menu

1. Introduction

This research follows the process of discovery and reflection that proceeded from my first use of 3D computer animation software in 2001, and my subsequent familiarisation with the software.

Research aims and methods

The aim of this research is to expand the practices and visual styles associated with computer graphics (CG) software by developing approaches to 3D animation that avoid reductive frameworks (such as standardised workflows and abstract mathematical models).

Using theoretical and practical investigations, I explore assumptions and biases implicit in the design, marketing and orthodox use of 3D animation tools. Informed by these investigations, as well as by reflection on my prior practice, the research proceeds via a practice-led enquiry that involves creating a series of short experimental 3D animations, as well as several customised 3D animation tools.

Background

After many years using traditional art media such as oil paint, charcoal and pencil, in 2001, while studying at RMIT University, I was introduced to 3D computer animation software. During this time I learnt how to use 3D Studio Max and made a short animated film called Drive (Moore, 2001). Making this film was a learning process and I didn’t have much control over the animations that I produced. I followed some tutorials, but mostly I experimented/played with the 3D tools and responded to imagery that the software came up with. Even if it wasn’t what I expected, I might accept what I saw, revise it or reject it altogether.

The fact that I had to compromise my “artistic vision” because of my lack of skill didn’t, at the time, seem like a good thing. But looking back on it now it seems that my unorthodox use of tools, my loose pictorial goals, and the way that my focus was on responding to what was given combined to form an approach to animation which sometimes resulted in imagery more intriguing than I could have designed in advance.

Figure 1.1: Stills from “Drive” (Moore, 2001).

Excited by the seemingly infinite potential of this new medium, I spent subsequent years working professionally in the 3D animation industry and learnt standard 3D animation procedures. Eventually, I felt equipped to convincingly illustrate any subject matter. However, this mastery, autonomy and empowerment came at a cost. Having gained a high level of technical proficiency, my 3D animations no longer felt like the result of a collaborative process because I was now the one calling the shots. It seemed that the software was now at the service of my creative design, my ideas and my imagination. However, instead of being full of discovery, the making process had become a necessary (and often tedious) set of steps required to reach a specified goal. A finished work either achieved this goal or it fell short; the work seldom exceeded or altered my expectations. Over time, I became aware of something else happening, which was indistinct, but disquieting, and unnerving.

My hard-earned knowledge of standard 3D procedures seemed to define my experience of the software, making it hard to play or experiment. It also influenced the way that I engaged with the world around me.

This research follows the process of discovery that proceeded from that initial spark of disquiet, inviting me to look below the software’s glossy surface.

Document outline

Section 1: Introduction explains the aims, methods and motivation for my research. This section introduces themes relevant to my intuition that computer software can have a reductive role in creative practice.

Section 2: Context of practice discusses a selection of 3D users and digital artists who expand the visual styles associated with computer graphics (CG) software by exploring qualities inherent in the medium and by being open to opportunities that emerge in the process of making. Unlike orthodox approaches to CG software, these artists don’t seek media transparency, either in their tools or in their work. We hear from artists using traditional media (such as Francis Bacon and Frederick Franck) who emphasise the importance of fostering an actively receptive comportment through their work, and who adopt particular creative strategies to do so. The work of digital artists discussed in this section indicates that an actively receptive comportment might also be important for those aiming to expand practices and visual styles associated with 3D software.

Section 3: Things themselves further examines the idea of an actively receptive comportment through a discussion of phenomenology. The basic project of phenomenology is to avoid the imposition of reductive abstract models and to “return to the things themselves” (Husserl, 2001, p. 168). This is a call to explore the inexhaustible richness of our everyday experience of things and notice what can’t be described in words, explained using scientific concepts, or represented using computational [algorithmic] models. The things with which my research is concerned include everyday physical things in the world around me as well as 3D software and the 3D animations which this software can produce. How, and to what extent, can I approach these things without being limited by standard practices and algorithmic models? The work of phenomenologist philosophers Edmund Husserl, Martin Heidegger and Maurice Merleau-Ponty together with that of the artists discussed in Section 2 and 3, suggests a number of approaches that will inform my practice-led enquiry.

In Section 4: Experimental animations I describe a practice-led enquiry aimed at developing new approaches to 3D computer graphics software. Drawing on themes outlined in previous sections, this practical enquiry involves creating a series of short 3D animations which cultivate an actively receptive comportment. Rather than focusing on achieving predictable and repeatable results, in these experimental animations I often work with computer glitches, treating them as aesthetic suggestions, to which I intuitively respond.

In many experimental animations I use 3D software to explore perceptual experience; sometimes working from pencil sketches and sometimes working directly from life. In addition, I explore ways to remap the user’s actions to a visible outcome so that activities involved in making a 3D animation are evident to viewers of a finished work.

