Handle multiple speakers on camera

Videos with multiple people speaking — interviews, panels, podcasts, roundtables — need a bit more setup so each speaker gets the right dubbed voice. This article covers how Dubly handles multi-speaker content and what you need to do.

Before uploading: set the speaker count

In the dub creation modal there's a Number of speakers dropdown with three options:

1 speaker — use this when only one person talks. Dubly skips speaker detection entirely.
2 speakers — use this for two-person interviews and dialogues.
Auto-detect — use this when there are 3 or more speakers in your video. Why this matters: if you set "1 speaker" on a video with two people, the second speaker's dialogue gets merged into the first speaker's voice — you lose the distinction. If you're in doubt, pick Auto-detect.

After transcription: assign voices

If your dub uses the Replace Voice cloning style and Dubly detects more than one speaker, your dub pauses in the "Assign Voices" state until you complete this step. You'll see a yellow warning badge on the dub in your list. Open the dub and you'll see a modal with one row per detected speaker:

Preview the original speaker (if available) — click the play button next to the speaker label to hear their original voice.
Preview each preset voice — click the play button next to a voice in the dropdown to hear what it sounds like.
Pick a voice for each speaker. The Confirm button activates only when every speaker has a voice assigned.
Click Confirm — the dub resumes processing.

For Original and Studio-Like modes, Dubly clones each speaker's own voice automatically. No assignment step is needed, so the modal won't appear.

Where Dubly struggles with multi-speaker content

Even with Auto-detect, the model works best when speakers are clearly separated in the audio:

Distinct voices. If speakers sound very similar, diarization may merge them into one.
Clean recording. Heavy background music or echo makes it harder to isolate individual voices. If your audio has a lot of crosstalk, consider editing the source video first — cut the worst overlaps or re-record affected sections before uploading.

Can i change the speaker count?

There's no way to change it after the dub is created. Delete the dub and start over with the correct setting.

Still stuck?

If you've assigned all voices but the dub isn't resuming, or if the speaker count seems wrong after diarization, contact support@dubly.ai with the dub title and a short description of what you're seeing.