[f_devd]

For those wanting a textual description: It seems to take a FFT of the audio and put it into rings (something like frequency = ring size), and those rings are then duplicated/projected to make spheres, the rings (above a certain size) are truncated to a cube shape. There is an option to have the rings move over the length of the sphere/cube so you can see the FFT over time.

[redox99]

More like rings, it's a spherical coordinate system where radial distance is frequency, and angles don't matter. Then the brightness of the color is based on the amplitude of each frequency. Finally, the function is truncated into a cube.