Diarization not working - AssemblyAI works perfectly

[rlucredio]

Hi, I have this audio (it's fine to share because it's fabricated using AI voices and fake company names). The diarization simply doesn't work, it outputs all utterances as speaker 0. I used the API playground and also my client app. I tried nova-2 and enhanced models, none of them worked.

I tried some other files which did work partially, and I wanted to do a controlled test using 2 ai generated female voices. The voices are similar but not equal. The diarization works out of the box on assembly AI.

If there's a certain combination of model/parameters I should try, please advise. Otherwise I'll spend more time exploring the competitor instead.

I really like Deepgram's speed and features and would like to use it, but if the diarization fails, it's a dealbreaker for me.

I saw some of the discussions, and in one of them, someone suggested using dual channel audio, but that's not an option for me and my client.

Thanks!
customer-call.zip

[deepgram-community[bot]]

Thanks for asking your question. Please be sure to reply with as much detail as possible so the community can assist you efficiently.
_{Consider joining our Discord community for more opportunity to engage with your fellow Deepgram users. You can earn points which can be redeemed for cool stuff by being active in our communities!}

[deepgram-community[bot]]

Hey there! It looks like you haven't connected your GitHub account to your Deepgram account. You can do this at https://community.deepgram.com - being verified through this process will allow our team to help you in a much more streamlined fashion.

[deepgram-community[bot]]

It looks like we're missing some important information to help debug your issue. Would you mind providing us with the following details in a reply?

• The programming language you are working in (e.g. JavaScript, Python).
• The deepgram product you are using (e.g Speech to Text, Agent API)
• A request ID that triggered your error or issue.

[rlucredio]

• The programming language you are working in (e.g. JavaScript, Python).
  I tested with the API playground direct from the website
• The deepgram product you are using (e.g Speech to Text, Agent API)
  Speech to text
• A request ID that triggered your error or issue.
  No need, if someone tries to transcribe that audio i provided, it'll probably fail. I just need someone with more experience than me to guide me, maybe suggest some sort of parametrization that works

[deepgram-community[bot]]

Diarization can be hit or miss, and sometimes there are combinations to consider to get it work ideally. We are working on new Diarization model that should improve the exp a lot, but I don't have an ETA on that release right now.

Multi-channel is a good option here it's probably the best way to go, but it sounds like that isn't going to work for you.

If you can wait until we release our improved diarization model we'll have better support for this feature, but if you cannot wait, I totally get it.

This message was sent by John Vajda from Deepgram, via our community automation.