Using openFrameworks and ofxFFT to generate 3D Meshes of sound data for use in 3D printing.
Download the app for OSX 10.9/10.10 – SonicSculpture_v1.0.0.zip (2962 downloads)
Download for Windows (not super tested, but should work – keep it in its folder! – Windows version provided by Oleg Pashkovsky – thanks Oleg!) – SonicSculpture_Win.zip (1961 downloads)
Released with this license:
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (terms)
The app above is available for free – under a it is by no means a finished product, but it should be usable for those willing to tinker – use at your own risk and expect to tweak settings! The app comes with settings that work for most songs, default maxes at 3000hz instead of 20khz – good enough for most music ranges. The app will also max out at about 5-6min of music capture depending on your computer – so don’t plan to sample an entire album just yet, unless you lower the sample rate. The final model also takes a bit of work to prepare for 3D printing. Read the included Readme for full instructions!
I am not registered with Apple as a developer ($99/year), so Gatekeeper may trigger and refuse to open the app because it is not code signed – feel free to build from the source if you would prefer. You can temporarily disable Gatekeeper in System Prefs – Security and Privacy.
If you do use it – please let me know if you have any success with it – would love to hear or see what you make. Please check my license and credit me where appropriate. I will try to make periodic updates.
I started this in May 2013 as an experiment in visualizing real-time music data as more of a physical structure than an ephemeral animation. It allows me to look at the structure of a song in a different way than a musical score or just a graph of the intensity of sounds and frequencies. I get a very different sense of the dynamic range and structure of certain songs when viewing them in this way. The app itself performs FFT audio analysis on incoming audio and spits out the intensity of different frequency ranges, allowing you to differentiate bass drums from guitar squeals or cymbal crashes. Once I have the model, I can then explore making physical objects out of songs in a more tactile way – different than burning a CD, pressing a vinyl or etching a wax cylinder where the shape of the sound is still imperceptible to your fingers. That fact that it runs in real time is also an intriguing way to look at live audio performances and things like that.
The shapes are set up like this: left is generally the bass/low end, right is high end. Red or orange marks are the loudest frequency band at that particular time. White to black shows the relative volume intensity of the particular time. You can adjust the number of frequency bands it is looking at and make it more coarse or fine.
Next steps for me will be to partner with a musician who is interested in getting some custom prints made of their songs. The models have a lot of detail to them, so printing them at a large scale at the size of a vinyl record or larger would be ideal, but that would take a lot of time to print or CNC. I have only printed ones that are the size of a coffee cup so far.
If you are a musician is who interested in getting some prints made of your songs – please get in touch. I would love to make some
You can see some images of sample prints below, they are about the size of a bangle or bracelet.
I have been doing this as a personal project and a side project where I work: www.fakelove.tv
Sample prints available on Thingiverse
This is also by no means a new idea – there are several pieces that explore the same principle that came before and after my implementation.
Examples:
Frequency Domain (game)
Robert Hodgin – Audio Generated Landscape (2008)
Lukasz Karluk – Holodecks – 3D printed audio + augmented reality app (2013)
Reify 3D printed sound sculptures (2015)