Also need to control pauces between paragraphs.

Should have global and specific controls for this ability.

They speak at an unnatural pace and rhythm. In addition to slowing down the general pace (words per minute), we should also be able to fine-tune in between words.

Make this a reality

Need to slow overdub voice to at least half the speeds. My use case is narration for non native English speakers. Current overdub voices, including my own even when recorded slowly, are generated unusably fast.

I would love to use the pre-built voices to record scripts, but it is too fast to understand easily. It would be great to have some controls on the voice generation like speed (faster - slower), pitch (higher - lower) and tone (excited, sad, somber, happy, instructive, quizzical, friendly, etc.) I am using 11Labs voice generation and importing it into Descript. It adds several time consuming steps to the workflow.

this is the main thing that is preventing me from using it!

We create audio using AI overdub voices from scripts to add to capture video and other visual media. Being able to adjust speed and timing of that audio to help align with the visual is important to our workflow and that is currently difficult to do. We also need this so when audio content needs to be updated that we can readjust the visual to audio alignment.

This is an important feature for us.