A rotoscoping experiment created in order to test the hypothesis – does an abstracted excerpt of motion still read as the original action? I rotoscoped a scene conveying intense emotion, from a very well known film. I then looped the action and repeated the images to create a kaleidoscope. I wanted to see if, even completely unrecognisable, the images would hint at the underlying action. I then removed the complexity over time – the final loop is just the original sequence.