Open an issue on the tracker (a free GitLab account is needed to post). It helps to include your distribution and version, the app version, and the steps that trigger it.
The common ones: MP3, FLAC, WAV, OGG, and AAC / M4A / MP4.
It runs offline on your machine. The first time you use it, it downloads a model once; after that it works without a connection.
Yes. Paste a link from many popular sites and it plays the audio. This uses yt-dlp under the hood, so keeping yt-dlp up to date helps when those sites change.
It's a reasonable estimate and usually close, but it can be wrong on some tracks. Treat it as a starting point rather than the final word.
No. Those changes happen live while the track plays, so your original file is never modified or re-encoded. Analysis like BPM and key is remembered in a local database, so it isn't recomputed each time.
It sits between a normal player and a full DJ app. It's built around reshaping one track at a time (tempo, pitch, key, groove) rather than beat-matching two decks. If you're looking for a two-deck mixing setup, there's better software out there for that.
No. It works fully offline and doesn't phone home. The only times it touches the internet are optional: playing from a web link, and a one-time model download the first time you use stem separation.
→ The AppImage won't start
Make sure it's executable (chmod +x) and that FUSE is available. On some distros you need libfuse2; otherwise run it with --appimage-extract-and-run.
→ No sound, or the wrong output device
Check your system audio (PipeWire / PulseAudio) and that the right output is selected there. The app follows the system default.
→ Playback is choppy or stuttering
Real-time reshaping is CPU-intensive, so the usual cause is the CPU in a power-saving mode. Set your system's power mode (CPU governor) to Balanced or Performance rather than Power Saver. Closing other heavy apps helps too.
→ Web playback from a link fails
This relies on yt-dlp. If a link stops working, updating yt-dlp usually fixes it, since those sites change often.
→ Stem separation is slow or runs out of memory
It's an offline, CPU-heavy process, and the first run downloads a model. Give it time on longer tracks; very low-memory machines may struggle.