Editing video podcasts with multiple 4K camera angles

DECEMBER 8, 2022

Releasing a podcast involves several steps before and after the interview. Before, it's mainly about booking your guest and preparing the topics you want to cover, or simply to write an outline if you're recording by yourself.

I've been recording interviews on video lately, and the latest episode and the following four, at least, will be released as an audio podcast and as a long-form video on YouTube.

For a recent interview with Alex O'Connor, I've edited the audio tracks following the same steps I always use. I edit each mono track separately in Adobe Audition and apply minimal noise removal, compression, limiting, and normalization filters. The filter that I discovered some time back that significantly improved my workflow was applying an Audiogate, which keeps high gain voices while removing the low ones. This helps to not only mute ambient and background noises but also to remove "bleeding" sounds from other speakers into a speakers' microphone, which prevents transcription services from creating duplicate transcripts and improves audio quality, as often secondary microphones get voices of other people with certain echo—they're far apart from other mics than from their own.

Having multi-cam video in 4K (2160p) makes the workflow slightly more complicated than it needs to be, but it makes it easier to edit and produce an entertaining video. I've tried Descript and DaVinci's multicam functionality and I'm happy with both. Yet Descript makes it easier to collaborate with other editors online, while DaVinci uses raw files, which can be huge if you're recording in 4K ProRes 422. (One out of the two video tracks for the latest podcast is 445 GB.) That makes it hard to share these files via cloud-sharing services.

Once the multicam project is set up, it's a breeze to edit. Plus Descript has a beta feature to auto-assign the correct camera angle based on who's speaking. Remember, Descript knows who's speaking as it's able to transcribe and discern from multiple speakers even when their audio is flattened in a mono track. In my last episode with Zach Kron, I ended up having to manually select which camera was to be shown when. But this is a feature that mostly works great.

I'll continue to share more as I polish this workflow, which involves collaboration with other editors that take care of cleaning up the interview and make it smooth.

My end goal is, of course, to release myself from as many steps of the post-production process as possible.