Section 5: Outcomes and findings discusses outcomes and findings from the practice-led enquiry, covering 3D animations, customised 3D animation tools, and a number of creative strategies used throughout the project. These custom tools and creative strategies all involve a shift in focus from control to response. Some tools and strategies involve responding primarily to suggestions from the software (including suggested form and suggested movement) while others respond primarily to everyday physical things in my local environment. Whether responding primarily to digital things or to physical things, 3D animation practice is appreciated as a type of conversation (or collaboration) between the user, the software and other physical things. In addition to 3D animations, custom tools and creative strategies described in this document, a major contribution of this research is to conceptualise a conversational approach to 3D animation practice which involves a shift in comportment from control to response.

Section 6: Discussion outlines an expanded conception of style, arguing that all things exhibit a style and it is useful to think of 3D animation as a stylistic exchange between the style of an animator and the style of other things. Reflecting on the experimental animations created in this research I notice that, in addition to my style of seeing, finished works also embody a distinctly non-human gesture reminiscent of digital objects and object oriented programming (OOP). This insight leads to a short discussion of object oriented ontology (OOO) which conceptually situates humans on the same level as other things (including physical, digital, real and imagined things). Like an expanded conception of style, OOO encourages us to think in terms of collectives and entanglements.

Ginger and Default Whippet

Throughout this document standard 3D animation practices and procedures are primarily discussed through an account of Default Whippet (Figure 1.2), a 3D animation completed at the outset of this research.

Figure 1.2: Default Whippet 3D animation.

Like many of the animations discussed in this document, Default Whippet is based on my dog, a whippet called Ginger. My constant companion throughout this research, Ginger is lying beside me in her bed as I type this document and every now and then I turn away from the screen and move my hand from the keyboard to stroke her fur.

The sketching process

Figure 1.3: Detail of page from my sketchbook.

Before commencing Default Whippet, I took my dog Ginger to the park and sketched her as she wandered around or lay on the grass. Observational sketching is not a typical way to start a 3D animation project and ultimately my drawings of Ginger were not very useful in the creation of the Default Whippet 3D animation. However, as a way of interrogating activities involved in standard 3D practices, observational sketching provides a useful point of comparison.

Requiring only basic tools – paper and pencil or pen – sketching is typically a simple activity but it can reveal a world of infinite complexity. While sketching, I often notice the way that shifts in attention or mood change what I see. I also notice that the activity of sketching is itself a way of seeing and it is different from how I normally see the world. While sketching I pay attention to what things look like rather than moving among them and hardly noticing them at all. And if I were to swap my pencil for paint brush, I would also see the dog differently (e.g. I might notice the texture, colour or tonal variation across Ginger’s coat rather than focusing on her silhouette). While sketching Ginger in the park, I visually explore her contours through the pencil which has become an extension of my body, i.e. my eyes travel across her form as I intuitively make marks on the page. I am mostly seeing and responding to the dog but I also look down at the page and respond to the sketch as an emerging figure or gestalt. Ginger moves around, but even when she’s still I notice her muscles twitching and her nose sniffing the air; it’s obvious that she’s making sense of a world that I can’t even imagine.

Figure 1.4: Sketchbook page showing some of the sketches I made of Ginger in the park.

Drawing can be a way of seeing or a mode of contemplation – but what about the resulting artefact; the drawing itself? The activity of sketching can open onto a world which is so complex, elastic and inexhaustibly rich that there’s no question of trying to pin it down or “accurately” represent it with graphite on paper. Sketches are often sketchy and this is especially true when, like Ginger, the objects of study are in constant motion. The finished sketch is indicative or allusive; it is an attempt to capture something which is obviously out of reach or beyond my grasp. In perceptual experience there is always more to see and I have the feeling that what I am seeing could always be different.

When it comes to the drawing as an artefact, qualities of sketchy incompleteness can hint at more than is explicitly depicted and hence leave the work open to interpretation. The sketch, like perception, is always incomplete.

Drawing as a way of seeing

In his book The Zen of Seeing: Seeing/Drawing as Meditation visual artist Frederick Franck describes his drawing practice as a mode of contemplation in which the world is experienced afresh at each moment. He says, “Drawing is the discipline by which I constantly rediscover the world. I have learned that what I have not drawn I have never really seen, and that when I start drawing an ordinary thing I realize how extraordinary it is.” (Franck, 1973, p.6) Franck says that each of his drawings is an “adventure” for which he can’t predict the outcome (Franck, 1973, p.6). He believes that it’s important to avoid drawing systems or recipes (such as those found in a book on “how-to” draw horses) because, although they may be useful for a “picture manufacturer”, they deprive you of knowing what things really look like; they stop you from really seeing (Franck, 1973, p.55). He says:

In order to draw a horse, draw horses until you practically become a horse – not “horses in general” but that particular horse you are drawing at a given moment. Until you feel the tense curving of its neck in your own neck! (Franck, 1973, p.55)

Figure 1.5: Drawing of a horse by Frederick Frank. Sourced from “The Zen of Seeing” (Franck, 1973, pp. 56–57); Copyright 1973, Frederick Frank.

