DIY Audio Spectrum Analysys - MonkeyListen!

[ehughes]

MonkeyListen – A DIY audio spectrum analyzer! (Designed with Altium)

This project uses the Kinetis K20D50 (on the FRDM-K20), a FRDM-OLED shield (128 x 64 pixel OLED) and an electret microphone to create a cool spectrum analyzer.

The firmware has 3 modes: Time Domain, FFT, and spectrogram (for a really cool real time time-frequency plot).   There is also the ability to compare and contrast a rectangular vs a hanning window.  Good way to start playing with FFTs.

Lots of instructional videos (4 hours!) included and all software/open is open for hacking.

Quick demo videos





Main project page:

https://community.freescale.com/docs/DOC-100149

No actual monkeys were harmed in the production of the project.

[lpc32]

So the primary function is sound processing rather than vis?

[ehughes]

A mix of both.  It is an education project to get newcomers some experience in FFTs in an embedded system (which is usually a difficult topic, even when  you have a nice library).   The visualization aspect adds some interesting I/O to "see" your voice to get a feeling of how things work.

From there someone could hack some other function in that does something cool with the frequency domain data.


I am starting a new project is an LPC4370 that will effectively do the same but do some interesting processing on ultra-sonic data.   That will get posted at some point as well.

[lpc32]

What could be interesting is some sort of open high-performance DSP platform, maybe hardware upgradable to a certain degree (optional multi-chip config?), and with a well defined software API for extra modules and functionality.

[lpc32]

What could be interesting is some sort of open high-performance audio DSP platform, maybe hardware upgradable to a certain degree (optional multi-chip config? connectable control devices?), and with a well defined software API for extra modules and functionality.