Upgrade Voice Cloning Module

I notice the more material you give the better, but it should be manageable to get it all done pretty quickly.

Back when I was planning on having a huge voice library and cloning voices these are the things i was going to implement.

Maybe around 15 or so different recordings.

Each recording would have a script that we give them. The scripts should be generated in a way that will last around 30-60 seconds and include various wording structures and grammar to get as many different sounds as possible. Having a script guides the user and gives less error for them trying to think of something random on the fly and it sounding like crap.

Giving them instructions on their environment and equipment they are using. Fo example, saying that they should try to be in a room that does not echo, have a microphone, not just talking into a speaker or something, to sit up straight and speak clearly, to turn off any fans or turn off all background noise.

Heck can even recommend downloading krisp.ai even if just for the free trial. And suggest a good (cheap) microphone on amazon if they don;t have a mic or something.

Then the SYnthflow UI should take them through a flow where it says Recording 1. Then you press play, and synthflow displays what the user should say. Then press stop when done. Then they can listen to the recording and either re record or move on to the next recording. And do that for maybe 15 or so recordings.

These voices should also be able to locked into the workspace, or give the ability to share to other sub accounts easily, or even placed in the marketplace you guys are building.

There should also be a Consent box as you must get consent to clone a voice. You should also have an easy way to display the voice ID.

Feature Type
-

Please authenticate to join the conversation.

Upvoters
Status

Backlog

Board

πŸ’‘ Feature Request

Date

9 months ago

Author

scott

Subscribe to post

Get notified by email when there are changes.