Of importance to Franck is the particular state of awareness that he cultivates through his practice and it’s clear from the quote above that, for him, drawing involves an empathetic kind of engagement. Franck calls his practice Seeing/Drawing and insists that for him seeing and drawing are not two distinct activities but rather “one single undivided act” (Franck, 1973, p.ix). We could say that Franck sees (understands, describes or explores) things (such as a horse, a landscape, a human figure) through his medium (pencil and paper). A 3D animator also sees things through their medium but it seems to involve a different kind of seeing.

A typical 3D workflow

Default Whippet (Figure 1.2) was made using a 3D character animation workflow which involves several distinct steps including modelling, texturing, rigging and animation (see Figure 1.6).

Figure 1.6: This diagram is comprised of stills from Default Whippet and illustrates some of the steps involved in a standard 3D character animation workflow.

Observational sketches of Ginger were ultimately not very useful in the creation of the Default Whippet 3D animation because they are idiosyncratic, inaccurate and incomplete. In accordance with the standard 3D character animation workflow (illustrated in Figure 1.6), I needed to create the dog at the origin of the Cartesian grid and in a generic or default pose. This means that the dog needs to be looking straight ahead with limbs straight and body symmetrical. This symmetrical pose can be seen in Figure 1.7 which shows the finished Default Whippet model.

Figure 1.7: The final Default Whippet model. Note that an internet connection is needed to access the 3D data associated with this Figure. Click on the arrow to load the model and then use the mouse to view it from any angle.

To achieve this, I engaged the help of a colleague who photographed Ginger while I physically held her in position. I then manipulated the photographs using Photoshop, creating a series of composite images which depict Ginger in a symmetrical pose as seen from the side, top and back (Figure 1.8). Collectively, reference images such as these are commonly known as a model sheet, and they act like a set of architectural plans or blueprints for the creation of a virtual mesh.

Sometimes imported into the software’s viewport, model sheet images are often used as a guide for building a mesh. When sketching, drawing and seeing are one activity and each drawing is an adventure. When animating with 3D software, by contrast, we are often working toward a predefined image.

Figure 1.8: The image above left shows composite photographic images of Ginger laid out as a model sheet. On the right is a viewport snapshot showing how these images were used as a guide to model Default Whippet.

As I progressed with Default Whippet I soon found that my photos of Ginger didn’t give me all the detail that I needed and so, like most 3D projects, I used the Google search engine to download a collection of source material from the internet. Figure 1.9 shows a sample of this collection which includes digital photographs, movies and anatomy diagrams. I referred to these images while modelling the dog and also during the rigging phase of the project, when I felt the need to study dog skeletons, musculature and anatomy as well as a variety of common whippet poses. Upon reflection, it seems strange that, throughout Default Whippet, I constantly referred to my vast collection of digital files (including downloaded images and movies as well as my digital photos of Ginger), but I hardly looked directly at the dog who lay right beside me as I worked.

Figure 1.9: Sample of the many internet images and movies which I referred to when creating Default Whippet.

What sort of creative medium is this that first requires me to manipulate the dog into a strange position and then to turn away from her completely in favour of images and movies sourced from all over the world?

Turning away from the local and contingent

Instead of responding to the dog in front of me, in Default Whippet I built a virtual world (or at least a virtual dog) based on digital reference. As described by Media theorist Douglas Rushkoff, this is a move “away from the local and toward dislocation” (Rushkoff, 2010, p. 41). Rushkoff says that this is one of the biases implicit in digital media. He explains that digital networks:

work from far away, exchanging intimacy for distance. This makes them terrifically suitable for long-distance communication and activities, but rather awful for engaging with what – or who – is right in front of us. (Rushkoff, 2010, p.41)

As an example of digital media’s bias away from the local, Rushkoff describes our obsession with smart phones and social networking sites like Facebook; he was one of the first to offer a now-familiar warning that these digital media have the capacity to alienate us from real world friendships (Rushkoff, 2010, p.43). Rushkoff’s warnings are relevant to contemporary creative practitioners because digital media (including computer software) impacts upon the way that practitioners engage with things in their local environment.

In addition to my collection of digital images, the production of Default Whippet was guided by the software’s help menu and information gained via online forums and tutorials. Throughout production I worked in a variety of locations; sometimes at university using a desktop computer, other times at home, and occasionally I took my laptop to a local cafe. My location made little difference to the outcome of a particular working session; it didn’t matter what was around me because I was absorbed in the screen world and was largely oblivious to my physical surroundings. For 3D users, as long as the software is installed and there is a stable internet connection, it’s easy to conclude that our physical location and context is not relevant.

How can a 3D user engage with their local physical environment? Instead of manipulating a dog into a default pose and then turning away from her completely, what other options are available?

