Hi Quinn! Coincidentally, I've been working on a sound visualizer which accurately translates sound to visuals, while preserving all the information and, what's way more important, it transfers the "sound symmetry" which we often refer to as "musical harmony". The visuals can be translated back to sound.
As someone with nearly perfect hearing and musical education background, I find the visuals oddly accurate when I also listen to the corresponding music. It can trivially distinguish human vowels, as those produce different recognizable shapes.
I've put together a few sample images and a live demo: check out "github soundshader" (the project that talks about ACFs). I'm really curious what you think about this.