Skip to main navigationSkip to main content
The University of Southampton
Winchester School of Art

Research project: The Image Machine

Currently Active: 
Yes

What do we mean by images of value, and how might we discover them systematically?

There is a huge but finite number of still images it is possible to present on a conventional TV: the number is the number of colours you choose to use – 256 is small but sufficient for general representational purposes – to the power of 414720. So, 256414720.

The Image Machine project has two phases: the development of a computer programme which will generate any and all of the images in the series from 1 to 256414720, in sequence or at random, as required, and the use of that programme to identify images of value.

The development phase takes data from artists and designers, and existing imagery held to be of value, and identifies key features of useful images, like horizons, facial characteristics (as recognised by quite basic modern cameras) in order to calculate the parts of the series more or less likely to yield ‘useful’ imagery, the more or less promising parts of the series for different image types. The programme incorporates a method of eliminating most of the near-repetitions which the series includes.

In the implementation phase, the programme presents the sequence as a nearly endless movie, in which each frame, say, is one image from the automatically edited series. The hardware is set up so that the movie only plays when a research team member is present to watch. The researcher’s role is to archive interesting frames, according to agreed criteria. That person has discretion to cause the series to jump to any new starting point they chose – by scanning a famous painting into the system, for example – but the system does not revisit images it has already shown.

Saved images will be available for researchers to appropriate and work with. The movie is shown continuously on a large public screen on the outside of the facility. If no researcher is present to monitor the programme, the screen displays the most recent frame the programme has produced.

As a way of discovering new images, the system is no more and no less systematic than having buildings full of artists simply experimenting with material, some with a plan, others with none.
As a control for The Image Machine experiment, artists work in an adjacent space, on a rotational basis, making and archiving digital images of the same pixel format, in Photoshop, say, or Illustrator, according to their own method and rationale, for inclusion in a parallel movie. Like the movie from The Image Machine, output from the control experiment is presented continuously on a large external screen. Unlike The Image Machine, outputs from the Control Experiment are not made available for appropriation, as they have individual authors and are subject to copyright.

Share this research project Share this on Facebook Share this on Twitter Share this on Weibo

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×