As well as turning away from the immediate physical environment in favour of internet images, 3D animation devalues the local and the particular by assuming that processes should be repeatable and that each work session should be exchangeable with the next. This assumption downplays contingent and contextual features of a particular practice situation including computer glitches, or surprise outcomes of any kind. As with all complex workflows, I encountered a number of surprising and intriguing images while working on Default Whippet. Some were the result of auxiliary processes (e.g. the mosaic design surrounding the digitally painted texture map in Figure 1.10). while others were the result of glitches or mistakes (e.g. the wildly deformed model shown in Figure 1.11 is the result of a skinning error, i.e. an error which arises when the mesh is not sufficiently bound to the underlying joint). These surprising images felt significant because they were not entirely random and they were also not entirely preconceived or designed in advance. I could not easily make sense of the images but they intrigued me.

Figure 1.10: Screenshot from Default Whippet 3D animation showing digitally painted texture map.

Surprising images (and movements) almost always occur in 3D animation production and they are often ambiguous – it’s hard to say exactly what they depict. It’s also hard to say where these images and movements come from: we can’t easily attribute them to the animator, to the software or to the computer hardware. It is common for a 3D animator to be dismissive of these outcomes and to focus instead on imagery that is deliberate, repeatable and easy to understand.

Promoting the assumption that surprising and ambiguous outcomes are mistakes to be avoided is one of the ways in which 3D software filters out complex and unexpected encounters (for both users and for viewers).

Figure 1.11: Screenshot from Default Whippet 3D animation showing skinning error.

In his 2011 book, The Filter Bubble: What the Internet is Hiding from You, online activist Eli Pariser describes how this tendency to filter out complex and unexpected encounters is implicit in digital media more generally (Pariser, 2011).

Filtering out complex and unexpected encounters

Pariser’s book examines the phenomena of web personalisation experienced by users of internet search engines such as Google, and social networking sites such as Facebook. According to Pariser, the result of web personalisation is that people are increasingly unlikely to encounter views other than their own (Pariser, 2011). Pariser worries that “In the filter bubble there’s less room for chance encounters that bring insight and learning” (Pariser, 2011, p.15).

As media consumers, we have always sort information that feeds our own interests and, prior to the ubiquity of digital media, we probably subscribed to a particular newspaper and changed the TV channel to avoid content that didn’t interest us. So what, if anything, makes the internet filter bubble different to prior media? For Pariser the reductive capacity of the filter bubble is more insidious than prior media because it is invisible (i.e. we’re not aware of its workings) and it is automatic (i.e. we don’t chose to turn it on) (Pariser, 2011, p.10). Collectively these features make it easy for us to assume that digital media are neutral, unbiased or transparent.

It seems that the filter bubble can cocoon us in a cosy and familiar world and, as long as we find what we (think we) are looking for, we probably feel that the world is at our fingertips. For users, 3D software can also provide a kind of cocoon because a turn away from the local physical world and adherence to standardised workflows can mean that the user is not open to complex and unexpected encounters (i.e. unexpected turns in their work). The 3D animator usually aims for an efficient production process and avoids surprises but, for many creative practitioners, surprise outcomes are crucial to their work (Bolt, 2007; Bond, 2012; Dijk, 2012; Kentridge, 2012; Sylvester, 1975).

The importance of chance and accident

Figure 1.12: Francis Bacon “Three Studies of Lucian Freud” (detail). Oil on canvas.

Emphasising the importance of chance and accident in his work, twentieth century painter Francis Bacon says “Half my painting activity is disrupting what I can do with ease” (Sylvester, 1975, p.91). Bacon uses existing painting techniques but is “trying to make out of them something that is radically different to what those techniques have made before” (Sylvester, 1975, p.107) and his paintings walk a fine line between abstraction and figuration. He often paints the human figure but he never renders it in detail and never in a conventional manner, instead Bacon’s figures are contorted almost beyond recognition and they seem to emerge out of and twist back into the paint. Bacon’s aim is to present the viewer with recognisable imagery while avoiding illustration and, rather than conceiving of the outcome in advance and working toward a clear goal, it’s important for Bacon that he stumble’s toward an image (Sylvester, 1975, p.17). According to Bacon, it’s important for an artist not to proceed in an entirely predefined, formulaic or rational manner “because the moment you know what to do, you’re making just another form of illustration” (Sylvester, 1975, p.59) and, in order to disrupt his own illustrative habits, Bacon employs a variety of methods such as using a very large brush (Sylvester, 1975, p. 16), throwing paint at the canvas or working while he’s frustrated or drunk (Sylvester, 1975, p.91). Through an exploration of qualities inherent in the paint, he elicits surprising outcomes and then responds to these outcomes in an intuitive manner. Like Franck, Bacon approaches each work as an adventure. Franck pays particular attention to things in the world around him (people, animals, and the landscape), while Bacon is more obviously attentive to the physical properties of paint.

For both Franck and Bacon, deliberately avoiding established recipes or conventions and being open to the complexities of material interactions helps them escape habitual ways of seeing the world.

