Need Advice on Voice Clone AI. Is free online voice cloning safe?

Descript is a powerful, all-in-one audio and video editing application that works very differently from traditional editors like Adobe Premiere. Its core philosophy is: edit your audio and video by editing the text transcript.

Overdub is Descript's AI voice cloning feature. It allows you to create a digital replica of your voice (or a voice you have permission to clone). Once created, you can type any sentence, and Descript will generate audio of that sentence in the cloned voice, which you can seamlessly insert into your project.

The primary use case is correcting mistakes without re-recording. For example, if you flub a line in a podcast, instead of going back to the microphone, you can just type the correct sentence, and Overdub will say it in your voice, matching the tone and quality of the rest of your recording.

Step 1: Create a new overdub voice

Step 2: Train your voice model. This is the most critical step for quality

Step 3: Wait for training. After you upload your audio samples, Descript's servers will train the AI model. This process can take several hours. You'll receive an email notification when your Overdub voice is ready.

Step 4: Using your cloned voice in a project. Once your voice is trained, Descript will then generate the audio for your typed text using the cloned voice and insert it into the timeline. The transition will be handled automatically.