topherdigital writes:
This is the second test and example of an audio visualization done in Blender. The video is a result of a python add-on for blender in development. Several fourier transforms were done automatically (104) on a logarithmic frequency scale. The speakers are shape-key driven with a driver tied to a broad base FFT. The mesh is an automation process that creates the mesh from data in the full range FFT. No additional modules or add-ons were used, only stock blender 2.61.
The audio track is "Réalité" from the album "L'autre endroit" by the artist "Silence" and is released under the Free Art Licence (FAL 1.3) and is available from Jamendo.
The video was rendered using OpenGL on an Nvidia 8500 GT instead of the traditional raytracers. Custom vertex and fragment shaders were used along with one point lamp. The scene totals 743537 faces (quads).
3 Comments
this combined with camera tracking could be really awesome... I'd really like to see a true tutorial for this, since it sounds very complicated...
A timelapse perhaps would do :)
Here is a good tutorial to start audio visualization in blender
http://cgcookie.com/blender/2011/06/02/building-an-audio-visualizer-in-python/
Would be cool to get a GLSL live bake from a VSE audio strip. Would make editing audio for level and frequency much cooler!