Can 3D animators aiming to extend the language of their (digital) medium learn something from practitioners such as Bacon and Franck?

When Franck talks about really seeing, he is alluding to an empathetic kind of engagement where he is responding to things based on feel rather than calculation. This type of response is difficult in 3D animation because standard workflows often involve creating a mesh out of context and according to a strict plan.

I created a model of Ginger at the “centre of the world”, in component parts and in a default pose (Figure 1.7). This ordered approach sucked the life out of the dog, but what other alternatives were available? In a pencil sketch or a painting, traces of an artist’s intuitive actions are often visible as multiple overlapping outlines or brushstrokes. It is harder for a 3D user to respond to things in an intuitive manner and (even if this is achieved) intuitive and iterative responses are not usually visible in a finished work.

3D software is a digital medium and a 3D animation doesn’t (by default) provide a record of the artist’s actions in the same way that a drawing or painting does.

From physical artefacts to symbolic representations

In a 2013 article “Media after Software”, media theorist Lev Manovich wrote:

While earlier reproduction technologies such as woodblock printing, moveable type printing, lithography, and photography stored media in ways accessible to bare senses … the electronic media technologies of the late 19th century abandoned these formats in favour of an electrical signal. Simultaneously, they also introduced a fundamentally new dimension of media – interface (i.e. the ways to represent and control the signal). (Manovich, 2013, abstract).

The shift to digital data and media software 100 years later extends this principle further. With digital computers, data is encoded as sets of numbers. This data is only really accessible via software applications because the software translates the data into sensory representations.

Figure 1.13: Illustration indicating how software modulates digital data to create a screen representation. This illustration uses images from xsens.com and orionstarmedia.com.

With computer graphics software, the way that a user’s actions are mapped in real-time to a visible screen based image is largely defined by human-computer interaction (HCI) researchers, software designers and programmers. This means that the “feel” of the medium, i.e. the “properties" that we experience when we use digital media (including 3D animation software) are largely defined by the software interface and, ultimately, by software and hardware developers.

Unlike a chemical based photograph, a pencil sketch or an oil painting, a digital image or computer animation has no causal or indexical link to the “real”, physical world (Mitchell, 2010, p.43). We can think of a photograph as the trace of light on film, a sketch as the trace of a pencil across a page, and a painting as the result of paint having been physically applied to a canvas. But with a digital image the link between a physical process and a visible outcome is looser, malleable – one could even argue, non-existent. For a 3D user the arbitrary mapping between process and outcome is perhaps best exemplified by the ongoing development of non-photorealistic rendering (NPR) techniques.

Operating in the same manner as stylistic filters found in image editing programs such as Photoshop, NPR rendering algorithms are often designed to achieve the look of traditional art media such as pastel, pencil or paint (Botkin, 2009; Meier, 1996; Shugrina, Betke, & Collomosse, 2006). NPR algorithms allow the 3D user to render the same virtual mesh in a variety of different ways which means that the same production process (i.e. the same physical actions performed by the user) can achieve entirely different visual outcomes. In this way the style of a finished 3D render is the result of aesthetic decisions/choices made by the user, rather than the inevitable outcome of particular physical actions or activities.

The sheer versatility of 3D software (and the fact that it can simulate the look of other media), coupled with an emphasis on deliberate user choices/decisions, encourages us to regard the production process as a means to an end, rather than as a series of intuitive responses to a complex and evolving practice situation.

With this conception of the role of process comes the assumption that it’s the artist’s ideas or explicit intentions that count.

Figure 1.14: This illustration shows the same (Default Whippet) 3D mesh rendered in three different ways; with textured Blinn shader, Toon shader and with Paint Effects charcoal linework. The processes involved in creating each of these images are very similar.

The reductive capacity of digital technology

Makers of ZBrush, one of many 3D animation packages, say that ZBrush allows you (the user) to “create limited only by your imagination” (Pixologic, 2015) and renowned computer animation studio Pixar proudly suggest that they use technology only as a means to an end (Jones, 2014). These sentiments of empowerment are echoed throughout the 3D animation community, from large companies to sole practitioners, but they downplay the agency of the software and promote the idea that computer software is at the service of a proficient user.

As 3D users we are encouraged by software manufacturers to feel empowered, to fully express our “creativity” (c.f. Digital-Tutors, 2014), but this narrative of empowerment masks the complexity of our relationship with digital tools.

When we use a smart phone, tablet, laptop or desktop computer we are almost always using software which has been carefully designed to make us feel like we are in control (Autodesk, n.d.-b; Microsoft, n.d.-a). Collectively, digital hardware and software are marketed as products that empower us to achieve our goals whether they are to stay in touch, to hook up, to learn or to create (Autodesk, n.d.-a; Microsoft, n.d.-b) There’s no doubt that advances in digital technology enable us to do things that we couldn’t do before. However, the computer software and hardware that we use don’t simply open us to new experiences and expand our capabilities; their effects also flow in the other direction, impacting the way that we see the world.

