A lifelong explorer, Herbert W. Franke was born in Vienna in 1927. After serving in WWII, he earned his doctorate in physics while pursuing a side interest in photographing caves—a passion and occasional profession that followed him throughout his life, across expeditions from the Mediterranean to the White Desert of the Sahara.
He was a writer, too, and in the late 1940s, he turned his focus from poetry to science fiction after an editor, not fully impressed with his verse, suggested that Franke’s scientific studies might allow him to offer realistic depictions of the far future. This was prescient advice, not just for Franke’s literary pursuits but for the work he would become best known for, as one of the first practitioners and theorists of computer art.
“My idea of art was and still is: to venture into unknown new territory,” Franke told The Brooklyn Rail in spring 2022, promoting a career-spanning solo exhibition that coincided with his 95th birthday. It was also his last: Franke died on July 16, but not before witnessing, thanks to the generative art boom enabled by NFTs, the widest-yet proliferation and vindication of his foundational ideas: that mathematics, art, and computer science go together beautifully.
Like his other explorations, Franke’s art began as a hobby. In the 1950s, unable to find work in physics, he joined the advertising department of Siemens & Halske, an electrical engineering company. After hours, he used the firm’s scientific equipment to conduct a series of aesthetic experiments. For instance, using extra-long-exposure photography, he captured the ethereal, Escher-like light and waveforms produced by an analog oscillograph, a device that visualizes electronic signals.
Over the decades, the look and feel of his work changed as he ventured wherever technological advances might offer something new. In the 1970s, he drew fractals with a pen plotter, a precursor to the modern printer that mechanically guides an ink pen based on computer prompts. Then, with the advent of home computers in the early 1990s, he modeled game theory in 2D diagrams that looked like millennia-long family trees. In the 2000s, using more technical software, he crafted 3D models of complex equations and placed them like statues in virtual worlds. Earlier this year, he even minted 100 works from his Math Art series—a brilliantly colored spin on Pop Art he produced in the 1980s and 90s—as NFTs. “It’s almost impossible to create something using generative tools and [for] it to not look like something [Franke has] done,” the artist Kevin Abosch remarked recently.
From the dawn of the computer age, Franke recognized something that most of us now take for granted: that the raw material of art—color, shape, sound—was also mere data. Using the computer as a tool and a creative partner, he manipulated that data to study its effects, all the while advancing the idea that the code and principles behind generative art were just as worthy of our appreciation as the mesmerizing outputs they produced. “I approached art as a scientist,” he told Outland. “I wanted to apply analytical methods to figure out why humans find some artworks attractive and beautiful and others not.”
Given the human brain’s limited capacity for processing information, how might varying levels of complexity impact a viewer’s experience of a piece? Is there an optimal middle ground to draw us in? And what sort of patterns might reward repeat interactions? These questions mattered to Franke, whether the subject was an image, an animation, a song, an essay, or something as monumental as existence itself: “How did organized life first establish itself on earth[?]” he asked in his 1971 book, Computer Graphics—Computer Art.
Though it was undoubtedly inventive, Franke’s early work was largely ignored—and that’s partly what led him to write the book in the first place. A comprehensive history of early computer art, the book situated the medium (and, implicitly, Franke’s own artistic practice) within a broad lineage of “mathematically conceived figuration” that stretched from scientist-artists like Leonardo da Vinci and Albrecht Dürer to Piet Mondrian and Le Corbusier. For future generations, Franke’s work would be the next link in that chain. By offering up that legitimizing historical precedent, he helped open the door to something more exciting: the production of entirely original digital forms, made possible through animation, combined media, and especially interactivity.
That new terrain has certainly been fertile ground for NFT artists lately—like Tyler Hobbs and Dandelion Wist, whose QQL collection (2022) distinguishes between its outputs and the work of its coders by design. Hobbs and Wist built a generative art algorithm that anyone could play around with, manipulating parameters like color palette, object spacing, and “turbulence” to create free previews to their heart’s content. Only 900 outputs can be minted, and collectors pay upwards of 14 ETH for the privilege. But exploring the tool is available to all, incarnating the Frankeian impulse in which the computer, in collaboration with a human creator, becomes a teacher of what’s possible.
In one of his last interviews, the self-described “dinosaur of computer art” spoke with Tyler Hobbs. They discussed Franke’s earliest experiments, his rational theory of art, and why looking into the future might not be as helpful for artists as understanding where we come from.—Duncan Cooper
Tyler Hobbs: It’s an honor and a privilege for me to have a conversation with you about your work. Thank you for agreeing to participate. Knowledge about the origins of algorithmic and computer-based artwork can only deepen our appreciation for those early works.
Herbert W. Franke: Thank you. It’s an honor for me, too.
TH: What was the context when you first started creating algorithmic artwork? What kind of computer was it, and what kind of access did you have to it?
HWF: In some ways, all of my art is “algorithmic,” because the artist can “construct” images in this way—even without a computer. In my case, it was photographs generated analytically with light. According to my definition, this is already algorithmic art.
