Algorithmic Photography

Creating photographs from video using digital algorithms.

Continuing his core interests of light, code, and time, Alex is creating a new body of photographic prints and video works using source material filmed around the world on his many travels.

The works are made by taking a video (anything from three seconds to three hours) and algorithmically combining the data contained in each frame down to a single resulting image.  The source video is filmed at 25 frames per second, so a three hour video contains 270,000 images.

The algorithms are analogous to how film works in chemical photography, where it responds to different qualities from the incoming images.

Alex watches as the image ‘develops’, which currently takes twice as long as the length of the source video, and decides when to extract the final image.

The images are then cropped, and occasionally straightened, but receive no further processing or have any kind of filters applied to them.

Copenhagen air and ground traffic (detail)

Some images contain pixelated artifacts as a result of the compression applied to the video by the camera.  These are an important part of the image that reveal the qualities of the technology being used.

Alex has been mainly using a GoPro camera to record the source video footage as it is small, rugged, and perfect for travelling with.  One of the interesting aspects of the camera is the lack of viewfinder meaning the video cannot be lined up or previewed in advance.  Much akin to using a basic pinhole camera, Alex has to guess how the image will look when setting up.

The algorithms are written in GLSL shader code, running under Fugio, Alex’s open source digital art programming system.

See the latest images on Alex’s Instagram account.

Enquiries for ordering of prints, exhibitions, or commissions, should use the contact form.

Leave a Reply