In this issue, I’d like to highlight my current system for editing my videos. In the best of times, it’s fun to develop a great visualization of what I’m talking about in my voiceovers. Most of the time, it sucks, and it’s the bottleneck of my workflow. By self-examining my editing workflow, I hope to figure out possible solutions and speed ups.
Step 1: Voiceover
Before I start any edit, I make sure to have a rough script completed. Then I do the raw voice over in some place quiet, like my car or a small fort made from sofa cushions. Not the most professional thing ever, but it gets the job done. I use Adobe’s podcast AI to help remove unpleasant noises from the raw audio and add some effects to it in edit to make it sound more even.
Once the raw voiceover has been processed, I start to cut down the audio to a more clean cut. I once tried using AI to do this part as well, but I couldn’t get the software to work quite right for my needs. I realized that having the cuts in the audio clip is useful for figuring out where my sentences start and end in the edit. Possible speed up here would be to suck it up and figure out how to get the AI to cut out the dead noise while keeping the edit, but it would be nice if this fix didn’t boost fixed costs for the company.
Step 2: The Slog
When I write the script, I often get ideas for what I should be showing on the screen to help show the point I’m trying to make. I don't really write any notes about it; I just hope I'll remember it when I edit. In other words, I don’t usually have visuals planned out in advance.
I just listen to whatever part of the audio I’m currently on, and I spend some time thinking about what would be a good visual for it. It’s worked… for now, but there are so many problems with it.
There are several occasions where I get to a point in the voiceover, and I have NO idea what to put down. This usually happens with more technical points of the video, like explaining probabilities or why a mathematical expression looks the way it does. I literally get stuck at a single moment for what often amounts to a half hour of wasted time.
My easy solution is to keep whatever is on the screen there for a few more seconds. But another rule I have is to try not to have the screen be static for more than 4 or 5 seconds, so this solution only takes me so far.
A harder solution would be to storyboard the entire script before I get to the editing phase. I've actually tried it before in the past, but sticking to a consistent upload schedule put too much time pressure on me. Now that my writing habit and new upload philosophy has made things easier, it's probably a good time to revisit this idea. I spend a lot of time finding visuals while I'm editing anyways, so why not just dedicate an entire phase to it.
That being said, I don’t know how I would implement this. Perhaps Figma would be a good choice here? Or google docs? I need to get text and figures together, and these figures will probably be hand drawn for speed.
I used to edit my videos in one monolith of a timeline. This created some inefficiencies because I would often need to find a clip that I previously put together. With timelines of 10-15 minutes, this became a huge time sink. Furthermore, I found that my computer could not export videos longer than 17 minutes. In response to all this, I edit all my videos in 3-5 minute sections now. It feels good to reach the end of a section and gives me some short-term motivation to edit for just a bit more.
Sometimes, I make major changes to the voiceover in the editing phase. I’ll do this when I feel like I could have described a teaching point better. I try not to do this a lot since it’s a big drag on time.
Step 3: Review
It might surprise you to hear that I previously did not review my videos after I finished the edit. I would be so sick of editing that I just wanted to get it out and uploaded onto Youtube. But nowadays, I require myself to give the video a single watch before I export. There’s always 2-3 small errors in this review, so it’s worth the extra time. In the past, I would have left these in, but I’m trying to be less compromising in terms of video quality.
Once this is done, it’s time to upload! Whenever I finish editing, I always schedule the upload to 10am the next day. I don’t upload consistently, but I do set a consistent time that a video goes up, based on where I’ve noticed my audience is.
If you also do your own video editing, I’d love to hear what kinds of adjustments you’ve made to make your edits better or more efficient. Thanks for your time, see you next week.
Christian
😵💫 What am I working on right now?
Advancement to candidacy. My exam is next week, so I’m practicing a presentation to convince my committee that I can plan a first-author paper.
Simulations for a paper in revision. Hopefully this one gets in! It would be my first first-author paper if it succeeds.
A video on “Correlation does not equal causation.”
🧐 What am I enjoying right now?
Books — I finished Million Dollar Weekend (affiliate) by Noah Kagan recently, and I’ve been listening to it in the background. My goal is to produce my first class this year, but I have no idea how to spread the word. The idea of “selling” still feels icky to me, but if I don’t advocate for my stuff, then no one will.
📺 Recent videos
Explaining Confidence Intervals and The Critical Region: an explainer video on the critical region/confidence interval method of making decisions for hypothesis tests. A (better) alternative to Fisher’s p-value method, and a necessity for understanding standard statistical analyses in research.