Live Glossa
OBS Studio translation guide

How to add real-time translated audio to OBS Studio

A production-ready workflow for capturing Live Glossa's translated voice and captions in OBS, preventing echo, and sending a language-specific feed to YouTube, Facebook, or RTMP.

Read the guide
The practical answer

How it works

Keep Live Glossa in Chrome or Edge so authentication, microphone permission, and live-session recovery work normally. OBS captures that browser window's translated audio as an application source, combines it with your video scene, and sends it to the audience. This separation is more reliable than embedding the signed-in control room as an OBS Browser Source.

  1. 1

    Prepare Live Glossa

    Sign in, select one target language, grant microphone access, and use headphones during setup.

  2. 2

    Create the OBS scene

    Add your camera, presentation, graphics, and an optional Window Capture source if viewers should see translated captions.

  3. 3

    Capture browser audio

    On Windows, add Application Audio Capture (BETA) and select the Live Glossa Chrome or Edge window. OBS 30.1+ can also capture audio through Window Capture.

  4. 4

    Remove duplicate audio

    Disable Desktop Audio if it captures the same browser. One translated source should appear in the OBS mixer.

  5. 5

    Test before going live

    Record at least 60 seconds. Confirm translated speech, caption alignment, volume, and the absence of English bleed or echo.

Why production teams use this workflow

Per-application capture isolates translation from notifications and unrelated desktop sound.
A dedicated output lets each language audience hear only its intended feed.
OBS scenes preserve existing cameras, presentations, lower thirds, and branding.
Private recordings expose routing errors before they reach a live congregation.

Windows and macOS audio capture

OBS 28 and newer supports Application Audio Capture on Windows. Select the browser process running Live Glossa and use the matching-executable option when changing tab titles makes window matching unreliable.

On macOS 13 or newer with OBS 30, use macOS Audio Capture or macOS Screen Capture with audio. Grant Screen & System Audio Recording permission. Older macOS versions may require a virtual audio device following OBS's documented desktop-audio workflow.

Avoiding echo and source-language bleed

Echo usually means the same browser output is captured twice—once as application audio and again through Desktop Audio. Disable one path. Acoustic echo happens when translated audio plays through speakers and re-enters the microphone, so monitor with headphones.

If English reaches the translated destination, inspect OBS Advanced Audio Properties. The microphone or main program mix may be assigned to the translated stream track. Route only the intended translated source unless bilingual output is deliberate.

FAQ

Common questions

Reach listeners in their language.

Start with a private rehearsal, verify the complete audio path, and then take your multilingual broadcast live.