Skip to content

$ cat ~/blog/depoaudio-court-audio.md

DepoAudio: getting court audio out of jail

Why I built a free, open-source desktop app that converts supported court audio locally, with experimental FTR/TRM handling.

updated Note 3 min read #depoaudio #legal-tech #open-source

If you work anywhere near litigation, you've lived this one: someone hands you a disc — yes, a disc — of courtroom audio, and ordinary media software does not give you a dependable complete session. .sgmca. FTR/TRM blocks. Multi-channel BWF with a sidecar you're apparently supposed to know about. The hearing happened, the record exists, and you may still need the originating player or a verified conversion path before you can use it.

I work at a court reporting firm. We produce the record. Watching the record get held hostage by file formats finally annoyed me enough to do something about it.

September 4, 2026 correction: this June introduction overstated Merge availability, installer signing, macOS support, and the app's network boundary. The details below now follow the published v1.0.2 release record (opens in a new tab), rather than unreleased source or intended behavior.

the rules I set

Court audio can contain confidential or otherwise sensitive material, so I set three constraints:

  • Recording processing happens on your machine. Core conversion and cleanup run locally. The reviewed v1.0.2 app has no recording-upload path; optional model installation downloads assets from GitHub without sending the recording. Your own backup and sharing settings still matter. The privacy details (opens in a new tab) spell out that boundary.
  • No accounts. You shouldn't have to create a login to listen to a hearing you were at.
  • Free and open source. This is plumbing. I do not want an account or subscription between a reporter and the recording.

what it does

DepoAudio converts supported SGMCA, BWF, and standard media into formats the rest of the world actually plays — WAV, MP3, FLAC, Opus. FTR/TRM handling is experimental because delivered variants differ, so detection is a starting point and every result still needs verification. When a variant cannot be identified or verified cleanly, use FTR Player or the originating system's approved export route. The app can also clean up recordings, and it keeps a local library organized by case and session instead of AUDIO_0047_FINAL(2).zip. Merge is not a supported workflow in v1.0.2.

the stack

Tauri 2 with a Rust backend and a React front end. Rust handles format detection, builds the FFmpeg filter chains, and persists the library; Tauri keeps the whole thing small and native instead of shipping a browser in a trench coat. The published v1.0.2 macOS build is ad-hoc signed and not notarized; its Windows EXE and MSI installers are not Authenticode signed.

why open source

Because “trust me” is not a security model you should accept from legal software. The whole thing is on GitHub (opens in a new tab) — auditable, forkable, yours to keep. If your IT department wants to read every line before it touches a deposition, they can. Honestly, they should.

The v1.0.2 release targets Windows 10 or 11 on x64 and macOS 13.3 or later, with a universal Mac build. Check the install guide (opens in a new tab) for platform details. If there's a format you're stuck with that it doesn't handle yet, the repo takes format requests — that's half the roadmap.

If you're deciding what kind of recording you received before converting it, start with the plain-English court-recording format guide.

— Andrew