N. Katherine Hayles' theory of technogenesis suggests that humans have coevolved with tools and technologies (Hayles, 2012). Hayles suggests that it’s never been about simply using technology with the users remaining unchanged.

Theories of distributed and extended cognition suggest that human cognition doesn’t just involve the brain or even just the body (Hollan, Hutchins, & Kirsh, 2000). These theories describe how we think with and through technologies and media. Whether typing this document on my laptop computer or sketching Ginger with pencil and paper in the park, it’s difficult to say exactly where my mind or my body ends and the technology begins.

As a 3D animator with experience in drawing, painting and sculpture, I intuitively feel that digital technology has an unparalleled capacity to reduce our experience of the world while purporting to expand it.

This intuition accords with the work of contemporary theorists who point out the reductive capacity of digital tools (Hayles, 2012; Lanier, 2011; Pariser, 2011; Rushkoff, 2010; Turkle, 2011). Like me, these theorists have mixed feelings about digital media; they are excited by its vast potential and also wary of its inherent dangers.

Rather than assuming that they serve as a means to an end, Hayles questions how the ubiquity of digital computers might be changing our practices and changing the way that we think. She suggests that our interaction with networked digital technology has rewired our brains for “hyper attention”, making it more difficult for us to engage in sustained, contemplative attention, or what Hayles refers to as “deep attention” (Hayles, 2007). Similarly, Rushkoff says that while computers copy our intellectual processes, they discourage more complex human thought processes (Rushkoff, 2010, p.23). Whatever the nature of the changes, it seems that we can’t just offload certain types of tasks to machines and remain unchanged ourselves because our interactions with machines (on an individual, societal and historical level) involve certain activities and encourage certain ways of seeing and thinking.

According to Rushkoff, digital technology is replete with simplified models of reality from restaurant recommendations and Google Maps to virtual reality communities such as Second Life (Rushkoff, 2010, pp.61-71). Rushkoff explains that “Because digital simulations are numerical models, many choices about them must have been made in advance. Models are necessarily reductive; they are limited by design. This does not negate their usefulness; it merely qualifies it” (Rushkoff, 2010, p. 71).

The problem is not that abstract and simplified models of the world exist; the problem is that if we don’t question them, they can simplify the way that we see the world. For a 3D animator the problem of simplification manifests in a variety of ways, including our propensity to approach real world things (such as a physical plant or a dog) according to simplified models provided by computer graphics researchers. These simplified models come in the form of 3D animation presets (such as Paint Effects trees illustrated in Figure 1.15) as well as standard 3D workflows (such as the character animation workflow used in Default Whippet and illustrated in Figure 1.6, above).

Figure 1.15: Images showing how to choose a birch tree from a number of presets and how to tweak parameters. These image from online tutorial “Designing and Animating a Birch Tree in Maya using Paint Effects” (Kumar, 2013) Copyright 2016 Envato Pty Ltd.

In response to the claim that Second Life will eventually be indistinguishable from real life, Rushkoff says, “I doubt there is a computer simulation on the horizon capable of accurately representing all the activity in a single cubic centimetre of soil or the entire sensory experience of clipping one toenail” (Rushkoff, 2010, p. 70). While sketching Ginger in the park I could shift my attention from the trees in the background to the shape of Ginger’s leg, or the hairs on her back, and seeing her includes the feeling of stroking her fur. I therefore agree with Rushkoff’s sentiment because an experience as ordinary as this is simply too complex and elastic to describe in mathematical or algorithmic terms. Rushkoff hopes that “By acknowledging the bias of the digital toward a reduction in complexity, we regain the ability to treat its simulations as models occurring in a vacuum rather than accurate depictions of our world” (Rushkoff, 2010, p.62). We can read about biases inherent in digital technologies but how can we experience these biases for ourselves?

Creation replaced by selection from a menu

In his 2001 Book, Understanding New Media, Manovich voices the concerns of many when he says that “in computer culture authentic creation has been replaced by selection from a menu” (Manovich, 2001, p.124).

Software manufacturers such as Autodesk and Pixelogic (Autodesk, 2012; Pixologic, 2015) want us to feel like we can make anything with 3D software but Manovich suggests that using digital media (or “new media”) involves choosing from a limited number of available options and that “New media objects are rarely created completely from scratch; usually they are assembled from ready-made parts ... In the process of creating a new media object, the designer selects from libraries of 3D models and texture maps” (Manovich, 2001, p.124). Manovich is referring to the way that 3D animators often work with a ready-made mesh that has been created by someone else. He is also referring to that way that we use simulation and mesh creation presets which allow us to quickly create predesigned virtual objects (such as trees as in Figure 1.15 above) as well as phenomena such as smoke, fire or ocean waves (as illustrated in Figure 1.16).

Figure 1.16: Screenshots showing how to create ocean waves using Maya's Ocean Shader and how to create smoke using Maya's Fluid Effects.

