A new history of computer graphics and software development
Rewriting the history of free software and computer graphics
Do you remember those days in the early nineties when most screensavers were showing flying 3D metallic logotypes? Did you have one?
In this article, I want to go back in time and briefly revise the period in the history of computer graphics (CG) development when it transitioned from research labs to everyone's home computer. The early and mid-1990s was the time when Aldus (before Adobe bought the company) was developing PageMaker for desktop publishing, when Pixar created ToyStory, and soon after 3D modeling and animation software Maya by Alias|Wavefront (acquired by Autodesk). It was also a moment when we got two very different models of CG development, one practiced by the Hollywood entertainment industry and one practiced by corporations like Adobe and Autodesk.
By recalling this history, I hope to be able to shed new light on the value of free software for CG, such as Blender or Synfig. Maybe we can even re-discover the significance of one implicit freedom in free software: a way for digital artists to establish relations with developers.
Where did Adobe and Autodesk come from?
It may seem natural today that CG software and GUIs empower digital creativity, but in fact, they were to a great extent the result of the attempts of the software industry to automate design and manufacturing.
In the early 1980s the U.S. felt threatened about losing its global economic power. The way it sought to overcome this threat was to envision software as a tool to reduce costs in manufacturing, by automating parts of it. CAD software was in great demand, primarily as a way to reduce the costs for manufacturing, but also to increase U.S. competitiveness. It sped up the development of algorithms for digital drawing and rendering. In this context, Autodesk managed to emerge as a market leader with its AutoCAD and outcompeted many other similar products with a cheap price and features. (At that point in time CAD software used to be developed in-house for large corporations and cost tens of thousands of dollars. However strange it may sound, Autodesk was among the first to make CAD software affordable for less wealthy companies at the time, but not, of course, for individual creators.)
This was also the moment in history in which computer graphics transitioned from being part of scientific work to becoming widely available for noncoding users. This move required producers to minimize maintenance obligations and costs by creating documentation and support services that detached users from developers. However, in the 1990s the business of CAD and desktop publishing software companies (like Adobe) stagnated and needed to diversify.
These growing CG software giants began to swallow smaller companies and incorporate these tools into their portfolios. As a result, individual CG creators got digital tools for drawing and animation, but dispossessed them from influence on technological development, a situation that art critique Boris Groys summarizes in his post, Art Production (March 18, 2013): "The post-industrial 'creative industries' presuppose the innovative, project-oriented and, in a certain way, autonomous working process. But on the other hand, the artists, designers, or writers use the means of production that they do not own or control." We may think that this is the default state of things that we end up with ever since the industrialization of work. But, this is not entirely true.
The Hollywood entertainment industry
While CAD software was moving CG away from science (at least for the end user) in the 1980s and early 1990s, the Hollywood industry was experimenting with renewing animation by new tech and by bringing users into software development. Initially, such experiments were rather limited. Somewhere around the 1970s, Lucas Film tried to experiment with CG for making visual effects. Disney tried in the 1980s to introduce CG as a way to automate parts of its animation production. These tests did not have a big impact because these studios had a very well established division of labor, technology, and production pipeline. Any changes caused by new technologies would disrupt the ongoing work process and require major restructuring that they were reluctant to do (cited below, Tai).
For this reason, their early experiments were more of a test of what it would be to have some CG in a movie rather than an attempt to actually develop new algorithms. This was changed by Pixar.
A landmark moment is the call made by John Lasseter to the CG engineering community, through his 1987 SIGGRAPH paper, to improve the visual appeal of animation by redefining Disney's principles of motion for 3D. This paper was the first one in SIGGRAPH that came from an artist and not an engineer. It showed the need to bring artists and CG developers together in a conversation and to jointly improve both technology and the visual expressions that came with it.
A parenthesis in history
In fact, the realization that artists need to be part of developing CG and computing was made in the 1960s through the work of an organization called Experiments in Art and Technology (E.A.T.). The organization tried to attract the computer industry (Xerox, IBM, Hewlett Packard, and Bell Labs) to fund experiments between electronic engineers and artists. E.A.T. thought that "corporations could extend their innovative capacity by providing artists access to their prohibitively expensive tools... Artists would realize their vision, engineers would learn to do different things—look at things differently, and companies would harvest the ideas and patents…." (cited below, Wisniowski). Some of the experiments that this organization managed to facilitate became exemplars of the nascent-at-the-time multimedia and technological arts.
Artists and engineers together would explore holograms, lasers, video, and computer graphics. For example, "Manfred Schroeder investigated the information content of visual images together with Leon Harmon and Ken Knowlton, to make one of E.A.T.'s most renown works called 'Computer Nude' a computer plot of a nude compiled from mathematical symbols" (cited below, Wisnioski).
