v0.6.0-beta48
If you have ever recorded a meeting in a room and been told there were no other voices in it, this is why.
Minutiva was listening to the wrong audio
When you record a video call, Minutiva captures two streams: your microphone, and the other side of the call. It separates and names voices from the second one.
To decide which to analyse, it asked whether that second stream had been captured — not whether anyone was on it. In a room there is nobody on it. The capture still succeeds; it just records silence. So Minutiva analysed the silence, found nobody, and reported “no other voices in this recording” — while every voice that was actually in the room sat on the microphone, unseparated.
It now measures which stream anyone was actually speaking on, and listens to that one.
If you see clients in a consulting room, this is the difference between speaker separation and none at all. A session recorded across a desk now comes back with the voices told apart, the same as a video call does.
An empty list explains itself
If there genuinely was nobody on the other side, the speaker panel now says so, rather than repeating a note about system audio that does not apply to the recording you are looking at.
Your existing meetings are untouched
Recordings made before this release were never measured, and are left exactly as they are. “We never looked” is not the same as “it was silent”, and treating it as such would have re-analysed your whole library against the wrong stream.
Auto-update
Existing beta channel testers get this within 24 hours. Or trigger now: Minutiva Private → Check for Updates…