Caption languages, and how the language is chosen EN

Rekorder can caption recordings in around 99 spoken languages. The language of your captions has nothing to do with the language you use Rekorder in. The interface comes in English, Spanish, German, French, Italian, Japanese, Portuguese and Romanian, but you can caption a recording in any supported language. Free and Pro workspaces get the same list. For the full list, see Spoken languages supported by Rekorder.

How the language is chosen

When you generate captions, the Generate transcript window asks for the spoken language of the recording. You have two options:

• Pick a language from the Language list. This is the recommended option. Rekorder transcribes the recording as that language.
• Leave it on Auto-detect. Rekorder works out the language from the audio. Use this when you're not sure what's spoken, or when the recording mixes languages.

Rekorder remembers the result. The next time you open the window for that recording, the Language list shows the language you picked, or the one Rekorder detected if you used Auto-detect.

Only the person who owns the recording can generate captions. For the full steps, see Generate captions for a recording.

Summaries and chapters are generated from the transcript, and Rekorder uses the recording's caption language for them too. See Generate a summary and Chapters: generated, added and renamed.

Recordings with more than one language

Rekorder expects one dominant spoken language per recording, but the recording itself can still contain more than one language - you don't need to record each language separately.

With Auto-detect, Rekorder looks at the first 30 seconds of the recording to work out the dominant language, then transcribes the whole recording as that language. This usually gets it right, but if those first 30 seconds are silence, music, or a different language than the rest of the recording, detection can pick the wrong one - see below to fix it.

Fix captions in the wrong language

Auto-detection can guess wrong, especially when a recording opens with silence, music or a few words in another language. If your captions came out in the wrong language, or look like nonsense:

1. Open the recording and go to the Transcript tab.
2. Select Regenerate.
3. In the Language list, choose the language that's actually spoken.
4. Select Generate.

Regenerating replaces the current transcript, including any lines you've edited. Your summary and chapters are kept.

If you see "You recently generated captions for a recording. Please wait a minute and retry.", wait a minute and try again.

Things to know

• The list has one language per recording. Choosing a language tells Rekorder what to expect. It doesn't translate anything.
• Changing your interface language doesn't change existing captions. The two settings are separate.

Related articles

• Spoken languages supported by Rekorder
• Generate captions for a recording
• Caption generation failed, or is rate limited
• Fix a mis-transcribed name, or delete a line