I make short demo videos for my side projects — the kind with buttery-smooth zooms that follow your cursor. Screen Studio does this beautifully on macOS, and charges $29/month (or $99/year if you commit) for the privilege. That's $348/year for software I open maybe four times a month.
So I spent one Saturday afternoon rebuilding the workflow with free tools. Eight months later, my Screen Studio subscription is canceled and nobody who's watched my demos has noticed a difference.
The Comparison
| Screen Studio | OBS + FFmpeg | |
|---|---|---|
| Price | $29/month | $0 |
| Smooth cursor zoom | Built-in, automatic | OBS "Auto-zoom" scene filter or manual keyframes |
| Platform | macOS only | Mac, Windows, Linux |
| 4K60 recording | Yes | Yes |
| Scriptable/automatable | No (GUI only) | Fully (CLI + ffmpeg) |
| Edit after recording | Basic trim | Any NLE, or ffmpeg cuts |
| Learning curve | ~5 minutes | ~2 hours (one time) |
The honest gap: Screen Studio's automatic zoom is genuinely smarter — it follows cursor intent (zooming to the button you're about to click). My OBS version zooms to fixed regions I define per scene. For a 2-minute demo video, the difference is maybe 20 minutes of extra setup. At $29/month, that's an hourly rate I can live with.
The Setup
1. OBS Studio (free, open source): one scene per "zoom region". I record full-screen at 4K, then use OBS's "Move" transition between a zoomed-out scene and zoomed-in crops.
2. The finishing pass — one FFmpeg command normalizes audio, trims silence gaps longer than 1.5s, and encodes for web:
ffmpeg -i raw.mkv \
-af "silenceremove=stop_periods=-1:stop_duration=1.5:stop_threshold=-45dB, loudnorm=I=-16:TP=-1.5" \
-c:v libx264 -crf 20 -preset slow -pix_fmt yuv420p \
-movflags +faststart \
final.mp4
loudnorm to -16 LUFS is the single biggest "sounds professional" lever — it's the podcast standard, and it's one flag.
3. Keystroke overlay (the part people think needs paid software): I use keycastr on macOS (free, open source) or screenkey on Linux, recorded as a separate OBS source.
The Part Nobody Talks About: Scriptability
Here's what actually made me switch permanently. My demo videos all start the same way — terminal, type a command, show output. I now have a shell script that:
- Starts OBS recording via
obs-cli - Runs my demo script in a tmux session with
asciinema-style pacing - Stops recording, runs the FFmpeg pass
- Drops
final.mp4in my uploads folder
One command, zero clicking. Screen Studio can't do this — it's a GUI app with no automation surface. When you publish demos weekly, the tool you can script wins, even if its zoom is slightly less magical.
Total cost of my pipeline: $0. Total time to reproduce one Screen Studio-style video: ~15 extra minutes.
I prototyped the FFmpeg silence-removal flags by describing what I wanted to MonkeyCode and iterating on the output — the audio filter syntax is notoriously cryptic: https://ly.cyberserval.tech/iIETXiF
What's your screen recording setup — are you paying for a tool, or have you found a free stack that's good enough?
Top comments (0)