Meeting transcription only seems to record my audio

[dccannon]

Hi there,

I'm sure I'm just not understanding how to configure this. I tried out the meeting transcription feature today while on a Google Meet call, and it just transcribed my voice and didn't capture anything from the other meeting participants.

I'm not finding any settings to control what device it's getting audio from.

Any help?

I'm on Linux Mint, installed via deb. Using my webcam as my audio source for text transcription, but the meeting audio would be via my headset (used for speaker and microphone). Like I said, it captures my voice but no one else's.

[gabrielste1n]

@dccannon is this still the case on 1.6.10 for you?

[dccannon]

Hey @gabrielste1n sorry I didn't get notified about your message. Still the case on the latest version as well. I just upgraded to 1.7.0 and tried the meeting recording feature again. It's still only recording my microphone, not the system audio. I'm not seeing any settings to control where it records from. Seems like a really useful feature if I could get it to work!

On Linux Mint, installed via the 1.7.0 deb file.

Any ideas?

[Sir-Sloth-The-Lazy]

openwhispr always listens to a single ALSA/PulseAudio source.
To capture the whole conference you have to point it at the monitor of the sink that is playing the remote participants (the headset’s output), not at the microphone you use for speaking.

# install the GUI mixer if you don’t have it
sudo apt install pavucontrol

# 1️⃣ Open PulseAudio Volume Control
pavucontrol &

# 2️⃣ In the “Recording” tab you’ll see “OpenWhispr - Audio Input”.
# Click the dropdown next to it and choose
# “Monitor of <your headset name>” (or “Monitor of Built‑in Audio Analog Stereo” if you route the meet audio there).

# 3️⃣ In the “Input Devices” tab make sure the microphone you want for speech
# is still set as the default source - OpenWhispr will mix both streams.

If you’re starting the app from the terminal you can also specify the source explicitly:

[dccannon]

Thanks for the help @Sir-Sloth-The-Lazy! Sorry that I didn't see your message here.

I have pavucontrol installed. While I'm recording a meeting, I do see a "Chromium input" on the Recording tab. It looks like that does control what OpenWhispr is monitoring.

If I change the dropdown to be my headset output audio, now it records my system sound, but it doesn't record my voice. How can I get it to transcribe both my microphone input and the system audio at the same time?

I would expect that it would use my my microphone preferences in tandem with the system audio (it would transcribe from two sources), but that doesn't seem to be the case.

[gabrielste1n]

This should be resolved in version V1.10.0.