Tree of Sound is a hobby project that started with the following question: Most applications for creating audio synthesizers or audio effects use the unit generator and patch cord metaphor, e.g. Max/MSP, Pd (Pure Data), and audio signal flow is represented as a directed graph. One of the downsides of this form of representation is that the resulting patches are often not logical and maintainable, because the form of representation obscures the direction of signal flow. The thought occurred to me what if one represented signal flow as a tree instead of as a directed graph, because that should make the direction of the signal flow always clear?

In the iteration of Tree of Sound show above, I experimented with using a 3D representation of audio signal flow that could be described as a Russian doll approach. One sees the modules that comprise the audio signal path as a tree. But parameters to those modules can also be modified by 3D trees, and switching to parameter mode reveals the nested trees (branches?) modifying the parameters of the main tree.