Current download: DepoAudio v1.0.2. Released 2026-08-20, with 4 published download assets whose filenames and SHA-256 digests were reviewed. Review release status →

Choose the problem

Start with the file, end with a specific action.

DepoAudio support is organized around the job in front of you. No recording upload is required to identify a file or use these guides.

01

I do not know what file I have

Use the browser identifier. It inspects only the first 8 KB on your device and gives the most specific evidence-supported next action.

Identify the file locally →
02

The file is SGMCA

Check whether the embedded Ogg stream is recognized, then use the v1.0.2 app for conversion. Keep the original.

Read the SGMCA guide →
03

The file is FTR or TRM

Conversion support is experimental in v1.0.2. Preserve every session segment, verify the complete result, and export from FTR software if conversion fails.

Follow the FTR/TRM workflow →
04

The file is AES or DCR

These formats need their creating or vendor player first. Export WAV, then use DepoAudio for conversion, channel work, or cleanup.

Choose the correct export guide →
05

The output is silent

Play the source in DepoAudio, then try Keep Original or Split Channels. A silent channel may be an unused or muted microphone in the source.

Review channel workflow →
06

I need a WAV or MP3

Use WAV for editing or an archival working copy. Use MP3 when size and broad playback matter; keep the source because lossy export cannot add detail.

Choose an output →
07

My transcript did not follow the audio or persist

Timed transcript lines follow the playhead; plain text needs timestamps before SRT export. Saved text is associated with the selected track on this device, so export SRT or TXT before moving the audio or clearing app data.

Review the transcript workflow →
08

Installation shows a warning

The macOS app was built with an ad-hoc identity and was not notarized; the Windows EXE and MSI are not Authenticode signed. Verify the exact filename and checksum before following platform steps.

Open install instructions →
09

A cleanup or analysis control is unavailable

v1.0.2 bundles its released local-analysis models, while DNSMOS is installed only when requested. Check the model registry for delivery, analysis bounds, platform execution, and known limitations; DeepFilterNet and de-reverb remain unavailable.

Check the capability status →

When a sample is genuinely needed

Protect the record before reporting a format problem.

  1. Try a synthetic reproduction or a non-confidential file first.
  2. Remove names, case numbers, paths, voices, and metadata that are not needed to reproduce the parser failure.
  3. Use the smallest excerpt that still reproduces the problem, only when you have authority to share it.
  4. Assume a GitHub issue and its attachments are public. Describe the format without attaching a record if confidentiality is uncertain.
Open the public issue tracker ↗ (opens in a new tab)