TrackPeel

Under the hood

How a conversion actually runs

TrackPeel doesn't have a server that touches your video. Every step below happens inside your own browser tab, using a real build of FFmpeg compiled to WebAssembly.

01 — Load

Your file never leaves your machine

Selecting or dropping a file just gives the page's JavaScript a reference to it on disk. Nothing is uploaded to any server — there's no upload step at all, which is also why file size is limited by your browser's memory rather than a network transfer limit.

02 — Extract

FFmpeg demuxes the audio track

WebM and MKV containers typically carry audio as Opus. A background worker running FFmpeg (WebAssembly) reads the file's audio stream and re-encodes it to MP3, without touching the video stream at all.

03 — Rebuild

The video is copied, not re-encoded

To produce the final MKV, FFmpeg copies the original video stream byte-for-byte (no quality loss, no re-encoding pass) and muxes it together with the new MP3 audio track into a fresh MKV container.

04 — Download

You choose what to keep

Once a file finishes, you get direct download buttons for the MP3 and the rebuilt MKV. Nothing is written back to your original file automatically.

Why there's a size limit

Browsers run WebAssembly inside a memory space that tops out well below what a modern desktop actually has installed — regardless of your computer's RAM, the tab itself can only address a few gigabytes. That's a limitation of running entirely client-side, which is also what makes it possible to skip a server and an upload step in the first place. Very large files (roughly 3GB+) may fail for this reason — that's expected, not a bug.

A desktop version without this size limit — using a real, native FFmpeg install instead of a browser sandbox — is in development and will be linked from the homepage once it's available.