Yes, Notta comes with a speaker identification feature. This article summarizes the conditions for using speaker identification and important notes. Please refer to this article if you have any questions about settings or if speaker identification isn't working as expected.
When is speaker identification available?
Speaker identification availability depends on the feature you're using and the transcription language. Here's a breakdown:
| Feature | Details |
|---|---|
| Screen recording | Speaker identification is available if enabled before initiating a screen recording. It supports all transcription languages and can identify up to 10 speakers. |
| Recording of browser tabs via the Chrome extension | Speaker identification is only available when the spoken language is Japanese. It's enabled by default, so users do not need to manually turn it on. |
| Real-time recording via microphone | Speaker identification is only available when the spoken language is Japanese, and up to 10 speakers can be distinguished. It's enabled by default, so users do not need to manually turn it on. |
| File upload | Speaker identification is available if enabled when uploading files. It supports all transcription languages and can identify up to 10 speakers. |
| Online meeting transcription via Notta Bot |
Available for all transcription languages, and there is no limit to the number of speakers it can identify. Identification is based on actual participant accounts in the meeting, so please ensure each speaker joins from a unique account. If multiple participants speak from a shared account, their speech will be grouped under that shared account. |
I enabled speaker identification, but it's still not working.
If speakers are not correctly identified, please first check the speaker notation format in the transcription record.
-
-
If speaker identification is off, all speakers will be labeled as "Unknown speaker".
-
If speaker identification is on, speakers will be labeled in the "Speaker + number" format, such as "Speaker 1", "Speaker 2".
-
If speakers are displayed in the "Speaker + number" format, speaker identification is enabled. However, if you notice discrepancies in the displayed numbers versus the actual number of speakers, or have other concerns about identification accuracy, please send your feedback to the Notta team.
Send feedback ⇒ https://support.notta.ai/hc/en-us/articles/17480242995099
Why did speaker labels disappear after upgrading my plan?
If the audio duration exceeds your current plan's transcription limit, the transcription detail page will display a message indicating incomplete transcription or a banner prompting you to upgrade to a higher plan.
When you unlock the full text after upgrading, the audio data will undergo re-transcription. Consequently, the original transcribed text, including speaker information, may be overwritten by the new transcription result and disappear.
For future recordings, as long as the duration of a single recording is within your current plan's limits, you won't need to manually unlock the full text and can view the complete transcription directly. In such cases, the audio data will not be re-transcribed, thereby preserving your original speaker labels.