A Total Portrait With No Omissions: Photogrammetry “at the point of a singular horizon”
This essay explores the pre-history of 3D scanning for Ren Gregorčič’s “at the point of a singular horizon” at the M16 Artspace in Canberra.
Consider the garden plot. It’s a small hole in concrete, filled with dirt, shaped and positioned in plazas and parking lots. Within those concrete frames, plants, soils, insects and microbes converge into tiny isolated worlds. We’re meant to look at the worlds inside them. But the frame isn’t as hard a border as we think. Concrete brings its own chemical and microbial contributions to these tiny ecosystems, its density determines the flow of water.
It’s here, at the imagined borders between natural, built, and virtual worlds, that Ren Gregorčič’s at the point of a singular horizon charts its strange terrain. In immersive images and video, slabs of concrete fuse with leafy greens. Shadows move across trembling weeds, competing with computational sunsets. At first the scenes seem natural, even calming. But a close examination reveals that they are, in a strange way, incomprehensible. These aren't bucolic woodland scenes. They are collages, assembled through a process of automated digital pastiche, reassembling and predicting the behavior of their sources: a garden, in a box, inside another virtual box, creates another kind of garden. And along the way, borders between concrete, digital and rendered realities collide and confuse themselves.
Gregorčič’s work is the product of Photogrammetry, what we might typically call “3D scanning.” Take a bunch of photographs while walking in a circle around an object, then connect the edges of those photographs until you — or a computer program — can map them to a three dimensional model of that object.
While photogrammetry emerged from a desire to document historical objects and structures before they disappeared, the method evolved, literally, from a collapse.
In 1858, Albrecht Meydenbauer, a young Prussian architect, was documenting a cathedral. At the time, this method involved climbing the building with measuring tapes and a compass, like a vertical explorer. When Meydenbauer nearly fell off the side of a spire, he wondered if there might be an easier way. He turned to an upstart technology: the film camera.
The idea was simple. What if he could measure the height of the building in pictures, and keep the climbing gear at home? Meydenbauer modified his cameras into technical instruments, relying on the concept of the vanishing plane. To understand this, imagine a road stretching away from you in a desert landscape. The lanes of the road never touch, but as it moves away from you toward the horizon, it appears to shrink into nothing. Apply this principle to the lines of buildings in photos, and a bit of algebra, and you’re able to calculate the edges of a building’s borders, and thus, their size.
Meydenbauer took countless photographs of buildings: same distance, different angles, piecing them together into two dimensional maps of three dimensional territories. One object was always kept in frame to provide a common reference point.
Along the way, he was creating an archive of the world that would endure beyond the loss of buildings and artifacts. He was providing a way to return to them, virtually, once they were gone — for example, Meydenbauer was able to “preserve” the ruins of Persepolis in his images. Today, we use the same process, though digitized, to step into virtual tours of lost World Heritage Sites in Syria: the borders of thousands of images stitched together to create a seamless, immersive recreation of lost places.
The field has dissolved into the everyday: you can do this to get candy for your Pikachu in Pokemon Go, walking around objects in the world to “scan” them into your phone. The history of film-based photogrammetry is still evident in strange places: the digital grid that we see in 3D image editing software, for example, can be seen in this anachronistic Vaporwave-esque image from 1924.
We’re also seeing the edges of the photo assembly here, the fitting together of multiple images to create one larger whole, a process of photo stitching that continues in today’s 3D modeling environments. Photogrammetry is also used is overhead documentation of spaces, such as aerial photography. In the image on the right, we see Air Force intelligence piecing together overhead film photographs to create a “live” map of enemy territory. Combined with digitals tools today, we’re able to automate this work into massive online maps, though the stitching lines of satellite photos is still evident. By bringing 3D modeling into the mix, buildings and topographies can be estimated for these maps, giving us a sense of elevation and measurement for the entire world.
The once theoretical concept of a life lived through screens moved from cyberpunk fiction to lived experience for much of the world in 2020. The technology has kept pace, and photogrammetry is far beyond the hand-aligned photographs taken in Meydenbauer’s day. Some transcend the idea of “images," relying on light and soundwaves to imagine a digital version of the world. This move from analog to digital photogrammetry pushes the math toward more complex algorithmic models. We can take more images at faster speeds, creating digital models of the entire globe from orbiting spacecraft. Photogrammetry has moved from a box camera outside of a German cathedral to an autonomous vehicle wandering alone on the surface of Mars. It is a boon for the archive and preservation of architecture, of bringing something lost to time or distance closer to us.
All of which brings us back to the garden plot, and rethinking our scale — moving away from distant observers of vast landscapes into intimate observers of small ones. How might it change what we see and how we see it? And what do we make of those borders then — those stitch marks between worlds of the camera frame?
Photogrammetry as a digital process moves from the simple film photograph we might associate with the small intimacies of a snapshot. It is about extrapolation of data from a photographic image, and then reassembling that data into a new, three-dimensional image; a process that is unnecessary for projects of intimate spaces. But this is what Gregorčič’s work does. It shows us the negotiations between different kinds of borders and frames — the physical, the natural, the data-driven — and compresses them into new ways of seeing.
It involves thousands of digital photographs wrapped around predictions of the landscape. The software looks at the images and models the terrain based on what it sees. Then it lays those actual images over the shapes of that imagined terrain. This produces images like those above, which are a collage, a sculptural form in digital space that emerges from the pastiche of bits and pieces of the original “thing” being photographed. The more data, the more precise the model gets put back together. Without a standard reference point — the grid of perspective in that mountain range from 1924 — the prediction slips. The ground falls out beneath our feet. When those assumptions and photographs do not connect, images are stretched and pulled into strange distortions.
When data is missing from one photograph, the system fills those gaps with informed assumptions from other photographs. That's why at the point of a singular horizon isn't glitch art: technically, there are no “errors” in these images. They are the system's best attempt to reconcile information competing at the boundary of the digital and observed worlds. In the garden plot, the concrete mediates the natural space. In rendering this as a 3D object, the observed and digital begin to seep. Renderings of the relationships between these boundaries is at once familiar and disorienting. There is a fuzziness in the process that you can see in these results, which feel like a computer dreaming of gardens.
I’m reminded of the nursery rhyme about Humpty Dumpty: the egg shells shatter, and all the King’s horses and men can’t put the egg-man together again. Photogrammetry does what knights and horses can’t: it can reassemble buildings from an archive of shattered pieces, thousands of photographs split apart and reassembled into a virtual Humpty Dumpty. The cracks still show in the seams, and Alice couldn’t have a chat with him: the original is broken. You can only remember the egg you have lost. After a global pandemic and environmental catastrophes such as 2020's bushfires, a digital archive of garden plots — natural spaces already shaped and defined by concrete — evokes something irretrievable. It feels to me like a science fiction archive of vestigial green spaces surrounded by concrete, as if it was all we had left of nature: digital time capsules for when even a garden plot is as far away as the architecture of Persepolis, an archive of a physical world teetering on collapse.
I come back to that core definition of photogrammetry and it’s lack of physical contact with the world we want to understand. These images are seductive, but distant: calming, meditative, but nonetheless, approximations of nature by a machine that cannot touch the world. If there was something whimsical and naïve about preserving the world through measuring sticks and compasses, the physical contact made it seem real and verifiable. Concrete touches the garden's edge, but the digital model watches from afar, telling what it sees from the logic of detached surveillance. at the point of a singular horizon reveals the unevenness of interaction at a distance. These works live between things, and emerge from the edges of a physical world and the collection of its measurements.