May 27, 2026
Lights, camera, extraction
Reconstructing a Mixbook movie from its data API with FFmpeg
No download button, no problem — commenters split over clever hack vs just screen-record it
TLDR: A Mixbook user discovered there was no downloadable movie file at all — the site rebuilds the video live in your browser — so they reconstructed it from the service’s own data. Commenters were torn between admiring the clever workaround and saying, bluntly, they could've just screen-recorded it.
A simple missing download button turned into a full-on internet detective story — and the crowd had opinions. The original post walks through how one Mixbook user tried to save a movie made from their own photos, only to discover there was no actual video file sitting on the site at all. Instead, the browser rebuilds the movie every time you hit play, like a flipbook assembled on the spot. So naturally, the author went snooping through the page, followed clues into another app, found the hidden data source, and used it to recreate the movie outside the website. Very resourceful, very "I will get my file one way or another."
But the real fun is in the comment section, where the reactions split instantly into two camps. One side was genuinely fascinated by the reveal that this wasn't a normal downloadable video at all. As ge96 basically marveled, the weirdest part is that it's "not a blob" — in other words, not the kind of ready-made file people expect. The other side came in with the ultimate buzzkill practical take: why do all this when you could just record the screen and move on? That deadpan energy from NDlurker gave the whole thread a classic online drama vibe: ingenious hack or glorious overkill?
That tension is what makes this story pop. Some readers saw a clever act of digital self-rescue from a platform with no export option. Others saw a very fancy route to the same result your laptop can do in two clicks. Either way, commenters turned a niche reverse-engineering post into a mini referendum on the internet's favorite question: should you do the smart thing, or the easy thing?
Key Points
- •The article found that Mixbook does not provide a downloadable rendered movie file because the browser builds the animation frame by frame during playback.
- •Inspecting the public preview HTML did not reveal any media URL; the page loaded as a JavaScript application shell backed by client-side API calls.
- •The author identified `memories.mixbook.com` as a separate Next.js app and traced the preview flow to an `AnimatedProject` component.
- •By examining referenced JavaScript bundles, the author located an API endpoint: `/api/v2/my/animated_projects/{projectId}?vk={viewKey}` on `https://www.mixbook.com`.
- •The endpoint returned JSON describing the movie structure, including duration, music, 43 animation segments, and 42 transitions, rather than an MP4 or streaming manifest.