RAUSCHEN
enter RAUSCHEN github page for more details
photography: Mika Stoerkel | Omar Zaki
Every digital image is a limited, but unimaginably large space of possibilities. All these
possibilities together are its probability space. In a regular 24-bit image file, each pixel
is defined by 256 shades each of red, green and blue, which comes out to 16,777,216 possible
colours. For a one megapixel image (1,000 pixels wide and 1,000 pixels tall) there are
16,777,2161,000,000 permutations. That number far exceeds how many atoms there
(probably) are in the observable universe (1080). Actually it is 7.87x107,224,639 times larger. Even if every atom in the universe was another universe full of atoms that are universes
full of atoms and so on, we would have to go 90,309 levels deep to finally have as many atoms as
there are permutations of a 24-bit megapixel image.
Most of it is just meaningless noise. But how do we find the infinitely small rest of it? You know,
the rest that contains (among other things) all photos, drawings or graphics ever made, will ever
be made or could ever be made? Usually, we go at it from the other direction: We take a picture.
We make a drawing. We create graphics. Yet we are always limited by the fact that meaning needs
to exist first, either in the world or in our minds, to be found within the noise.
Or we could just leave it up to chance.
RAUSCHEN ("noise") is a real-time emergent media system exploring the probability space of a 1,000×1,000
image. It generates textures through a variety of techniques such as white or Perlin noise to recursively
feed into a modular palette of per-pixel algorithms. They each represent effects that, running
in quick succession, synthesise exponentially random patterns. While contemporary image generation
models use training data to impose meaning onto the noise, RAUSCHEN aims to chart new areas hidden
within it. In the end, a human decides what is worth keeping.
How can we find meaning within all this noise?
