Separate capture from the final loop
The interface records a longer source, then selects a short output window. A detailed timeline, frame-step controls, crop tools, captions, and an optional image overlay make the chosen moment editable. Opening a local video provides an alternative to screen capture.
Capture duration and final GIF duration are distinct controls. That avoids making the user record an exact half-second performance just to create a short reaction.
Own the export pipeline
The editor uses browser media APIs, Canvas, and an original GIF encoder. It supports forward, reverse, and boomerang playback, a shared palette across frames, and explicit frame-rate and dimension choices. Encoding runs as cancellable background work and reports actual file size.
The source states that media stays in the browser. There are no upload endpoints or FFmpeg dependency. Copying a still image is clearly distinguished from copying an animated GIF.
Cropping is nondestructive: reopening the crop tool changes the selection over the original recording, and canceling the preview discards only the pending crop edit.
Make limits visible
An animated GIF used as an overlay becomes a still, output dimensions do not upscale a smaller source, and file size depends on content. These details help users choose a useful export without guessing. An included demo exercises the editor and encoder without requesting screen access, while the original recording remains separate from the finished loop.