In Default Whippet I chose to build the dog mesh myself, and I didn’t use simulation or mesh creation presets, but the process still involved using pre-designed digital objects such as shaders, lights etc. This is typical; using 3D software always involves creating pre-designed objects, customising and combining them to suit a project’s specific needs.

Like the drawing recipes that Franck avoids, digital objects can act like a kind of prototype, providing a basis for the way that 3D users describe things in the world.

The progression toward realism

Manovich describes how the development of 3D computer graphics is presented as a progression toward “realism”, and how for the computer graphics community realism means “the ability to simulate any object in such a way that its computer image is indistinguishable from a photograph” (Manovich, 2001, p.168). Comparing 3D graphics with photography and cinema, Manovich makes the point that, in photography a camera records an already existing reality, but in 3D computer graphics we need to construct a virtual reality “’from scratch’ in order to photograph it with a virtual camera” (Manovich, 2001, p.175).

Constructing a photorealistic virtual world is an enormous task and this goal continues to drive a vast array of computer graphics (CG) research (Yuksel, Kaldor, James, & Marschner, 2012). With the ongoing development of new simulation, lighting and rendering algorithms, 3D software is becoming increasingly sophisticated while also becoming easier to use. For computer graphics pioneers, using a computer meant knowing how to program but today there’s no need to understand the code behind our increasingly “realistic” creations. Does the progress of CG research and the corresponding increase in user choice mean that users become increasingly empowered? Or does the ongoing consolidation of “synthetic realism” (Manovich, 2001, p. 168) mean that the software’s implicit assumptions and simplified models are more difficult to discern, more difficult to question and more difficult to disrupt? Heidegger’s concept of technological Enframing suggests that the latter is true.

3D animation and technological Enframing

More than 50 years before the work of contemporary media theorists such as Rushkoff and Pariser, German philosopher Martin Heidegger wrote The Question Concerning Technology (Heidegger, 1977), urging readers to question assumptions and biases implicit in the technologies that we use. Heidegger warned that technology has the capacity to dictate how we see the world and that if we consider it as a means to an end or as neutral then this capacity is hidden from view. He said:

We are delivered over to [technology] in the worst possible way if we regard it as something neutral; for this conception of it, to which today we particularly like to do homage, makes us utterly blind to the essence of technology. (Heidegger, 1977, p.4)

Heidegger explains that the late modern era is haunted by Enframing which is the “essence of technology” (Heidegger, 1977). The word Enframing (Ge-stell in German) suggests a type of framework according to which we interpret everything that exists. For Heidegger, Enframing is a type of compulsion that "sets upon man to order the real as standing-reserve" (Heidegger, 1977, p.19). In other words, Enframing compels us to consider everything (including ourselves) as quantifiable and exchangeable resources which can be endlessly ordered, optimised and stockpiled ready for use. As a 3D software user, the notion of interpreting the world according to an existing framework – reducing things to what is calculable and exchangeable – is particularly relevant.

Standardized workflows, lighting models and simulation tools can all provide frameworks which have the effect of strictly guiding the 3D user’s attention and actions.

3D animation software is a digital medium and a 3D animation is comprised of numerical representations rather than physical artefacts. This means that a 3D work, as a digital file, can be easily duplicated exactly and infinitely. It also means that the work’s constituent parts (i.e. the data or digital objects within a digital file) can be “switched about ever anew” (Heidegger, 1977, p.16).

Figure 1.17: Screenshots from Default Whippet.

In creating a 3D scene (or virtual world) the user tends to duplicate data wherever possible because it keeps the digital file in order and it reduces the animator’s workload, often mitigating the need to do things twice. For example, when creating Default Whippet, I worked on half the dog mesh and the other half was a mirrored duplicate (see Figure 1.17a). If I wanted to create a scene depicting many dogs, each dog would likely be a copy of the first (see Figure 1.17b). As a virtual form, I can share the dog among projects, among users or even sell it on the internet.

Digital animation before and after software

Images in Figure 1.18 illustrate the state of digital animation in 1972. These were the early days of CG research, when interacting with a computer meant programming. The images on the left show a group of Ivan Sutherland’s students at the University of Utah in the process of digitising a VW car. Operating like a manual 3D scanner, the students painted lines over the car and measured the location of resulting coordinates with a yardstick (“Computer History Museum,” n.d.). By entering these coordinates into the computer, they created the mesh shown in images on the right.

Figure 1.18: Images of Sutherland’s students measuring the coordinates of a VW car, and renders of the resulting car mesh. Image sourced from computerhistory.org.

In the process of digitisation, there is a lot about the car and its context that is left out (e.g. the particular shade of green experienced on a sunny day or the car’s significance to its owner, Sutherland’s wife) and this fact is probably obvious to Sutherland, to his students and also to viewers of their work. To our eyes today, this VW mesh is rudimentary; it’s more like a sketch of a car than a convincing representation. But in the 40 years since this digitisation project, there have been many advances in computer graphics software (and in networked digital computing more generally).