This 1967 image was one of the first computer-processed prints in the history of digital graphics, and it was exhibited at the Museum of Modern Art in the first computer art exhibition to be arranged in New York, which was named "The Machine as Seen at the End of the Mechanical Age." Through such works, the engineer was also redefined as "an artist who begins with an idea or need" and then uses his "special tools," constrained by time and budget, to bridge the "creativity gap" between theory and reality (Wisnioski).
Back to the entertainment industry
Ok, so, fast forwarding (rewinding?) back to the early 1990s, what Pixar, and later Dreamworks, SONY, and Ghibli realized was that there was no way to renew a visual medium without messing up with technology and involving artists and users in the process. Ever since, even if their general production process still resembles a factory with a hierarchical management, the development of artist vision and technology for each film go hand in hand.
Historically, they have tried to downplay the importance of technological development in public opinion, because the industry wants people to believe that their products are made by exceptionally talented artists (something very evident in some of the writings of Ed Catmull, Pixar's co-founder, for example).
This is changing now, with Pixar itself celebrating its technology and the possibility to bring together artists and developers through videos like The magic ingredient that brings Pixar movies to life. Yet, the Hollywood industry makes money not out of technology, but out of the content that it brands and circulates. So, we end up in a situation in which we have software companies that develop CG software, but users not being able to experiment with the algorithms, nor can they contact developers and experiment with alternative techno-art visions.
We now have a Hollywood industry that further pushes the development of CG algorithms and what can be done with them without making the tech available (or only sharing it to a very limited degree). There have recently been signals that this may be changing; for example, Ghibli made their Toonz animation software available for Linux users. But so far, sharing technology remains an exception.
The significance of free software for CG
On the backdrop of this history, free software like Blender, Synfig, Krita, and other projects for CG gain significance for several reasons that stretch beyond the four freedoms that free software gives.
First, free software allows the mimicking of the Hollywood industry's models of work while making it accessible for more individuals. It encourages practice-based CG development that can fit individual workflows and handle unexpected circumstances that emerge in the course of work, rather than aiming at a mass product for all situations and users. Catering to an individual's needs and adaptations of the software brings users work closer to craft and makes technology more human. Tools and individual skill can be continuously polished, shaped, and improved based on individual needs, rather than shaped by decisions "from above."
This sense of craft gets stronger as artists get in a direct contact with the developers of their tools of work. When they do so, artists and users start guiding technological development, and software engineers change into being servants rather than masters of technological development. I have written more about this in Free software beyond radical politics.
Perhaps the most important quality of free software for CG is precisely this one: to encourage connections between users and producers of technology, reshaping the ways in which technology gets produced nowadays. What happens when these connections get established?
Food for further thought
My observations about the effects of these connections are based on some research I have been doing on the Blender and Synfig communities. One obvious outcome is that users gain a greater freedom and advantage to move between freelancing jobs across the world, because technology and a community to help are always easily available.
Surprisingly, most Blender users whom I have met in the past three years do not turn this into their advantage and try to escape the freelancing condition nor do they experiment radically with creating new aesthetic expressions with CG. Instead, the goal of many seems to be to try to reproduce the stylistic visions of Hollywood and bring them to the products they make for television, advertising, and other industries. There are some exceptions, of course, and there are artists who experiment with the medium (for example, the work of the ELI KURUS collective: Platonic Solids 3-D scans of plasticine sculptures 2015), but as a rule they end up being somewhat marginalized and receive less recognition by the community.
The freedom of users to connect to developers seems to bring so much satisfaction that it shifts away focus from problems of employment. The rush for catching the new job and updating portfolios seems to remove time from reflecting on what could be changed in the form of work, which occurs when the form of engaging with the technology has changed and been brought back to craft.
After all, short-term contracts and constant movement between continents might be a cool lifestyle, but they benefit economically mostly the industry that decreases its production costs and responsibilities for creators. Technology and a specific organization of work have given the birth of this very entertainment industry that follows the factory-like model and pushes everyone to constantly move around on short-term contracts. I wonder why more people, and not just the Blender Institute, are not trying to create alternative models of work for making computer graphics.
Tai, P.-y. (2012). The Principle of Animation: History and Theory of a Social Technology. Doctoral Dissertation. University of California Irvine, Irvine.
Wisnioski, M. H. (2012). Engineers for Change: Competing Visions of Technology in 1960s America. Engineering studies series. MIT Press, Cambridge, Mass.