Hello Product Hunt friends!
Today we are pleased to release the latest and greatest version of Descript, Descript Pro.
Pro features several amazing features, foremost among them Overdub, the text to speech voice cloning technology we announced in private beta almost a year ago. As of today, anyone can upload 10 minutes of audio (although we recommend closer to 30 minutes) to create an ultra realistic text to speech model of themselves... and it's integrated into the Descript editor, which we think brings this tech beyond the level of novelty and into something that's actually useful.
I make the product release announcement videos for Descript and frequently get editorial feedback from my team - with Overdub, I can just delete the words I got wrong and type in new ones - no need to re-record. It's amazing, it really changes the way you think about audio by making it as malleable as text. There's an Overdub in the "Pro" announcement video attached to this post, see if you can pick it out! Actually, don't waste your time, you'll never get it, it's too good!
The second coolest Pro feature (IMO) is advanced filler word detection. Now with a single click you can remove not only um and uh, but "you know,", "like," "so," repeated words, and a dozen other filler words that are contextually sensitive. - i.e. we can tell the difference between "do you know the time?" and "I'm, you know, excited about this feature."
In our internal testing, we've found that you can reduce the length of meeting recordings by about 10% just by a one-click zap of the filler words... not to mention how much more listenable they become :-) That means if you speak 5,000 words a day at 150wpm, you spend roughly 333 continuous days in your lifetime spouting filler words! (It's possible my math is wrong, I did not double check it because the number is so dramatic that I don't want it to be wrong).
The third cool Pro feature is custom audiogram branding. You can add custom backgrounds and change the colors used in your audiograms. You can also remove the Descript branding, if for some unfathomable reason you would want to do that. There are other ways you can make audiograms, but Descript's audiograms are crazy easy / fast to make. And they're a great way to share your Overdub voice!
Descript Pro has a few final features that are less theatrical but we know regular users will appreciate: 1) increased transcription rate limit to 30 hours / user / month, 2) batch export features.
That's all for now. If somehow you read this near 9am pacific today (7/28), I'll be doing a live demo of the new features at http://descri.pt/07-28-20. Otherwise I'll be hanging around answering questions.
Andrew
@andrewmason Descript is awesome. I wish I could remove all the gaps in my audio by one click. I try to record videos without a script and tend to pause between words of a single sentence when thinking what to say next. It ends up being a huge editing labour — which I could remove all the gaps in one go.
I’ve been using overdub in my process for awhile now and I did audio for an entire promotional video using overdub to create my voice cause I’d lost my voice in a week long teaching gig and it sounded so amazing most people couldn’t tell it wasn’t me! Descript has been an amazing tool to have in my toolkit and I’ve enjoyed being an early adopter and can’t wait to see how it continues to grow.
Really hope you guys make a linux version soon. (already voted on the feature request).
There's never been an easier time to make linux apps.
(Or work with Wine to make it compatible)
Product Hunt