Empty Frames

Let me pick up on something from a previous post:

"The digital video encoder has never been happier than when given the job of encoding films by Chantal Akerman or James Benning. These are computer-friendly substrates"

The punch line being that the "ideal" computer image would be an empty frame. And the most odious image, according to the essence of computation, would be white noise, an image of pure entropy. I'm not sure if Gregory Chaitin has ever visualized his omega number, but that's what it might look like. The omega image.

In fact Jason Lariviere at NYU has been plotting the bitrates of different films, showing the contrast between the kinetic style of someone like Michael Bay and the slow films of Andrei Tarkovsky or Béla Tarr. For you DH fans, this provides a primitive way to measure aesthetic complexity in digital cinema, and likewise to show images exhibiting greater or lesser compatibility with the compression codec. Suffice it to say that Michael Bay aspires -- asymptotically, impossibly -- toward the omega image.

Of course this leads to a series of risks and potential pitfalls -- have the films been encoded in similar ways? do we have an effective baseline against which to measure? what's our control group? And certainly a film can be encoded at any bitrate desired, be it low or high, meager or plentiful, cold or hot. So it's important to be cautious about jumping to conclusions. But at the very least we have a hypothetical measure of aesthetic complexity, or "cost."

(For maximum perverse entertainment one could subject certain films to "restrictive" bitrates, like encoding Revenge of the Fallen at the bitrate of The Sacrifice just to underscore how debilitating it would be for Bay to degrow into Tarkovsky.)

Along similar lines, I've been thinking a lot about "cost" in media, not monetary cost of course but the cost of aesthetic complexity. (Cost isn't the right word, but I can't think of a better one.) So for instance, highly detailed textures are "free" for the camera obscura and its derivative forms (analog photography, or even, in reverse, the magic lantern), but they are "expensive" if you try to render them with a computer. Photography essentially starts at the aesthetic level of the world, but computers start at zero. And everything above zero comes at a price measurable in electricity usage, time, cpu power, resolution, etc.

Simply apply brute economic realism to the computer and the "ideal" image of computation quickly shows itself. The ideal computer image is an empty frame. Ideal computation is emptiness.

Contrast this with other episodes in art history, where empty frames are in fact quite difficult to achieve, and are often only the consequence of tedious illusionistic techniques. Try taking an empty photograph, for instance; it's practically impossible. As artist Liz Deschenes has shown, the more you try to empty the photographic image -- removing the camera, removing the object -- the more lush and vivid it becomes. There's always some grain or schmutz or something. And the same is true for audio recording. Silence is a concept; it's not a reality, at least not in analog representation.

In other words, analog representation has a very hard time being empty... because it's already predefined as the thing that isn't empty. Whereas digital representation always aspires toward an empty frame... because the digital means framing and nothing else. (This is one way in which digitality and phenomenology are incompatible. Digitality is pure "conditions of possibility" but without thereness. Phenomenology is an inquiry into conditions of possibility but always from within the real experience of thereness.)

Now the lovely irony...recall how entropy is central to Claude Shannon's definition of information. Information is defined directly in terms of entropy; more entropy means more information. So the omega image above is an amazing contradiction. From the perspective of information theory it's maximal information. But from the perspective of computation -- and according to Chaitin's definition -- the image is literally uncomputable. So the most uncomputable is also the most informatic.

So if you want to know the essence of film, luxuriate in the visual richness of Josef von Sternberg or Max Ophüls. But if you want to know the essence of the computer, just watch the last section of Antonioni's Eclipse.