Voice styles are only available on classic AI voices that previously had voices created, and will be deprecated in a future update.
You can adjust the intonation of a classic AI speaker with voice styles.
Assigning a voice style
To assign a style:
- Click on the associated speaker label in the Speaker label editor to assign it to the project
- Re-select the speaker, and click the ellipsis button which will now appear to the right of the selected speaker
- Scroll to the bottom of the list and choose from the list of existing styles.
Check out our Creating and using an AI speaker guide for more details on using AI speakers in Descript.
Intended use of voice styles
Please note that voice styles are intended for when you are creating AI speech from scratch, not when using Regenerate with existing audio. When editing existing audio, Descript’s AI automatically creates a custom style based on the audio surrounding the edit so it blends seamlessly with your existing audio.