This is taken from a random image on TV. I don't know the performer.


DiceTV 'dices' the screen up into many small squares, each defaulting to a size of 16 pixels by 16 pixels.. Each square is rotated randomly in one of four directions: up (no change), down (180 degrees, or upside down), right (90 degrees clockwise), or left (90 degrees counterclockwise). The direction of each square normally remains consistent between each frame.


The key controls are simple:

Key Result
<Space> All squares are randomly assigned new orientations.
C The square size is halved in both dimensions. This results in a finer grained image. The size will not go below 1 pixel by 1 pixel, for obvious reasons. (At that size, no effect is visible.)
V The square size is doubled in both directions. This results in a coarser grained image. The size will not increase beyond 32 pixels by 32 pixels.

How to enjoy

DiceTV all but obliterates the optic flow of a video image. Only the largest and simplest motions survive. To maintain visual continuity, choose video streams in which the subject stands out prominently from the background, and both foreground and background are relatively simple in complexity. Small squares are less disruptive than large ones.


Sam Mertens

Copyright (C) 2001-2006 EffecTV crew.