What is Overdub?
How do I use Overdub?
To use Overdub, you first need to set up your Overdub Voice. 10 minutes is the necessary recording length to create your Overdub Voice, but for production-ready voice quality, we strongly recommend recording 30 minutes or more of training data.
Once you’ve completed recording your training data, you'll need to wait 12-24 hours for your voice to be trained. Once it's ready, you’ll be able to attach your Overdub Voice to a speaker label inside any Descript composition.
Now, any time you type text associated with that speaker label, Overdub will automatically generate the corresponding audio.
Is there a shortcut for using Overdub?
Yes! Command D (Mac) or Control D (Windows).
What are some ways that I can be using Overdub?
These videos feature several different ways to use your Overdub Voice. We encourage you to get creative and experiment!
Does Overdub work with longer passages of audio, for example, a video voice over?
Yes! The Overdub feature is designed for both editorial corrections within existing audio, and also from-scratch audio creation for long-form narration. When generating sentences from scratch, Overdub tends to work best with shorter sentences. If you’re creating an Overdub sentence with multiple clauses separated by commas and you notice the quality suffering, try experimenting with punctuation — replacing commas with periods, for example.
Can I edit Overdub audio the same way I edit normal audio?
Yes, you can! To do this, you need to convert Overdub audio to normal audio. Do this by selecting the Overdub Audio in the transcript, right-clicking, selecting Overdub, and selecting “Convert Overdub to normal audio.”