- Always double check the final rendered video for hickups before publishing. The video speed controller chrome plugin helps a lot with screening.
- Video assets are huuuge. A typical episode takes hundreds of gigabytes of render files. To edit efficiently, I have to squeeze out every byte from my tiny disk. In the long run, I will have to invest in better hardware, which I haven't done in a while because Computers are reasonably fast for most tasks today - except recording.
- After a video is shot, you can not change the narrative anymore. You'll wake up in the morning thinking: "oh God, I totally forgot THAT". But the only thing you can do is add a comment. Therefore, a script is vital to keep track of things I absolutely have to say during the episode, but I don't use word-for-word screenwriting.
- Every programming mistake during the recording is very expensive. In post, you have to decide whether to cut it out or leave it in there. Cutting it out can be tricky, as you might only discover your mistake minutes later and leaving it in might confuse or even frustrate your viewers ("why do you waste my time???").
- If I have a long block of uncut material in my video editor, it's probably not because there are no hickups, but because I forgot to edit that.
- Never let the recording equipment get in the way of the narrative. In other words: shoot with what you have. I wanted a second camera but didn't want to spend a lot of money on it. So I just use an old webcam for now. The quality might not be perfect, but it's still better than not having it. I'm also shooting alone and sometimes I would love to have somebody zoom into a scene when I explain things. For now I'm simply emulating this effect in my Video editor using crops or a Ken Burns effect. This will dilute the recording quality a bit, but it adds so much depth and liveness.
- Adding features as I go: for example in the beginning I manually formatted the show notes.
Seeing that this took so much time, I wrote a small CLI tool to render it from
a YAML file. Over time, I added more output formats for the same content:
YouTube, Twitter, the
READMEand so on. The important part is that I did that over the course of many episodes instead of doing lots of work up-front. - From day one I took down all observations and learnings. Many things feel obvious to me now, but they were sure not when I started out. I'm thankful for taking notes, because I could never go to a beginner mindset, even if I wanted to. I want other people to have an easier time when learning how to make their own show.
-
Doing a single episode takes me ages. I would never have thought it takes so much time. Here's a rough timetable:
- Topic preparation (2-5 hours)
- Audio/Video setup (30min-1h)
- Recording (1h-5h)
- Writing show notes (30min)
- Editing (10-20h)
- Publication (1h).
- Creating a title image (15min)
- Updating the website (15min)
- Sharing on social media (15min)
- Feedback and questions (15min)
-
It's pretty easy to find excuses for skipping an episode: live gets in the way, there's more important stuff to do, the content doesn't feel just right yet, there was a technical problem while recording,... don't fall into that trap. I try to avoid skipping the schedule at all costs. Consistency is king and viewers reward it. I treat "Hello Rust!" like a TV show: Even if I know that no episode will ever be perfect, I still publish at the scheduled time (which can be pretty stressful).
- There's a ton of royalty free music out there, but it's hard to find because it's not properly categorized or the preview is cumbersome. Right now I use freemusicarchive, although it's not the most convenient to use.
- Metadata is quite important to people: one person suggested to add keywords to the end of the title to improve discovery on YouTube. That was a great idea! People rarely search for the show directly, but instead they find it through keywords.
- Similarly, people were frustrated about the flow of the show: some said I was rushing over some topics, others thought that it was dragging along. What I did was adding a lot more links to entry-level topics for onboarding beginners and adding chapters to the videos so that people could jump to the interesting bits right away. Similarly, I suggest to watch the videos at 1.5x the speed if you feel bored. That's a nice variation to live-coding where you're kind of stuck with one pace.
- Presenting on stage or in front of the camera is a performance art. You need to show people that you got this and that their time is well-invested. The worst thing you can do is bore them, so skip the chase and get right to the point.
-
Bootstrapping a YouTube channel reminds me a lot of bootstrapping a band: you learn as you go and your first shows are probably awful. This soul-seeking takes a while until you find your "inner voice" and your audience.
-
It's tempting to arrange the show like your viewers presumably want it to be, but it's very important that you stay true to yourself. If you want to revamp the show, just do it. Not everybody will like it, but it's your personal project and it shouldn't become an additional job
-
In the beginning, it was hard to let my personality shine through. I'm usually a very enthusiastic person, especially when talking tech, but in the videos I sometimes look dull. Some people even commented that it was "painful to watch". It took me a while to figure out what was the reason: I was scared to reveal my clumsy self. What helped me was watching other people and learning their tricks on how to loosen things up:
Peter McKinnon is a charismatic storyteller. That's the level of enthusiasm I'd like to show. He's using a ton of b-roll to support his narrative.
The background music in Hot Ones is top-notch. It really creates a thick atmosphere.
- The audio and video editing of Masterclass is second to none. Just watch a trailer of Chris Hadfield's masterclass on space exploration to get a glimpse of that.
The gist is, music creates a stronger atmosphere (pizzicato works good for my clumsy style), b-roll loosens up long explanations (e.g. showing the website of a project while I talk about it), camera movements and zooms make the content more lively. Sometimes I wish I had a dedicated camera-person to make that part easier. A second camera might be nice to record additional footage.
I found that not many coding channels think about delivery. Many have great content but they fall short on the execution part which makes the videos less immersive and entertaining. That's why I spent time on learning how to edit. There are tons of nice resources out there like How Does an Editor Think and Feel? and This Guy Edits.
- Like everything else, video editing is a rabbit hole. From multi-camera setups to LUTs, there's sooo much to learn and experiment with. But it's also a lot of fun! It's important, that the main focus is always on the content, though.
- Lighting is also a rabbit hole and so is audio.
- Invest in good gear. If you're serious about this, you will spend a LOT of time in front of the screen, editing text / cutting videos. It's wise to spend some money on a good screen, mouse, keyboard, and so on. Unfortunately, it's a never-ending race to find the perfect gear; and it can get quite expensive. I suggest you look at what sucks the most and fix that first.
- Rendering is still a time-sink in 2018. Plan in some time for that. I usually aim to render the master video over night.