PixelBay

Screen recording tools optimise for capture, not for what happens afterwards. They produce a single continuous file, so a fumbled sentence forty minutes in means either re-recording the whole session or fixing it in an edit that costs more time than the recording did.

For anyone recording regularly — tutorials, demos, product walkthroughs — the editing tax is the real cost of the tool, and almost no recorder is designed to reduce it.

I make videos for fun, and I kept hitting the same wall with every recorder I tried. The workaround was always the same: record in short chunks manually, name the files carefully, and stitch them together later.

That workaround is a product. So I built it.

Solo: product definition, interface design, and implementation. Built natively for macOS in Swift, with Claude and Codex as the development tooling.

Three decisions define the product, and each one attacks the editing tax directly.

Segmented scene-based capture — a recording is a list of scenes, not one file. A bad take is re-recorded on its own without touching the rest.

Keyboard-shortcut zoom triggers — zooms are marked live, during the recording, instead of being added by hand in an editor afterwards.

Appending new takes to an existing project — a recording session can be resumed days later and land in the same project rather than a new file.

In active development and in daily use for my own videos. Building toward a public release.