Projects

​​​​​​​​​​​​​​

Tonemap.live — Building a New Type of Practice Aid from the Ground Up

 

​

 

 

 

 

 

Tonemap.live began as an experiment: could a tuner provide more than a slice-of-time analysis, and help a musician identify intonation trends in their playing over time?  

​

The first ChatGPT sketches were on my laptop: a Python proof-of-concept using aubio for pitch detection and matplotlib to visualize frequency drift over time.  As an app on my laptop, it worked, but I realized that it wasn't very portable, and between App Store approval, downloading, installing, and maintaining, would have many obstacles to wider adoption.

​​​

So I started researching, and discovered it may be possible to implement in a browser.  I used ChatGPT to work directly with the Web Audio API and JavaScript.  I built an early HTML version using Aubio.js, which could analyze live microphone input.  From there I moved to an in-file McLeod Pitch Method pitch analysis, and a universal web version started to come to life.

​

As the prototype evolved, I moved everything into Visual Studio Code with Codex, integrated it with GitHub, and connected it all to Cloudflare for hosting.  Revisions and features are now much easier and faster.

​

ChatGPT also helped design the Tonemap.live logo: the 4×4 grid of colored squares representing the Tonemap grid.

​

The result is a single-page app that opens instantly, runs entirely client-side, and is private to your device.

​​

What's next?  I'm hoping to add different note names for wind instruments, persistent user settings, and more.  Your ideas, perhaps? Let me know!

​

-Jeremy