On this page
Every time you record in OBS, the plugin can save a second file per language: the same video with the translated voice over the original sound. After you stop recording, it re-syncs that file so every translated sentence starts exactly when it was spoken on screen, and adds subtitles.
Who it's for#
Teams who record services and later upload them to YouTube, a website or a podcast in another language.
Before you start#
- OBS is paired and translating. See Set up audio, languages and start/stop in OBS.
- The language is one of the device's translation languages.
- Your OBS recording uses a normal format and quality. Lossless quality or a custom FFmpeg output is not supported, because the translated copy reuses the main recording's video.
- Enough free disk space: each translated copy is about the same size as your main recording.
Step 1 — Add a language#
- In the dock, go to Languages.
- Pick a language in the list at the bottom and click Add language. The dock says "Added …: its recording copy is on. Tick "Stream" to stream it too."
- A panel for that language appears with Record a … copy already ticked.
You can add up to 5 languages. Nothing is added to your scenes, and your main recording is not changed.
Step 2 — Choose the recording options#
In the language's panel:
- Record a … copy — on: every recording also writes a translated copy.
- Re-sync recording after stop (on by default) — rebuilds the file after you stop (see below).
- Keep live-mixed copy ("_live") (on by default) — also keeps the file as it was mixed live, for comparison. Turn it off to have it deleted once the synced file is ready.
- Captions in recording (on by default) — adds a subtitle track and a
.srtfile. - Show captions by default (on by default) — players start with the subtitles shown (MKV). Viewers can still turn them off.
How loud the original sound is under the translation is set with Original audio and Under translation; see Keep translation in sync with video.
Step 3 — Record#
Start and stop recording in OBS as usual. The panel shows "Recording → …" with the file name. Pausing the main recording pauses the copy too.
What re-sync does#
Live, each translated sentence reaches the recording 3–6 seconds after it was spoken, like a live interpreter. When you stop recording, the plugin:
- Waits a few seconds for the last translations ("Waiting a few seconds for the last translations, then syncing the … recording…").
- Rebuilds the file in the background so every translated sentence starts exactly when it was spoken in the video, over the recording's own original sound. The dock shows "Syncing … recording… 60%".
- Finishes with "… recording synced: …" and how many sentences were placed.
It takes about a minute per hour of recording. The video is copied, not re-encoded, so quality is unchanged.
Details worth knowing:
- If two sentences would overlap (translations often run longer than the original), the later one starts right after the earlier one.
- A sentence spoken while the recording was paused is left out, because it isn't in the video. A pause no longer splits the synced file.
- The silence a synthetic voice starts with is trimmed, so the voice itself starts on time.
Where files are saved and how they're named#
Translated files go in the same folder as your main recording (OBS Settings → Output → Recording Path). They use your main file's name plus the language code. For a main recording 2026-10-04 10-30-00.mkv and Spanish:
| File | What it is |
|---|---|
2026-10-04 10-30-00_es.mkv | The synced translated recording (with subtitle track) |
2026-10-04 10-30-00_es.srt | Subtitles as a separate file, e.g. for a YouTube upload |
2026-10-04 10-30-00_es_live.mkv | The live-mixed copy (if Keep live-mixed copy is on) |
2026-10-04 10-30-00_es_live_part2.mkv | Further parts of the live copy after a pause or automatic file split |
If the main recording is MKV, copies are .mkv; otherwise they are .mp4. Subtitle tracks use SubRip in MKV and mov_text in MP4. Subtitles show at most 2 lines of 42 characters per sentence.
Without re-sync (turned off, or it failed), the live-mixed file keeps the plain _es name and gets subtitles timed to when each sentence played live.
The Dashboard's device page shows each translated recording's status and file name, read-only.
Troubleshooting#
"No translated recording — add a language under Languages". You're recording but no language is added. Add one under Languages.
"… recording is off — tick "Record a … copy" under Languages". The language is added but its recording copy is off.
"Error: the main recording has no shareable encoder (lossless or custom output) — use a normal recording format". Change Settings → Output → Recording Quality away from Lossless, or stop using a custom FFmpeg output.
"Re-sync failed (…). The live-mixed recording is kept: …". The reason is in brackets, for example "the disk is full" or "the main recording file could not be opened". The live-mixed file is still usable.
"Not re-synced: the Custom mix has other sources on this track. The live-mixed recording is kept." Recordings using a custom audio mix are not re-synced.
FAQ#
Does this change my main recording?#
No. Your main recording stays exactly as before. The translated copy is a separate file.
Can I upload the .srt to YouTube with my original video?#
Yes. The .srt is timed to the synced translated file, which uses the same video timeline as your main recording.
Do I need to be streaming?#
No. Translated recordings work while only recording, if Translate when includes recording. See Set up audio, languages and start/stop in OBS.