nonsite

Claude Shannon’s important work on information theory in the 1940s gives us the theoretical bounds for rate–distortion theory and how much compression can be achieved through redundancy using lossy compression methods.

I’m doing some experiments to see what king of images I get when I shoot at a low resolution and then compress using jpeg compression and then enlarge. I’m curious how much interpolation the output devices need to make in order to create a large scale version of a low resolution image.

I’m reminded of seeing a Sigmar Polke retrospective I saw many years ago in the early 1990s at the old San Francisco Museum of Modern Art, where in the large gallery Polke had made two enormous vinyl banners on each of the two main walls. From a great distance the images held together, but when you approached the work up close you could see the stochastic dot patterns of the inkjet printing.