One Live Stream, Many Languages: Extra Audio Tracks on DCAST

A conference keynote in English, watched by an audience that also speaks Spanish, German and Ukrainian. The usual answer is four streams, four links, four chats and four times the chance that something goes wrong. On DCAST it is one stream: the picture and the original sound come from your main encoder, each language arrives as its own audio feed, and viewers pick the language they want with one button in the player.

This article explains what the feature does today, who it is for, and where its current limits are — so you can plan an event around it with no surprises.

What "extra audio tracks" means

A normal live stream has one picture and one soundtrack. With extra audio tracks, the stream keeps one picture but carries several soundtracks side by side. Viewers see the same video and choose which soundtrack to hear.

On DCAST you add the tracks yourself, in your stream's settings, on the Feeds tab under Languages. Every time you press Add language, DCAST creates a new track and gives it:

• a name viewers see in the player (for example "Español" or "Deutsch" — up to 64 characters, any language you like);
• its own connection details — for an RTMP stream, a stream key that belongs to that track; for an SRT stream, its own Stream ID.

You can add up to 16 extra languages to one stream. The main stream's own sound always stays available to viewers as Original.

Who sends the languages: you, not a machine

DCAST does not generate translations. Every language track is audio that you, your team or your interpreters produce. That is a deliberate choice: an interpreter who knows the subject, the speaker and the audience is still the standard for events where words matter — medical training, legal briefings, product launches, church services, sports commentary in the local language.

In practice, each language is a separate, lightweight feed:

1. The main encoder (OBS, vMix, a hardware encoder) sends the picture and the original sound, exactly as it does today.
2. Each interpreter — in the same room or at home — sends their voice to the key of their language track.
3. DCAST puts all of them together under one player.

The interpreter does not need to send video. Only the audio of a language feed is used; any picture in it is ignored. If your encoder insists on a video track, send the smallest one it allows.

Recommended settings for a language feed

These are the values we show in the i tooltip (Recommended stream settings) next to the Extra audio tracks heading:

The one thing to avoid is sending full-quality video to a language track. It costs the interpreter's upload bandwidth and adds nothing, because the picture is thrown away.

Timing: before, with or after the main stream

Language feeds do not need to be started in a strict order. A track can connect before, together with, or after the main stream. This matters on event day: an interpreter can connect early and wait, and a late interpreter can still join without anyone restarting the broadcast. The one rule: add all your languages in the Feeds tab before you press Start.

While you prepare, the Feeds tab shows a small signal dot next to every track — green while that track's feed is arriving, grey when it is not. You can see at a glance which interpreters are connected before you press Start.

What viewers see

Viewers do nothing special. They open the same watch page or the same embedded player as always.

• When a stream carries two or more audio tracks, a language button appears on the player's control bar, showing the name of the track that is playing. On narrow screens it shrinks to an icon.
• Original — the main stream's own sound — is first in the list and plays by default.
• Choosing another language switches the sound; the picture keeps playing.
• Streams with a single soundtrack look exactly as before. There is no empty button and no extra menu.

Because the language choice lives inside one player, everything else stays shared: one link to promote, one chat, one viewer count, one paywall if you sell access. A viewer who paid for your event does not need a second ticket to hear it in their language.

Which protocols are supported

Extra language tracks work for streams that you send over RTMP or SRT. Each language track uses the same protocol as its stream: an RTMP stream receives its languages over RTMP, an SRT stream over SRT (the Feeds tab also lists every SRT parameter your encoder asks for).

Streams sent from the browser camera, over WHIP, from an HTTP source or from a file do not have the Languages tab today.

Current limits — read this before you plan

We would rather you hear these from us than discover them during your event.

Plan the multilingual part as one continuous broadcast. The replay keeps every language — but if you press Stop and then Start again on the same stream, the replay offers the extra languages only for the part after the last Start. For an event with a break, keep the broadcast running through the break (for example with a holding picture from your encoder) rather than stopping it.

The replay is the copy your viewers watched live. After the stream ends, viewers of the replay choose a language with the same button. For streams with extra languages, the replay is not re-encoded after the broadcast yet; it keeps the quality of the live stream. If you need separate audio files of each language, record each interpreter locally.

Interpretation quality is yours. DCAST carries the interpreter's audio faithfully; it does not correct levels or delay. Ask interpreters to test their feed with you before the event (the signal dot and the player make that easy).

Why this matters for creators who sell their own events

If you sell tickets or subscriptions, every extra language is a new market that does not need a new event. A course recorded in one language can be taught live to three audiences at once. A conference can sell international passes without a second production crew. And because everything sits on one stream, your analytics, your chat moderation and your access rules stay in one place.

How to get started

1. Open your stream's settings and go to Feeds → Languages.
2. Press Add language for each language and rename the track to what viewers should see.
3. Give each interpreter the connection details shown on their track's card.
4. Watch the signal dots turn green, then start your broadcast.
5. Open your own watch page and try the language button, as a viewer would.

Our step-by-step guides show the encoder side in detail for OBS, vMix and the Larix mobile app.