A walkthrough of Descript's voice cloning feature. Record a short script, wait 30 seconds for training, and generate speech in your own voice. The demo creates a speaker called ROS3 and shows the full process from setup to playback.
To turn a script that you've written like this in descript into an actual voice using AI, click right here and type in a name. So I'm just going to call this ROS3 because I've already trained it on a couple other voices. I'll hit create speaker ROS3, and it doesn't do anything yet, but it just labels it. And if I click on my name, come over here, click on the three dots, and then click on this thing that says enable speech generation. Click on that. And what you'll have to do at this point is read this short script. And this is what'll train the AI on all the different sounds that it needs in order to be able to replicate your script in your voice. So what you'll do here is you'll hit start, record, read this, hit stop when you're done, and then It'll take about 30 seconds to train it. I'll go ahead and do that. I want descript to create an artificial version of my voice that I can use to create speech that sounds like me. I am training my voice by reading the following statement. Every day we tell stories, ask questions, and share jokes. And stop. And then you can listen to it again. You can re record, or it says, by submitting, I authorize descript to enable speech generation. Hit, hit, submit. Authorizing speaker down in the bottom right. And there we go. Speak speaker authorized. And then you'll see the blue text. The text that I typed in or that AI typed in is now pulsing to let me know it's processing. And in just a moment, we'll be able to listen to it. And there we go. It's done. So let's listen. Let's hear how it sounds. Hey there, fellow Internet nerds. Gather around your screens, because today. So there we go. That's my AI voice. If you like this one, be sure to watch this video next.
