One YouTube video, 13 languages
Dubbed audio tracks in the narrator's own voice
YouTube lets a single video carry several audio tracks, and viewers pick their language in the player. The hard part has always been making the tracks. We made twelve of them for one film with Seed Audio 1.5, each in the original narrator's voice and each exactly as long as the video, and uploaded them to the same video.
The tracks and steps below come from our own channel, not a mock-up.
The result on YouTube
“The Last Ordinary Day in Pompeii” on the Ordinary Lives channel. Open Settings → Audio track in the player: English is the original, the other twelve were translated with Seed Audio 1.5.
- Audio tracks
- 13
- Video length
- 9:00
- Uploads
- 1 video
- Retiming needed
- None
Why one video beats twelve channels
Views stay together
Every language adds to the same view count, likes and comments, instead of splitting a small audience across separate uploads.
Viewers choose
The player offers the tracks in its audio menu, so a viewer in Seoul or Oslo can switch to their language without leaving the video.
No new edit
A track that is exactly as long as the video lines up with the picture as it is. You upload audio, not a new video.
From one video to many audio tracks
Our workflow for the Pompeii film.
Export the voice
If your project keeps narration on its own track, export it without music. That is what we translated, then we laid our own music and effects back under each language.
Translate in parts
Seed Audio 1.5 takes 4 to 360 seconds per request, so cut longer videos at pauses. Our 9-minute film went in as 4:43 and 4:17 and came back the same length.
Mix each language
Put the music and effects under every translated voice, at the same levels as the original, and export one audio file per language.
Translate the title and description
In YouTube Studio each language can also have its own title, description and thumbnail. We wrote them for all 12 languages.
Upload the tracks
In YouTube Studio open the video, go to Languages, add a language and upload its audio file. Repeat for each language and publish.
Check in the player
Open the public video, switch through the audio tracks and listen to the first minute of each.
What we learned
- Listen for names. In one of our test films a character's name changed form in the Russian translation; a glossary in the translation request prevents it.
- The YouTube Studio language dialog takes a moment to open; clicking before it finishes loading hits the wrong control.
- Use the regional variant your viewers expect: we chose Castilian Spanish for this film, and Mexican Spanish is a separate option.
- Translation starts from English or Chinese, so the original audio has to be in one of them.
YouTube multi-language audio FAQ
Can one YouTube video have several audio languages?
Yes. YouTube supports multiple audio tracks on a single video, and viewers switch between them in the player's audio track menu. Our Pompeii film carries 13.
How do I make the dubbed tracks?
We used Seed Audio 1.5 video translation, which re-voices English or Chinese speech in another language with the same voice and returns a track as long as the original.
Does the dubbed track have to match the video length?
A track the same length as the video lines up with the picture with no editing. Seed Audio 1.5 kept our tracks within 0.03 seconds of the source.
Which languages should I add first?
Start with the languages your analytics already show among viewers. Seed Audio 1.5 can produce 30, so the limit is how many you want to check by ear.
Do I need a separate video per language?
No. That is the point of audio tracks: one upload, one set of views and comments, and the viewer picks the language.
Related
Let your next video speak more languages
Translate the voice once per language, upload the tracks, keep one video.
Dub your video