To start with Descript, you need to feed it some audio. The software can record your podcast, or you can feed it an audio file. I gave it an old podcast I recorded and let it rip.
Descript‘s machine transcription did a very good job of transcribing the podcast, showing an error about every other sentence. It was also able to split the two people talking in the podcast.
Once the podcast is loaded, then the fun can begin. There are two editing modes: transcription and audio.
When editing the transcription, it works much like a word processor. You can click and hear the audio, and then type to correct what was heard. It goes pretty quickly, and may I say, it’s almost fun.
The magic happens when it comes time to edit the audio. Taking out the ‘ums’ and ‘uhs’ is as simple as highlighting the word and hitting the deleted key. Was there some extraneous words in a sentence? Highlight and delete. It works very well.
Descript has a Mac, Windows, and Linux version. There are limited options when using a browser, so students on Chromebooks or iPads will need access to a machine running a desktop operating system. A lot of Chromebooks can run Linux apps now, but I haven’t tested it. Linux support is still in beta, so I wouldn’t rely on it either.
Pricing is reasonable. Three hours of transcription is available for free. The next tier is $10/month/user which gives you access to everything but team features. Final tier is $15/month/user for team features. Paid accounts also offer features such as timeline support for video editing production.