Today the processes used to create a mesh are more intuitive and the outcomes are more convincing. Rather than entering Cartesian coordinates into the computer via the keyboard, contemporary 3D users can “sculpt” a car of up to a billion polygons using a process “that feels incredibly natural” (Pixologic, 2015). With the aid of various lighting and shading tools, a contemporary digital representation of a car can (more) easily approach photorealism. Today we can create photorealistic CG cars. In fact, contemporary advertising often prefers to render virtual cars than to photograph real cars (Southern, 2012). Journalist Alex Southern’s online article from 2012, “Real or Rendered? How 3D Imagery Is Changing the Way You Shop” (Southern, 2012) describes the growing trend toward the use of computer graphics within the advertising industry. Southern states that in 2012, 80% of automotive advertising was created digitally (Southern, 2012).

Figure 1.19: A car advertisement which uses a computer generated car (Southern, 2012).

Car manufacturers were one of the first to use digitally created products instead of photographing the real thing, but this trend is not limited to the automotive industry. A 2012 article in the Wall Street Journal predicted that in 2013 furniture manufacturer IKEA would be creating 25% of its content for catalogues and brochures using 3D software (Hansegard & Jens, 2012).

For companies such as IKEA, computer graphics provides a way to save money and also to achieve visually flawless images (Hansegard & Jens, 2012; Southern, 2012). As a contemporary 3D user, if there is anything about cars that my virtual models leave out then this omission (i.e. the sense that there is an omission) is no longer obvious to me or to viewers of my work.

Figure 1.20: This illustration shows a 3D render (i.e. a computer generated image) of an IKEA chair, a more complex 3D scene, and an IKEA 3D artist at work. Images sourced from cgsociety.org.

As described by Heidegger scholar Graham Harman, Enframing “drags … things into a colossal gridwork and reduces them to calculable and manipulable surfaces …Things now have meaning only insofar as they are subjected to this universal grid of presence” (G. Harman, 2009, p.6). Harman’s use of the words "surface" and "presence" allude to the dogmatic and totalising nature of Enframing, i.e. the way that it obscures other understandings and fails to acknowledge the existence of meanings that don’t come from us.

Harman’s description of the totalising nature of Enframing doesn’t seem entirely applicable to activities undertaken by Sutherland and his students or to their “boxy” digital car because of the novelty (i.e. the “non-naturalness”) of this 3D scanning process and because of the obvious inadequacy of the outcome (i.e. the “sketchy” car model). But

Harman’s description of Enframing does seem applicable to practices and outcomes associated with contemporary 3D software because, as 3D users, we feel like we can make anything and our creations can be very convincing.

Writing more than 60 years ago, the technologies that Heidegger referred to in his essay include nuclear power plants, mechanised farming and hydroelectric dams. He contrasted these modern technologies with simpler technologies such as an old wooden bridge (Heidegger, 1977, p. 16). Subjected to the demands of the hydroelectric dam, a river becomes a source of power and “appears as something at our command” (Heidegger, 1977, p. 16). Like the hydroelectric dam, a wooden bridge is also an intervention into the landscape but, according to Heidegger, it doesn’t reduce the river in the same way. Unlike the dam, the bridge allows the river that it spans to flow unimpeded and the bridge also highlights some of the river’s intrinsic qualities, bringing them to our attention.

In an essay from 1951, “Building, Dwelling Thinking”, Heidegger describes how a bridge:

does not just connect banks that are already there. The banks emerge as banks only as the bridge crosses the stream. The bridge designedly causes them to lie across from each other. One side is set off against the other by the bridge. (Heidegger, 1971, p. 150)

The problem with modern technology, as Heidegger sees it, is its capacity to reduce our experience of the world in a way that prior technologies never did. A bridge allows the river’s intrinsic qualities to shine forth, bringing them to our attention, while a hydroelectric dam reduces the river to watts.

3D computer graphics are now a ubiquitous part of the networked digital world (Guha, 2015, p. 3). For animators, 3D software has the capacity to reduce our experience of the world in a way that prior creative media never did. Along with our growing dependence on 3D graphics, reasons for this include the complexity and versatility of 3D software and the fact that its imagery can be convincingly “real” (Konnikova, 2015; Rushkoff, 2010, p. 69). For 3D users, it’s easy to assume that the secrets of the visual world are at our fingertips and, that with enough knowledge and dedication, we can make images as convincing as real world things. By contrast, an activity such as sketching (as outlined above) can open onto a world of inexhaustible complexity - using a pencil to both explore and describe things simply highlights the impossibility of an exhaustive description. Simple technologies such as paint and brushes, or pencil and paper can call attention to the inexhaustible complexity of real things and, in this way, they are like Heidegger’s wooden bridge. 3D software, on the other hand, is more like a hydroelectric dam, reducing real things to data.