As a general rule I think adding music to a project as early as possible makes your edit better, because you can fit the speech around the music and vice versa, rather than waiting until the very end of the edit to add music into a finished edit. The ducking tool helps with that.

Audio of the draft script was temped by an AI narrator, which meant we could record the script just the once, after we'd refined it; 'ducking' meant Descript would automate music levels around speech, giving us a rough automix; and the basic EQ/compressor made it an easier listen.

3) I've used these for years, but I think they're still underused: the audio effects panel (ducking, high-pass, compressor, limiter) and AI voices. Used together, they meant we could send rough edits for feedback which sounded much closer to the finished product.

2) The 'remove retakes' feature. The series running time is 272 mins and we have a lot of script, recorded over about 7 hours in the studio in total, so 'remove retakes' + 'shorten silences' really helped speed up script editing.

Having video which was vision mixed live, and then working on a single video and audio edit up to the last stage of the process, hopefully saved time. It's the first time I've worked like this but it was pretty straightforward.

… this meant that once our edit was done, it could be exported from Descript both to a Premiere timeline (for Sky colleagues to produce the video promos) and to Reaper, where I could swap the video for the original isolated audio stems before sending it to the sound designers.