The translation keeps repeating itself (audio loop)

What it looks like: listeners hear the same sentence translated again and again, the transcript fills with the translated language instead of the speaker’s, and your usage climbs even though the speaker has stopped.

Why it happens: Glossa’s translated voice, played through the listener page or sent to your mixer or PA system, is reaching the audio input you send to Glossa. Glossa hears it as new speech and translates it once more.

How to fix it

• Feed Glossa from a bus that carries only the speaker’s microphone, for example a pre-fader aux or matrix send, rather than from the main mix.
• Play the translation only to earbuds or to an output that is not routed back into that bus. If you send a listener page to an FM transmitter or assistive-listening system, route that browser’s audio straight to the transmitter.
• When testing with a second device in the same room, use headphones on it, or the two devices feed each other.
• If the speaker switches between two languages, set the speaker language to Multilingual rather than translating the interpreter’s voice back again.