It is minus forty degrees and two o’clock in the morning. Two of us are out here, pacing up and down to try and stay warm, on a small area cleared of snow in the centre of a frozen lake in Lapland. Twenty miles away, on another frozen lake, the second filming team is doing exactly the same thing, with a duplicate set of specialized low light, low temperature camera gear. All our equipment is focussed on the same point in the heavens, carefully aligned against the point map of the night sky. As the hours slowly pass we try to memorise the names of the stars that we use for our alignments: Dubhe, Merak, Phecda, Megrez, Alioth, Mizar Alkaid and, of course, Polaris. This is a waiting game.

Night after night we are out on that thick ice, listening to it boom and crack under our feet as it expands and contracts its way through the bitterly cold winter. Our ears are very aware of the sounds but our eyes are always looking upward, as my aching neck testifies.

We are here, under this enormous black sky, to attempt to become the first people to successfully film the Aurora Borealis stereoscopically – in 3D*. This is by no means an easy task. It requires the fortuitous alignment of several rather unpredictable variables. First and foremost we need the weather to be clear, allowing a view of the open sky. Secondly we need the sun to be creating sufficiently large bursts of solar wind – charged particles that blast the earth’s magnetic field – to generate auroral activity. Thirdly we need our complicated, multi-camera filming equipment to function correctly, even at these extremely low temperatures when plastic becomes precariously brittle and vehicles have a tendency to just stop working. Finally we require the humans working on the project to also function correctly. There is no doubt a graph somewhere that describes the relationship between human error and temperature. The line must get pretty steep as the thermometer edges below minus 30. Restrictive clothing, continually cold feet, and the need to remove gloves to operate tiny buttons, night after night, leads to understandable lapses in concentration and the occasional thinning of patience.

The Aurora Borealis had never been seen in three-dimensions before. The phenomenon, also known as the Northern Lights due to the northerly latitudes they are mainly associated with, hovers anywhere between one hundred and five hundred miles above the surface of the Earth. As the human eyes are only separated by a relatively insignificant six centimetres we are unable to derive any stereoscopic spatial information from the spectacular curtains of light waving and swirling above us.

That is why our second team was out on the other frozen lake twenty miles away. What we were doing was replacing the human eyes with cameras and then virtually expanding the human head so that the ‘eyes’ were now twenty miles apart. When presenting the final film, the footage from one camera is fed into one eye of the audience and the footage from the other camera into the other eye. The viewer’s brain then reforms a stereoscopic – 3D - image of the Aurora which appears shrunken to a size that allows for clear comprehension of its shape, structure and motion as it floats before the projection screen. It feels as if it we could reach out and touch its delicate curtain-like bands of pulsating light.

The Aurora project was a collaboration between myself and George Millward. George is an atmospheric physicist and musician. We have worked together on several video and installation projects over the years, both in the UK and Australia, but always with George wearing his musician’s hat as apposed to his scientist’s hat. Some of these past projects have been 3D – stereoscopic - films. In 2001 we were selected to show what was probably the only 3D video to have been screened at the Sundance Film Festival in Utah. These earlier films were complex, multi-layered affairs with a definite ‘everything-including-the-kitchen-sink’ aesthetic. Often with an underlying scientific theme.

In 1999 George showed me some filmed footage of the Aurora that had been kicking around the research lab at UCL. From its snaking and curling movements it was obvious that it was a three-dimensional phenomenon and that it should be possible to film it as such. We also realised that by filming the Aurora stereoscopically we would be recording the largest thing it is possible to film in 3D from the Earth. In 2001 funding was secured for the project from the National Endowment for Science, Technology and the Arts – NESTA.  The next three years were spent in the pursuit of this one goal. This meant working with George in his scientific context for the first time. Complexity of image content was replaced by complexity of logistics and pairs of heavily insulated boots.

My background is photography. I was, however, never satisfied with the image-making capabilities of off-the-shelf camera technology. I was, and still am, very inspired by image makers who are also technical innovators. Photography’s early years are full of such characters and an enjoyment of this history encouraged a rather Heath Robinson approach to picture creation. Vans were converted into giant cameras, kites were used to haul large cameras into the air, 3D pinhole cameras were created, strange colour processes were ‘invented’, etc.

By the mid 1990’s I had discovered computers and digital imaging which quickly led into digital video. I realised very quickly that the accuracy and freedom that non-linear video editing allowed would make it possible to simultaneously work with footage from two cameras forming a stereoscopic pair. The stereoscopic principles of the 19th century, first put to use on daguerreotypes in the 1840’s, could easily be transferred into the digital arena. On making my first migraine-inducing forays into the world of 3D video, I looked around to see a blossoming ‘golden age’ of stereoscopy immerging. Developments in computer gaming, CGI, Virtual Reality and screen hardware technology was all converging into a vibrant and robust new medium.

The growth of television in the 1950’s and then the development of video players in the 1980’s had a devastating effect on cinema ticket sales. In reaction to this the film industry, in each of these decades, made films in 3D that could not be experienced using the new competing systems. This latest burgeoning 3D scene, in contrast, was a technological and visual evolution rather than a backlash against a commercial problem. It was developing across many spheres simultaneously, within the arts, sciences and entertainment. Within the illusional space of the stereoscopic films I was now making I could interweave 3D medical imaging data sets, stereoscopic video footage, historic and contemporary stereo photography, animated models from CGI and hovering graphics layers. The resultant films could now be viewed on home computers, be projected to large audiences, be viewed through VR headsets or watched glasses-free on the new autostereo monitors being developed.

The Aurora project was in some ways a step away from the complexity of options that this new medium offered. One single idea. One very large idea. One that brought with it complications of production rather than meaning. Four people in the arctic. Digging cars out of snow banks following the swerve to miss the reindeer herd. Weeks and weeks of cloud cover. Boiled meats at eleven o’clock in the morning.

And the Aurora does not like to be filmed. She certainly does not like to be filmed in 3D and only reluctantly revealed her shape when she was confident we had earned the privilege of seeing this wonder of the natural world in an even more magnificent way.  When the entire sky lit up with the wildly dancing auroral flames, the cold and discomfort were forgotten as we fell back into the snow and stared in true amazement.




* Hans Nielson and Tom Hallinan from the Geophysical Institute at the University of Alaska made stereoscopic observations of the Aurora in the 1970’s. Due to the technical complications they found the experiments did not create the dramatic 3D images they had hoped for. We had greater success because the development of digital video post production techniques allows for the careful adjustment of the alignment of the footage, to compensate for inherent inaccuracies within the filming process.


The 3D Aurora project was made possible with funding from The National Endowment for Science, Technology and the Arts (NESTA). Additional support came from The Arctic Academy (Finland), Sodankylä Geophysical Observatory (Finland) and Apple Computers.