How I Make Notion Tutorials: Apps, Tools, and Process!

High-quality software tutorials come from treating audio, screen capture, and face-cam footage as separate, editable assets—then assembling them in post with a disciplined workflow. The process centers on recording voiceover in Adobe Audition for cleaner, standalone WAV files, while screen activity is captured in OBS Studio at 4K. That split matters because it protects the final edit from common failure modes (like audio issues in OBS) and gives maximum flexibility when syncing, cutting, and rebalancing in Premiere Pro.

On the screen-capture side, OBS is configured to record only the exact Notion area that should appear to viewers. Instead of relying on heavy “punch in” editing later, the Notion window is positioned to fit the OBS frame, and the Notion interface is zoomed using Control + Plus until the UI is readable at typical YouTube or phone viewing sizes. The sidebar is collapsed for this specific tutorial, and the zoom level is chosen by mentally simulating how a viewer sees the smaller player window. To make interaction cues unmistakable, two helper tools are used: Mouse Highlight, which visually tracks the cursor and also mitigates an OBS cursor invisibility bug in text-selection mode, and Carnac, which overlays keyboard shortcuts (like Control + C) at a consistent on-screen location.

The face-cam approach also prioritizes edit control. Rather than recording face footage with Loom, the workflow uses a Canon C70 camera with a 20 millimeter lens positioned above the monitor. This produces higher-quality video and—crucially—lets the face-cam be moved, resized, masked, or even temporarily hidden so it doesn’t block UI elements. The assembled edit can therefore show the face cam larger when narration needs it, or cut to face-only when the UI isn’t the focus.

Lighting and audio gear are built around reliability and sound quality. Aputure 300D Mark 2 lighting with the Light Dome and grid provides soft diffusion and directional control, while a Shure SM7B handles the main voiceover through a Focusrite Scarlett 18i8 audio interface (with a Cloudlifter used to boost gain for the SM7B). A Schoeps CMC641 (Hyper Cardiod) shotgun-style mic sits as a high-end backup and can be used when the main mic shouldn’t appear in frame. The lighting stack includes a reflective bounce card for facial fill, a Godox light with a Furnale modifier for hair light, and Nanlite Pavotubes for colored background ambiance.

After recording, footage is organized using a centralized server folder structure tied to Notion. A numbering system links Notion “completed projects” to the correct server assets, enabling quick retrieval of B-roll from other projects. Editing begins from a pre-built Premiere Pro template (Creator’s companion-style defaults), which already includes intro stings, titles, and common elements suited to TF explains’ more formulaic structure.

Because Premiere Pro can be slow with high-resolution media, the workflow uses proxies: ProRes-quality 4K screen and 4K face-cam files are edited via lower-resolution proxy versions, then swapped back to full quality at the end. Audio previews and WAV forms are generated to sync multiple audio tracks, and the face-cam is masked and positioned using copied clip attributes. The result is a tutorial timeline that’s ready for final cutting, with the template and asset structure reducing repetitive setup work.