distro-planet

Harmonicon adds harmonica bending trainer with live pitch feedback

Key Takeaway

The open-source music practice tool Harmonicon has gained a dedicated bending trainer that visualizes pitch in real time to help harmonica players learn note-bending technique.

Summary

Tomaz Canabrava has detailed a new bending trainer screen in Harmonicon, an open-source harmonica practice application built with Rust and the Bevy game engine. The trainer addresses the core difficulty of learning harmonica bends — the technique is invisible and beginners cannot judge by ear whether they are doing it correctly. The tool provides real-time pitch visualization, structured practice drills, and progress tracking across all harmonica techniques including overblows and overdraws.

How the bending trainer works

  • Pitch detection via FFT: The trainer uses an FFT-based pitch detector, with the current example set to "C harp · FFT"
  • Three-part UI: Settings strip at top, target card in the middle showing current bend target and performance, harmonica diagram on the right displaying all possible notes by hole and technique
  • Real-time pitch rail: Visualizes the player's pitch relative to target with a moving marker and dot trail showing recent history, revealing technique issues like scooping, overshooting, or unstable holds
  • Six practice modes: Free exploration, find and hold, bend and release, repeated bends with metronome, bend ladder through multiple depths, and overbend response drills
  • Adaptive drill system: When Drill is enabled, the trainer automatically selects targets based on practice history, prioritizing less-practiced or poorly-controlled bends
  • Progress persistence: Session data and drill progress are saved between sessions; the harmonica diagram doubles as a progress map with color-coded reliability indicators

Why it matters

The tool demonstrates a practical application of Rust and Bevy for audio-focused creative software, and fills a specific gap in open-source music education tools where instrument-specific learning aids remain rare.

Original reference: tcanabrava.github.io →