The first computer I used was an analog computer built for me by a fellow physics student in 1954. It allowed me to perform basic arithmetic operations, such as subtraction and multiplication. I used it to generate oscillations that I could display on a relatively small oscillograph [a device which displays and records electrical current and voltage]. However, because its screen was only five centimeters in diameter, I moved the camera in front of it with the aperture open, resulting in curve multiplications and superimpositions.
As a student, I wondered why some scientific images are perceived as “beautiful.”
TH: Before the first algorithmic work, there was no clear idea of, or model for, what “computer art” aesthetics might look like. Did you find yourself using references from mathematics or nature as “nearby” starting points for your designs?
HWF: Of course. That was essential for me. I do have a physics background. As a student, I wondered why some scientific images are perceived as “beautiful.” That immediately made me think that the fundamental principles of aesthetics are not limited to art but are quite universal, and can thus be applied to scientific visualizations as well.
Simultaneously, it was clear that these principles could be grasped mathematically. From the very beginning, I was searching for this mathematical code of aesthetics. However, when I designed my art model in the 1960s, which was based on information theory, the sensoric input of information was fixed, but this was not enough for me. In contrast to other scientists who dealt with information aesthetics at this time, it is information psychology. For example, emotion, which I consider to be an essential component of art perception, cannot be defined solely in terms of pure information flow. Emotions are different neuronal phenomena—also with a hidden code, of course. But this code is much more complex, and we are far away from understanding it. In aesthetic perception, information flow and emotional processes must be brought together.
TH: I would assume that your earliest work relied on plotters, and that screens and printers came later as output formats. Is that correct? If so, how did that transition affect your work?
HWF: As said before, my earliest works used cameras and oscillographs, so there were already screens involved. The first two digital series—Squares (Quadrate) (1967/1969) and DRAKULA (1970/1971)—were created on a plotter [a computer hardware device used to print vector graphics]. However, as early as 1973, I used a system named Sicograph, which was located in the Siemens research laboratories, where I could occasionally use the latest medical technology for artistic purposes in the evenings.
The Sicograph was one of the world’s first systems for digital picture processing of analog X-ray images. The mail output device was one of the first inkjet plotters worldwide, but it was also equipped with a small monochrome preview screen on which the Serie Grün [a series of single images for a photographic soft-cross fading system with two Kodak projectors] was created. A follow-up computer developed by Siemens, the Bildspeicher N [a device used to render diagnostic imaging], already had a large color screen as an output device. I used this to produce series such as Einstein Digital (1974) and Digitale Impressionen (1973).
TH: When I create algorithmic artwork today, I rely heavily on having a quick feedback loop. I can change the code, run it, and, within a few seconds, see how the work has changed. I know you didn’t have that luxury, especially early on. Can you describe your feedback loop and how that affected your working process?
HWF: With some of the machines I worked on in the 1970s, immediate feedback was quite possible for me. And, to be honest, it was very important to me. I always found it appealing about computer art that you didn’t have to wait as long to see the result, and then have problems with subsequent optimization, as you did with traditional painting. You could work interactively with algorithms—I found that very nice. All my programs developed on the PCs from the 1980s onwards were created interactively.
I’ve always been fascinated by ornamental art and I’ve tried to find and apply such mathematical codes [to it].
TH: Did working on algorithmic artwork change your perception of traditional artwork or cause you to be interested in a traditional art form that you had previously ignored?
HWF: I’ve always been fascinated by ornamental art and I’ve tried to find and apply such mathematical codes [to it]. I’ve also looked to mathematicians for aesthetically translatable structures and ordering principles, such as Stephen Wolfram’s cellular automata.
TH: What makes an art algorithm good, in your opinion? Did you have particular aesthetic goals that you tried to achieve?
HWF: I have been very intensively concerned with the question of how aesthetics come about. Throughout the process, I gained insights for my rational art model based on human perception, which I have naturally applied to my art—both visual and literary. A mere algorithm is usually boring, so it must be broken, for example, by random processes. It is also important to provide the viewer with complexities that require different perception processes, resulting in a work that has an impact that extends beyond the moment.
TH: Did you have opportunities to show your early work in exhibitions? If so, how was the work received? I know the general history is that early computer art was poorly received, especially by art critics. Did that match your experience? Did that change how you felt about the art form?
HWF: My first solo exhibition, Experimentelle Ästhetik, took place in 1959 at the Museum für Angewandte Kunst in Vienna and featured my early works of photographic art up to oscillograms I created with my analog computer. Then there were numerous exhibitions, including the large exhibition, Ex Machina in the Kunsthalle Bremen (2007), which featured my extensive collection of computer art. Wanderer between the Worlds was a major solo exhibition at the Zentrum für Kunst und Medien in 2010. My most recent solo exhibition, Visionary, took place at the Francisco Carolinum in Linz, Austria (2022).
TH: How do you think computers may change humanity’s aesthetic preferences over time?
HWF: I’d ask a different question: How do computers actually impact people’s perceptions in general? Because the effect of aesthetic perception extends far beyond art. Certainly, the modern world’s information overload, accelerated pace, and decreased ability to focus on something for an extended period of time have an impact on art. However, in my opinion, these are rather minor aspects of the package. We are basically Stone Age people, so art still operates on the same basic principles as it did back then. And it will continue to do so, although the world around us is changing dramatically.