I got the idea to do the Framed podcast series around December of last year. Someone had sent me a link to a podcast on Kit’s case on YouTube. Like several other podcasts done on the case, this one painted Kit as a monster guilty of murdering his three neighbors. And like many of those other podcasts, there were numerous factual and logic errors as well as omissions of important details that pointed in an entirely different direction.
I had a hard drive full of content: courtroom footage, images, secretly recorded audio, my recorded interviews, documents, and more. Why not put out my own podcast using all of this source material, I thought.
It was about this time that I decided to enlist the help of an expert. That expert was Microsoft Copilot. For those who may be unfamiliar with Copilot, it is Microsoft’s AI application. I ran the idea by Copilot along with a few suggested titles. There were other podcasts with Framed as the title but adding – The Kit Martin Story – as a subtitle was enough to make the title unique, according to my expert.
With Copilot solidly behind the endeavor, I looked into renting a podcast studio. That was quickly ruled out due to cost ($150/hr) and lack of flexibility. I settled on using a greenscreen. I next had to decide on a camera. The options were my webcam, an Osmo 3, or my Lumix GH6. The video tests from my webcam showed that it did not have the quality I needed. The Osmo 3 had descent quality, but I could never make eye contact with the lens. It always looked like I was looking off screen. The best camera for the job was definitely the 35 MM GH6. Copilot suggested various camera settings such as white balance, format, etc.
Next, I had to write a script with some narration taking place on camera and some off camera. I ran every word through Copilot. Once I had the script for the first episode, I decided to record and edit the episode in order to iron out the process for future episodes. I knew going in that I would be combining video from various sources with different resolutions and formats. Copilot guided me all the way.
Here is a short video showing what the process was like.
Once I had a decent take, I edited everything in Adobe Premiere Pro. Once again, I had an expert one click away.
After I got through the first episode, subsequent episodes went much faster. From idea to completed series was right around four months. I had a couple of test screenings. I went back and reshot a number of scenes because of errors caught by those early screeners. I gave Copilot my three options for releasing the series: April, May, or June. Copilot suggested May. I would get an early bump from people who knew about the project, then in June I could have a second push building off the momentum gained in the May release and tie the series to the five-year anniversary of the trial.
I had been using stock music from Adobe for all episodes. I was working on episode five when I stumbled across a music video on YouTube from an artist named Kelly Boesch. I wrote about her in a previous post about the Framed theme song. When I asked Copilot about Kelly, I learned that both her music and videos were AI generated. Copilot even gave me a list of platforms that Kelly uses. So, I signed up on one of those AI music generation sites and tried it out. After just a few hours, I knew I was going to redo all of the music for every episode. I’ll have more to say about that process in my next post.
The bottom line is that doing all of this on my own was only made possible because of AI. If I had to learn every skill the conventional way, it would have taken me four years, not months.