Skip to content

Releases: marirs/tono

Release list

Tono v0.1.0

Choose a tag to compare

@github-actions github-actions released this 27 Sep 16:21

Tono turns your music into instrument-specific practice videos with backing tracks, fingering guidance, a count-in and adjustable practice speed.

Downloads

Choose your operating system and processor. Offline ZIP includes Tono, Python, FFmpeg and the ML models. Extract it and keep its files together. Small executable downloads its matching runtime on first use.

Platform Small executable Offline ZIP
macOS Apple Silicon Download Download ZIP
macOS Intel Download Download ZIP
Linux x86_64 (glibc) Download Download ZIP
Linux ARM64 (glibc) Download Download ZIP
Windows x86_64 Download EXE Download ZIP

Native Windows ARM64 packages are not available. macOS binaries target macOS 14 or newer; compatibility with every OS version has not been validated. Packages are not yet Apple-notarized or Windows code-signed.

Run

Tono currently uses a terminal. Rename the small executable to tono (tono.exe on Windows); on macOS/Linux, run chmod +x tono if needed. The offline ZIP already uses this filename.

./tono song.mp3 --instrument ae01 --tempo-scale 0.75
./tono song.mp3 --instrument ae20 --easy-fingering
./tono --help

On Windows PowerShell, use .\tono.exe instead of ./tono. Initial setup needs several GB of free space; the small executable also needs an internet connection. No separate Python or FFmpeg installation is required.

Included

  • Instrument profiles for Roland Aerophones, Yamaha digital saxophones and Alto Venova, plus guitar, bass, piano, ukulele, recorder, flute and violin.
  • Easy Fingering for AE-01/05/10/20, YDS-120/150 and AE-BRISA in Brisa mode. It preserves intervals and shifts the backing with the melody; incompatible passages are rejected. Unsupported profiles warn and continue normally.
  • Song-region selection, stem separation, transcription, practice-speed control, metronome and video titles.

Validation and limitations

All five platform packaging jobs passed Rust checks, actual ML execution, relocated runtime setup, doctor --ml, and a validated demo render. Per-platform SHA-256 files, runtime manifests and smoke-test MP4s are attached.

Transcription and instrument fingerings still need musical/on-instrument review. Profiles remain verified: false; Easy Fingering limits controls, not rhythmic difficulty. This is the first public CLI release, not a GUI application.

Built from commit daee8f1. Source and dependency distribution details are in docs/DISTRIBUTION.md; runtime archives retain dependency inventories and license files.