What’s new
Changelog
How DepoAudio got here — the headline changes in each release. The current build is v0.8.0. The full, dated log and every download live on GitHub Releases.
- v0.8.0 Current June 2026
Smart Cleanup goes public
- AI audio cleanup is out of preview and in every build — free.
- Remove background noise (Fast or Best Quality with DeepFilterNet3).
- Reduce room echo and reverb, enhance narrow-band clarity, and de-clip.
- Balance speaker volume and score recording quality (1–5) on Scan.
- Turn detection: who’s talking, how many speakers, speech-to-silence ratio.
- Hardware acceleration on Apple Neural Engine, Ryzen AI, Intel AI Boost, or GPU.
- Model manager: download only what you need, every model SHA-256 verified.
- v0.7
Smart Cleanup preview
- First on-device AI cleanup models, in preview.
- Scan a recording to spot noise, uneven volume, and clipping.
- Everything runs locally through a bundled ONNX Runtime.
- v0.5 – 0.6
Merge, library & acceleration
- Merge multiple recordings of one event — Best Quality or Mix All.
- Auto-filing audio library, organized by case and participant.
- Built-in player with color-coded, editable speaker tracks.
- Automatic detection of installed court reporting software and jobs.
- v0.4
The converter
- Convert SGMCA, FTR .trm, and BWF court recordings.
- Output to WAV, MP3, FLAC, Opus, or M4A.
- Three output modes: stereo mix, keep original, or split by participant.
- Batch processing with flat memory use and no file-count limit.