Basic structural concept
The patcher reads in 50 (or so...) videos, and plays 10 second clips randomly based on a probability table generated by how often viewers are watching the specific clip. We achieve this using one of the cv.jit objects (the name escapes me) that detects heads. For videos that have been watched frequently, we apply a discretely controlled distorting effect (I might want to pt jit.sobel in there, I think I'm addicted.).
After the first week, maybe create a new video table slot in the patch that adds the tweaked out mirror from 10.17's lecture, thereby adding to the concept that we are tacitly complicit in our media.
Any Thoughts? I might need a little help steering in the right direction with content; my initial reaction is to make it be on the dark of side pop-culture and rather cynical, but i don't know if that's being a little too cliche.
Comments
thoughts:
1. the content is where the art is. that is where you need to put most of your effort.
2. edge detection algorithms like jit.sobel are hard to make happen gradually; consider mixing sobel'ed video with the normal video and slowly crossfade
3. we must make sure that your project for the Experimental Video course does not conflict with this one. i expect two separate projects for the independent study and for experimental video.
4. i suggest that you begin making a patch and that you begin preparing your content clips as soon as possible. you must work with the material and try different things, let it mature, and all that takes time...
5. look at the 'coll' object for storing a list of your videos
Posted by: Ali | October 18, 2007 2:04 PM