Turning Knowledge Into a Screencast
When I record educational videos, the goal is high value-to-minute content. That means no time wasted on intros about what we'll learn (the title and description cover that) and no "smash that like button" requests at the end. I start straight into the material and wrap up with a quick review of what was covered.
Start With the Finished Example
The process begins with a completed example of whatever concept I'm teaching—often based on something from real work. I then strip that down to its fundamentals, removing distractions and unrelated concepts so the focus stays tight. This reduction process is instructive in itself: it forces me to clarify my own understanding and sometimes corrects assumptions about the material.
While simplifying, I make sure not to remove anything that people would commonly encounter in practice. If something gets cut that people will likely hit anyway, I keep a mental note to mention it during instruction.
Work Backward to the Starting Point
Only after I have the finished version do I construct the starting point by removing the important parts while leaving irrelevant scaffolding in place. For example, if the lesson isn't an introduction to an abstraction or module, I'll leave import statements in so viewers don't have to watch me type them.
In my workshop material, starting points live in "exercise" versions of files, marked with emoji (🐨) to guide learners. Accompanying markdown files provide details and links to prepare people for the exercise.
Break Out Extras
Each video stays focused on one core concept. I scan the finished version for things that are only tangentially related to the main lesson—those get their own video or exercise. In workshops, that means they become "extra credit."
The term "extra credit" is a bit misleading there. Some important lessons live in those sections. They're labeled that way not because they're optional in importance, but because in a workshop context, learners don't need to complete them independently to accomplish the exercise's purpose: getting their mind engaged with the problem early on, not grinding through every piece solo.
I also assess how heavy a concept is. If a lesson would run past roughly six minutes, it probably covers too many things at once and should be broken into smaller parts.
Pre-Recording Prep
Most of what I record is material I've taught many times before—in workshops, blog posts, or talks. Being deeply familiar with the content means I can explain it confidently without fumbling. When a video is genuinely the first time I've explained something, that's fine; it just means several retakes before the explanation lands right.
Before hitting record, I set the screen resolution to 1280x720 with HiDPI (effectively 2560x1440 with larger controls). I adjust editor settings so the font is legible even on mobile, and there's no wasted space on screen—often that puts editor and preview side by side. Notes go on a second screen if needed. Frequently the only reference I need is a diff of final versus starting version, via git diff --no-index before.js after.js. I'll occasionally jot down small notes, but never a full script.
The Recording Session
I record video and audio together, not separately. That takes practice but yields a more natural feel. I also make sure the cursor isn't in an awkward spot and disable cursor blinking before starting.
If something goes wrong mid-take, I lift my hands off the keyboard and trackpad, mentally rewind to where I want to restart, restore everything to that state, and go again. Scrolling complicates restarts, so sometimes creative cleanup is needed—or an all-new take when things get too tangled. If the typing is good but the narration isn't, I'll repeat just the spoken part until I get a clean take and overlay it during editing.
Editing and Export
Screenflow handles both recording and editing. These days most videos get edited for me, but I still set audio to mono—a single speaking voice doesn't need stereo—and use the "Smooth Volume Levels" feature to normalize sound. Video exports as 2560x1440 MPEG-4 at 30 FPS.
Further Reading
There's plenty more to the craft, but two resources cover the rest well: John Lindquist's course on recording badass screencasts, and the egghead instructor guide. Both are packed with useful ideas even if you're not planning to publish on that platform.
