Clever concept, but watching that video really hurts my eyes. If he had slowed it down a tad it may be easier to register.

You need to sit REALLY far from your monitor to appreciate this!

this brings to mind the ASCII videos that are made entirely with text & symbols.

http://www.youtube.com/watch?v=RENthHHKiGA&feature=related

This is a fun exercise, but there’s nothing special or revolutionary about it.

Way back in the nineties, I had a few programs that turned photos into ACII graphics. A few years ago, I downloaded a shareware app (I forget the name) that did the same, only it allowed the user to supply a designated folder of images that would be assigned as the “cells” of the resulting image. The program reduced the resolution of the user’s source image to something low, such as 160 X 120, replaced each pixel with one of the images in the folder most closely matched to the visual characteristics of the target pixel, and output a very large image that appeared, from a distance, to be a pixellated version of the original. Upon closer examination, the individual pictures serving as pixels could be discerned. A famous online use of this technique was a portrait of John Ashcroft, which, when zoomed in, was revealed to be comprised of pornographic imagery.

The use of the described method, applied to a sequence of frames (in this case derived from random video clips), was apparently the way (>’.’)>=O____l_*__O= was created.