sky Nov 12, 2025, 9:31 AM

#

Dubbing studio: So when I generate audio with "dynamic duration", the original audio clip is ignored?

desert irisBOT Nov 12, 2025, 9:31 AM

#

AI-generated Answer for soundofmusso

Short answer: Dynamic Generations do not “ignore” the source, but they do stop enforcing the clip’s original duration. Instead, the system stretches or shrinks the generated speech to match the text length, which can overrun the current clip region and affect sync. If there isn’t enough space before the next clip, the generation may fail.

What to do:

To preserve the original timing, use Fixed Generations (the default). Right‑click the clip and switch back to Fixed, then regenerate. (elevenlabs.io)
If you changed clip length or settings, regenerate any “stale audio” so the output reflects the current timeline. (elevenlabs.io)

Sources:
Dubbing Studio | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation

mighty sky Nov 13, 2025, 4:56 AM

#

Where can I submit a bug report for manual dubs? The CSV file is not processed correctly in certain cases.

desert irisBOT Nov 13, 2025, 4:56 AM

#

AI-generated Answer for soundofmusso

Submit it via the ElevenLabs Help Center “Submit a request” form. Choose the relevant category (e.g., Customer Support or Problem Tickets), set Type to “Bug report,” and attach your CSV plus any screenshots and the affected project details so the team can reproduce it. (help.elevenlabs.io)

Sources:
Submit a request – ElevenLabs

mighty sky Nov 13, 2025, 4:58 AM

#

desert iris

is it possible to change the target language after a project is created?

desert irisBOT Nov 13, 2025, 4:58 AM

#

AI-generated Answer for soundofmusso

Yes—with Dubbing Studio you can add more target languages to an existing project.

In the Studio UI: open the project and use the “+” in the Dubs panel to add another language; you can then toggle between languages in the timeline. (elevenlabs-sdk.mintlify.app)
Via API: POST /v1/dubbing/resource/:dubbing_id/language with the language code to add a new target language to the project. This only adds the language; you still generate transcripts/translation/audio afterward. (elevenlabs.io)
Billing: you’re charged per output language. For Studio projects, ElevenLabs recommends creating the project with one language and then adding additional languages inside Studio so edits stay in sync; translation for added languages is billed per character plus standard TTS generation. (help.elevenlabs.io)

Note: There isn’t a “switch” that converts an existing dub’s target language; instead, add the new language (and optionally remove the old dub) within the same project. (elevenlabs-sdk.mintlify.app)

Sources:
Studio - ElevenLabs
Add language to resource | ElevenLabs Documentation
How much does Dubbing cost? – ElevenLabs
Studio - ElevenLabs

mighty sky Nov 13, 2025, 5:14 AM

#

in manual dub, can i add more speakers after a project has been created?

desert irisBOT Nov 13, 2025, 5:14 AM

#

AI-generated Answer for soundofmusso

Yes. Manual Dub projects open in Dubbing Studio, where you can add more speakers after creation:

Use the action buttons at the bottom of the timeline to add a new audio/Voiceover track; Voiceover tracks create new Speakers. (elevenlabs.io)
You can then assign clips to that new speaker (or drag existing clips between tracks to reassign speakers). (elevenlabs.io)

Sources:
Dubbing Studio | ElevenLabs Documentation
Dubbing Studio | ElevenLabs Documentation

mighty sky Nov 13, 2025, 5:16 AM

#

what's the difference between a dubbed speaker track and voiceover track when adding speakers in manual dub?

#I am using dubbing studio, and often have overlapping dialogue. Would creating a new dubbing projec