HapTone is the first smart tuner truly built for blind and visually impaired musicians. Halfway between an adapted musical instrument and a connected device, it combines an ultra-precise audio sensor, an embedded microcontroller and multi-modal feedback — vibration, voice, sound — to make tuning reliable, fast and autonomous, without depending on any visual display.
This article is a project update as of May 2026. HapTone was first introduced in June 2025 in our founding article “Tune AI: the inclusive adapted tuner”; since then, the project has entered hardware and software prototyping in collaboration with Polytech’ Paris-Saclay. Two students from the school, Manuelle Wafeu and Yannis Ngako, supervised by Amine Khettat (CEO of Blind Systems, who proposed the project to the school), defended in April 2026 a mini-project dedicated to HapTone. They presented the functional architecture, hardware integration and signal-processing pipeline detailed below.
The problem: a simple gesture made inaccessible
Every consumer tuner relies on a needle, LEDs, or a screen. For a visually impaired musician, this closes off something essential: the ability to practise their instrument independently. The existing workarounds — smartphone apps with a screen reader, tuning by ear, or asking a sighted person — are slow, imprecise, or simply exhausting on a daily basis.
Tuning becomes especially difficult in noisy environments — rehearsal rooms, stages, music classrooms. Standard microphones pick up everything, and the tuner needle becomes unreadable just when you need it most. That is exactly the barrier HapTone sets out to remove.
Our solution: multi-sensory tuning
HapTone replaces the needle with three simultaneous feedback channels:
- Haptic vibration — a vibration motor driven by a haptic driver, the DRV2605L from Texas Instruments, encodes pitch accuracy as distinct patterns (short pulse = too low, long pulse = too high, brief confirmed pulse = in tune).
- Voice synthesis — the system announces the note name and the deviation in cents, through a speaker.
- Audio tone — a continuous reference tone modulates its frequency depending on the deviation.
Sound acquisition: MEMS microphone or piezo sensor, your choice
- 24-bit MEMS I2S microphone — standard acoustic capture, ideal for quiet environments, wind instruments and voice.
- Piezoelectric sensor (SEN-VIB01) — direct contact with the instrument. Selected specifically for its robustness to ambient noise: it stays reliable on stage, in rehearsal rooms or in group classrooms.
Functional architecture
The system core is an ESP32-S3 microcontroller (from Espressif Systems), running FreeRTOS. End-to-end latency below 50 ms — essential so haptic feedback feels synchronous with the gesture.
Final product architecture: two paths under discussion
From an industrial standpoint, we are still weighing two final configurations for the tuner. This decision is not trivial: it will impact the retail price, the ergonomics and therefore the economic accessibility of the product. At Blind Systems we fight every day against the “expensive product for the disabled” fallacy — visual impairment must never be a reason to pay more.
Configuration A — Standalone tuner (built-in vibration motor and speaker)
- Pros: no smartphone needed, instant power-on, fully autonomous, accessible to musicians who don’t own a smartphone (or don’t want to use one while playing).
- Cons: bulkier housing, higher BOM cost (LRA motor + speaker + larger battery), more frequent recharging.
Configuration B — Minimal clip paired with a smartphone (Bluetooth)
- Pros: tiny housing (sensor only), much lower BOM cost, the smartphone provides vibration and speaker, the dedicated app can evolve independently of the firmware.
- Cons: smartphone required, Bluetooth pairing step, latency depends on the phone, the user depends on a third-party device to play.
Our current intuition leans toward a two-product line-up — HapTone Solo in configuration A for those who want full autonomy, and HapTone Mini in configuration B for the lowest possible price. Before we commit, we are opening this discussion to the community: your feedback will weigh heavily. What you would say at the moment of buying matters more than anything.
Demo: piezo-sensor pitch detection
This first demo, recorded by the Polytech team, shows pitch detection through the piezo sensor in direct contact with a string.
Demo: MEMS I2S-microphone pitch detection
This second demo illustrates pitch detection in acoustic mode through a MEMS microphone, useful for wind instruments and voice.
Detection algorithm: YIN, chosen for robustness
Rather than combining multiple approaches, the Polytech team picked a single algorithm chosen specifically for its robustness: the YIN algorithm, published by Alain de Cheveigné and Hideki Kawahara in their landmark 2002 paper (JASA). YIN is significantly more robust to missing harmonics and noisy signals than a classical FFT — exactly the use case of a musician playing in real conditions. Its principle: difference function, cumulative mean normalisation, YIN thresholding (typically 0.1) and parabolic interpolation. From the detected frequency, the firmware identifies the closest tempered note (88 piano notes, A4 = 440 Hz) and computes the deviation in cents.
Who is it for?
- Blind and visually impaired musicians.
- Music students with adapted curricula.
- Professionals in noisy environments (stage, studio, rehearsal room).
- Inclusive music education teachers and institutions.
Development status
HapTone is currently a working prototype, validating the acquisition stack (MEMS microphone and piezo sensor), the YIN detection pipeline and the haptic feedback. Next steps: pick configuration A or B (or both), PCB industrialisation, enclosure, companion mobile app, CE certification, and broader user testing.
Open source: building it together
HapTone is a fully open-source project. Schematics, firmware, 3D enclosure models and technical documentation are published on GitHub.
⭐ Explore the project on GitHub
Join us: contributors, volunteers and patrons
- Embedded developers (ESP-IDF, FreeRTOS, DSP)
- Hardware designers (PCB, enclosure, mechatronics)
- Mobile developers (iOS, Android, accessibility)
- Musician testers
- Patrons and institutional partners — HapTone is currently personally funded by Amine Khettat (CEO of Blind Systems), pending support from patrons or institutional grants. Your contribution will help accelerate industrialisation and deployment to schools and conservatories.
- Communication, translation and documentation volunteers
Four ways to reach us — pick the one you prefer:
💬 GitHub Discussions 🛠️ Open an issue ✉️ contact@blindsystems.org 📝 Contact form
Tell us about your background, your availability and how you would like to contribute — we will get back to you quickly.
HapTone is a project by Blind Systems, an organisation dedicated to digital accessibility for blind and visually impaired people, run in collaboration with Polytech’ Paris-Saclay. Inclusion in action, innovation through passion.
