The entirety of the build is minimal, consisting of a GY-MAX4466 electret microphone module, a KY-040 encoder for some user control and an ESP32 attached to a Neopixel strip. The only additional electronic parts are some passive resistors to limit current on the data lines and a capacitor for power line noise suppression. [Stephen] uses various enclosures from Thingiverse for the microphone, rotary encoder and ESP32 box to make sure all the modules are protected and accessible.
The magic, of course, is in the software, with the CircuitPythyon ulab library used to do the heavy lifting of creating the spectrogram and frequency filtering. [Stephen] has made the code is available on GitHub for those wanting to take a closer look.
It wasn’t very long ago that sound reactive LEDs used to be a heavy lift, requiring optimized FFT libraries or specialized components to do the spectrogram. With faster and cheaper microcontroller boards, we’re seeing many great projects, like the sensory bridge or Raspberry Pi driven LED spectrogram, that can now take spectrograms and Fourier transform calculations as basic infrastructure to build on top of them. We’re happy to see [Stephen] leverage the ESP32’s speed and various circuit Python libraries to create a very cool LED car hack.
Video after the break!
Source link
