Software
Podcast Editor Descript introduces a new pro level with access to its AI voice dual functionality.
Descript, the maker of artificial intelligence-powered audio editing tools, is launching a new subscription process today called Descript Pro that includes access to its Overdub feature.
Overdub, first published in beta last year, is a synthesis device that allows you to effectively create a double AI voice for yourself. That way, if you make a recording error, the program will correct it while the final product still sounds natural and organic. It's sort of like making a deep-fake audio, but you do it to correct your own mistakes instead of attempting to impersonate someone else.
Descript says Overdub is now an official feature with the launch of Descript Pro, which is available if you opt for a $30-per-month (or $24-per-month) subscription. Prior to today's announcement, Descript offered a free tier, a product tier for $14 a month (or $10 a year) and a team tier for $18 a month (or $15 a year). Both of these paid plans are bundled together with their cheaper package, or what Descript called the Creator Plan, for $15 a month or $12 a month billed annually. This contains everything but Overdub and fewer transcription hours needed.
Besides Overdub, Descript's key selling point is that it re-imagines audio editing as a kind of collaborative Google Docs-style software, so you can edit the podcast transcript or any other form of audio recording as if it were a Word document and have the changes mirrored in the playback. Descript also acts as a video editor and transcription service, which includes AI and human-aided transcriptions up to 30 hours for the Pro plan and 10 hours for the Creator one, with the option of buying additional hours.
removing 199 ums, uhs, likes, you knows, and other filler words from a recorded meeting pic.twitter.com/NpnCTdQRyg
— Andrew Mason (@andrewmason) July 16, 2020
The company claims that its latest pro tier also provides "unique access to advanced word filler handling, custom audiograms, and customizable publishing and export options." Filter word handling is a particularly useful-sounding feature, as Descript claims it can be used to "detect, ignore, and delete filler words like 'you know,' 'like' and 'kind of'" to make audio recording sound cleaner and more competent.
CEO Andrew Mason says to The Verge, "In our internal study, we find that eliminating filler words from Zoom meeting recordings decreases the duration of the recording by around 10%."
