Converts audio to LED lights. Chooses a color based on the frequency of the audio, chooses the brightness based on the loudness.


A friend of ours told us he could not find an app like this anywhere and we all agreed that it would be fun to try to tackle.

What it does:

Provides more customization, immersion, and general impact on your listening experience than typical lights that pulse to the beat of the music. It takes more than the volume of the sounds into consideration, the frequencies can give a lot of information.

How we built it:

We split the work up into several sections. They involved analyzing and filtering the audio data, making the user interface, making the lights display properly given a signal, and making sure that we can piece those parts together.

Challenges we ran into:

Numerous software issues and hardware limitations. Instead of ideally having an LED strip to display the lights, we had only a handful of single-colored lights on a bread-board. One of our laptops started running very slow early on in the event and it caused us to miss out on several hours just waiting for it to install software and compile throughout the event. We also had issues setting up the library we were using to analyze the data as well as issues installing Raspbian on the Raspberry Pi.

Accomplishments that we're proud of:

Learning the math behind audio analysis and being able to filter the data.

What we learned:

We learned that even if we know a language or IDE, it does not necessarily mean it will make our lives easier. The software and hardware is dependent on the task, not the other way around.

What's next:

Use metadata to provide more accurate lighting for the mood, or genre. Also, allow user to store their preferences, like different color themes or palettes, brightness, etc. And the ability to transmit audio files wirelessly via Bluetooth or WiFi.

Built with:

Raspberry Pi, Wi-Fi Nano USB Adapter, Bread Board, Python, Qt, Raspbian, Essentia

Prizes we're going for:

Misfit Shine 2

HAVIT RGB Mechanical Keyboard

Team Members

Quentin Bethune, William Sullivan, Kyle Toohey, Ethan Johnson
View on Github