Abstract 

Keywords: Motion Prediction, MPEG, Digital Video, Video Compression, Statistics, Ballistics.

Prediction theory emerged during the WWII in order to improve anti-aircraft artillery and resulted in algorithms devised to statistically predict airplanes and missile paths. Although today predic- tion is the backbone of the video compression, the historical and technical connection between this mathematical theory and contemporary imaging technologies has not been sufficiently determined. Using a media archaeological approach this paper discusses how the implementation during the 1990s of prediction algorithms to video compression has generated an entirely new type of moving images. I argue that the consequence of turning each displayed picture into a rigid grid and its construction into the statistical prediction of the pixel’s values is dramatic because it renders the temporal coincidence of all pixels within the frame unnecessary. On the surface there is no change. Yet, using prediction a video codec such as H.264/AVC has turned the frame into an address where chunks of pixels coming from different moments in time are put together. At the coding level, prediction has banished the frame. The elimination of that basic unit of all moving images, not only miniaturized video but it also has had ontological consequences for the image that are not yet fully understood.

• Ricardo Cedeño Montaña (Bogotá, Colombia, 1976). His background spans the fields of media history, media art, and design. His current research interest revolves around technical media and the history of knowledge with a particular focus on imaging techniques. He holds a PhD in Cultural History and Theory from the Humboldt-Universität zu Berlin, Germany (2016), an MSc in Digital Media from the Hochschule Bremerhaven, Germany (2009), a degree in Multimedia Creation from the Universidad de los Andes, Colombia (2003), and a degree in Industrial Designer from the Universidad Nacional de Colombia (1999).

Full text (PDF) p. 415-420