Transcription

​

Generate source captions

Add a video, open it in the workspace, and choose the language spoken in its audio. Start transcription after checking the credit estimate. Transcription creates timestamped source captions and speaker assignments for review. Those captions provide the source text for translation and dubbing.

​

Wait for processing

Submitted work runs on the server. You can leave the page and return later to check the video and job status. Processing time varies with the source length, service availability, and queued work. A fixed length threshold does not determine whether you must keep your browser open.

​

Review the source

Open the source caption track and check:

• Names, numbers, and specialist terms
• Missing or incorrect speech
• Speaker assignments
• Segment timing and splits

Correct the transcript and click Save before starting translation. Save updates the selected track’s edits rather than creating a new historical version for every change.

​

Prepare a better source

Clear speech with limited overlap gives better results. If the source has heavy noise, music, or missing audio, use a cleaner recording where possible. See Caption Editor, Audio Quality Tips, and Translation.