Transcription
Learn about transcription and provider-specific settings
Here is where you can set the default engine for transcription.
Speechmatics is our default transcription provider, however you have the flexibility to choose another provider from dropdown menu. Our supported transcription engines include: Amazon Transcribe, Whisper (Scriptix), and Speechmatics.
Depending on the provider you select, each engine may offer several transcription settings to customize the transcription process according to your preferences. These can be accessed via your Connectors page
Please find the settings currently available by provider:
Speechmatics
Profanity filtering
When enabled, this feature allows you to identify, redact, or obfuscate profanities in the transcriptions. This can be useful in contexts where you need to ensure the language in your transcriptions is appropriate for all audiences.
Disfluency filtering
Disfluencies - a set list of words that imply hesitation or indecision such as 'hmm' or 'um' - can be automatically removed from your transcript. This setting is automatically enabled for all teams. Deselecting this turns off disfluency removal. When disfluencies are removed, they do not show up in the transcript output at all.
Audio Events for non-speech sound identification
When enabled, this feature is designed to enhance the captioning of audio within media by using AI to automatically label non-speech events such as [music], [laughter], [applause], or other ambient sounds. When enabled, audio events are added to the second caption track and can be particularly useful in providing context to the transcriptions.
