Serious diarization issues
Multiple instances of other people being labeled as speaker zero even when I have no other speaker tags and a voice profile added. Reviewed audio files and something is chopping up the files and cutting speech parts off and that directly aligns with the changes back to speaker zero when it's not me.
At this point the way this is working it is no better than a personal note taking device where only I speak to it. My old compass AI device did that just fine. I want something I can build off of that will actually improve my life but I can't build off of broken transcripts and a device that can't recognize when I'm speaking vs someone else when my voice profile is the only one it has to go off of.
What is being said and who said it is the closest thing to truth the whole system has to go off of. This should be the foundation, it should be rock solid before anything else.
This request was merged into another request
Comments3
MooseSoup
Aug 11, 2025
Do I have a defective unit or is this all software?? I just had a 6 minute phone call on speaker. Raw transcript only says 1 minute 30 seconds. Transcript in app says 1 minute 9 seconds. Cannot tell the delifference between me and the caller merged some of what I said and what he said together and finally identified my voice on the last snippet it took of me. Then I go and look in my Google drive and listen to the audio and there's 2 30 second audio clips. Both are exactly the same clip…. I just don't know what to do anymore.
Nathan Sudds
Aug 10, 2025
@MooseSoup can you share more here about what you were sharing on the community meetup, that you felt once you added Google maps Voice and some other people you felt it started getting worse at the diarization.
Also can you share more about the phone you are using as well? To confirm.
MooseSoup
Aug 10, 2025
I'm using an iPhone at work and an android at home. Same Omi, same account, just turn off Bluetooth on the one that Omi shouldn't be connected to and turn Omi on and connect.
I don't think it's related to adding the speaker tags anymore. I started listening to the audio files and there's skips and cuts all over the place and when I can track down the skips and cuts it almost always lines up with where the new speaker zero (that isn't me) gets labeled. I don't know why the audio has cuts and skips in it especially cutting off actual speech not just cutting out silence.
On top of that the whole issue of my voice transcript being added to my daughter’s is just because the audio file cut from her to me real quick with no pause whatsoever so the stt probably didn't even have a chance. I don't know if there's timing that these things have to stick to but what I'm hearing on my audio is all messed up and it all lines up perfectly with what I'm seeing in my transcripts.
If I have a speech profile and no other speakers tagged I don't think I should have to go back and retag myself in over half of every transcript.