-
-
3D CAD Design of FretNot
Inspiration ✍️
One of our members is an avid guitar player, and the the rest of the group either tried to or have been interested in learning how to play.
What it does 🎸
FretNot is a clip on attachment to your guitar that projects lasers onto the fretboard to help you learn the finger positionings of chords. It is controlled via a web application that allows the parsing of song lyrics and chords, and can cycle through the song lyrics with their respective chords automatically. The idea is that it's like the training wheel for guitars, where you get the muscle memory in using it which eventually lets you play the chords without it.
How we built it 🛠️
FretNot was built using an ESP32 alongside a Bluetooth connection to a web browser front end. The web server utilizes TypeScript to parse the images into the proper format and preview them automatically. The enclosure was custom built in OnShape and printed out using an Elegoo 3d printer.
Challenges we ran into 💢
One challenge we ran into was the limited amount of output pins on the ESP32. There are lots of frets on the guitar and only so many outputs for the laser diodes, and while we thought about linking several ESP32s together and having them communicate via Bluetooth, we only had brought one board along. To compromise, we ended up choosing specific chords to display with a specific song to preview. Doing this would allow us to show the functionality of FretNot and how we could scale it up in the future. Additionally, we had to wrestle with the enclosure not being properly fitted together when printed, which was especially difficult as each print took a substantial amount of time in this limited time hackathon. We ended up allowing more tolerance in the measurement, which ended up working better as it was easier to fill the space rather than have too much of it.
Accomplishments that we're proud of 🏅
This is a first time hackathon for three of our members, so being able to have a finished product in 24 hours is astonishing for us. Additionally, we were really proud that we were able to take a shared interest of ours and put that into developing something that can really help people.
What we learned 📝
For the software side, we learned how Low Energy Bluetooth is transmitted through the ESP32 and how that communication is formatted with the connected device. Additionally, we learned how and OCR algorithm works in TypeScript, and the tradeoffs of using that over something like machine learning. On the hardware side, we learned how to model real life objects such that attachments can be made around the real life parameters of those objects. We also had a fun time learning more OnShape.
What's next for FretNot 🔮
Currently FretNot only supports 4 guitar chords, C, E, F, and Fm. Scaling up, we would want support for all chords of the guitar, which would involve a lot more lasers. Additionally, we would want to add more support for the types of files that can be inputted for song practicing. Currently, it uses an OCR to parse the lines of text, thus it requires a specific format of song and chords. For the sake of robustness, we could potentially switch to machine learning for parsing for a wide variety of song formats, such as tabs and sheet music.
Log in or sign up for Devpost to join the conversation.