Experiments in Extreme Format Shifting and Physical Object Creation

What I want to explore is how I can adapt this to other media? 3d printing? How would a heightmap in Rhino affect this?

Now, I have all these and the process for creating them - .mp3 to .wav to .png to edited .png back to .wav back to .mp3

Can I convert the .wav to a spectrogram that can be printed then re-imported back to .wav to .mp3?

What can the introduction of Perlin (or other!) noise to the .png do to the .wav audio?

What if I radically compress the .png - what does that sound like?

What about other extreme format shifts back to audio?

Combination of an image and an audio file?

What about a pen plotter?

How to make 3d?

Or flip it: image -> wav, edit the wav, -> back to image?

Maybe: painting -> .wav -> edited .wav -> new painting?

Can I create a cycle of iteration through integration of a 3rd transformation?

How would a print to a transparency re-importaed back to .wav affect the audio?

How far can I push the .png and still retain recognizeable audio - and is that important?

What about hte reverse - how far can I push a .wav and still retain recognizeable image - and is that important?

Could this (or a shorter sample) be converted to gcode that could create a 3d print that could be photographed then converted back to .wav to .mp3 from a 3d print?

Sort of riffing off this but in my own radically different direction of extreme format shifting: https://mixmag.net/feature/spectrogram-art-music-aphex-twin

(unlisted videos)

Three early and basic tests using public domain Beethoven

Moonlight Soanata 20x blur:

https://www.youtube.com/watch?v=5rjNtUdboI0