Responding to the Call of Things
A Conversational Approach to 3D Animation Software
5. Outcomes and findings
Through an iterative process of enquiry, this research has developed approaches to 3D animation software which disrupt habits and conventions of practice by fostering a comportment of active receptivity.
Outcomes from this research include a number of short 3D animations as well as seven creative strategies (Playing with Software, Playing with History, Animation as Relation, Working from Sketches, Working from Life, Colour as Light and Geometry as Shadow ) and five custom tools ( Auto Expression, Auto Keyframe, Modelling as Animation, Shadow Play and Plein air Still life). Appendix C of this document presents research outcomes and findings as a list.
A conversational approach
Surprises, glitches and unexpected outcomes inevitably occur during any 3D animation project, but these are usually encountered by the user as accidents or mistakes to be avoided. I was intrigued by surprising images that occurred while making Default Whippet but I found it hard to incorporate them into my work. In part this is because the modular nature of a many standard 3D workflows means that improvisation is possible only within limits.
My comportment also tended to prevent me from being receptive to the unexpected; because, despite appreciating surprises when they occurred, with Default Whippet my main focus remained on implementing correct procedures and achieving repeatable and predictable results.
The experimental animations, by contrast, were less concerned with achieving specific outcomes and more concerned with achieving specific qualities of practice.
With those animations, rather focusing on my ability to control a situation (which felt like a way of reducing it), I wanted to feel that I was responding appropriately to a situation that is too complex to control and too complex to entirely comprehend or understand.
By shifting the emphasis from outcome to process and from control to response I was able to foster a comportment of active receptivity. This is Heidegger’s bringing-forth, and it can be contrasted with the challenging-forth associated with technological Enframing.
I refer to my approach in these works as “conversational” because it involves listening carefully to the call of things and responding in a way that feels right or appropriate.
All of the experimental projects presented in this research involve responding to the medium (i.e. responding to suggestions from the software) and some of the projects also involve responding to physical things (e.g. dogs, books, people). Whether responding primarily to the software or to physical things, a conversational approach also involves responding to the work as it unfolds. This means responding to emergent content, and making changes that seem to help the work become what it wants to be.
At a certain moment, or moments, during production, a work exhibits its own authority/identity. At these moments judgements of the work are not (entirely) based on its resemblance to reference images or to physical things. I have found that it’s easier to notice these moments when avoiding internet images and Working from Sketches or Working from Life. It also helps to appreciate the viewport display, and not just see it as an indication of a final rendered outcome.
The thingliness of digital objects
A conversational approach to 3D animation attunes the user to the agency of the software and reveals that this agency extends beyond ideas implicit in the software’s design. A conversational approach recognises that 3D software is capable of more and different than it was designed for and that it always has the capacity to surprise.
Lanier (2011, p. 133) suggests that digital objects (including computer software and its constituent parts) are always designed for a specific purpose. This may be the case, but approaching these objects with an actively receptive comportment means encountering them as digital things, in the Heideggerian sense. As a general strategy, Playing with Software helps us encounter digital objects as things because it shifts attention away from what a tool has been designed to do and encourages us to focus on what a tool actually does. We miss qualities inherent in 3D software if we encounter digital objects only as they were designed to be encountered. Whereas, on the other hand:
Playing with Software reveals digital objects as things, i.e. conceptually inexhaustible with the capacity to continually surprise.
We can be attentively present in the domain of things, digital and otherwise, by disrupting our own expectations and our own habitual practices. In this research, I have Played with History to elicit suggested forms, and I explored Animation as Relation to elicit suggested movement. Rather than seeking an efficient production process, I have continually invited surprising outcomes in order to disrupt my own working habits and also to disrupt my own easy interpretation of a work.
Unlike Pixar’s unambiguous imagery, designed to communicate a clear storyline, I have pursued image, form and movement that feels significant or meaningful, but is not so easy to describe in words. Rather than seeking media transparency, or using technology as a means to an end I found that the fact of the digital medium can inform animation content and be (more) apparent to a viewer of a finished work.
It is implied that being “limited only by your imagination” (Pixologic, 2015) is a good thing, but:
I have found that, by exploring the vitality of the accident, animations and images emerge that are sometimes more interesting, intriguing or poignant than I could ever design, imagine or dream up in advance.
Beginning in poverty and seeing essences
According to David O’Reilley “The problem [with 3D software] is that there is simply too much power and very little control, essentially you get too much for free” (O’Reilley, 2010, p. 1). All tools have their associated conventions but I share O’Reilley’s concern that 3D users get too much for free. Cézanne had to learn how to implement pictorial perspective, it didn’t come embedded in his paint brush, and we can assume that his studies of traditional techniques helped him recognise perspective as a convention. Ultimately Cézanne was able to play with this convention and develop his own, looser version of perspective; the version that Merleau-Ponty calls “lived perspective” (Merleau-Ponty, 1993a, p. 64).
For early computer graphics researchers such as Newell, Sutherland and Phong, finding ways to digitally describe objects meant looking at things in the world around them in a novel way. Although their approach to these things (e.g. a teapot or a car) could be described as “objective”, because they attempted to disregard context and meaning, these researchers examined their own perceptual experience and, in abstracting simple models from complex experience, they were likely aware of the limits of their own models. But for those of us who use tools based on their algorithms, 40 years after they were designed, it is easy to forget the limitations of these models. For contemporary 3D users who don’t need any knowledge of programming, there is a danger that we may mistake algorithmic models for an explanation of the way the world is, or as the definitive way of accurately describing it. As 3D software becomes more sophisticated and easier to use, this danger grows rather than diminishes.
This research finds that the acquisition of programming skills and technical knowledge is liberating for a 3D user in a variety of ways. Several programming languages have been used throughout this research, but most programming has been done using Python. In learning Python, an object oriented language, I have begun to appreciate coding as a practice of defining essences. With an understanding of some of the practices involved in the creation of 3D software, I can now appreciate that seeing (and defining) essences is different from working with essences defined by someone else. As well as becoming familiar with practices involved in creating 3D software, programming skills and technical knowledge can reveal conventions and assumptions inherent in standard uses of 3D software; they can foster an appreciation of these conventions, as well as a healthy disregard for them. An example of this is the way that acquiring technical knowledge helped me to escape sculptural metaphors which are common in 3D animation packages.
Designed to be intuitive and to hide its algorithmic workings from a user, ZBrush (Figure 5.1) is an excellent example of software built on a sculptural metaphor. A ZBrush user is not encouraged to know anything about the technical specifications of a virtual mesh; they are instead encouraged to approach 3D modelling as though they were working with clay (Pixologic, 2015). Many users find this liberating and associate this non-technical approach to computer software with artistic expression and creativity. In contrast to these users, I have found that liberation is achieved by gaining awareness of what goes on behind the user interface.
A 3D mesh exists as a dataset which specifies certain points in virtual space, and the polygons that connect them. In my research I gained an appreciation of this fact through activities such as creating a polygon mesh by writing code instead of moving icons in the viewport. An example of such an activity is shown in Figure 5.2. Although a mesh dataset includes a virtual third dimension, unless it is 3D-printed it will be used to produce a 2D image, not a 3D object. This is true even for stereoscopic images, which use a slightly different 2D image for each eye.
Sculptural metaphors imply that a mesh should be stable and coherent or even that it should look like it could stand up to gravitational forces. But 3D modelling is not sculpture and there is no reason for a mesh to look like a completed form; there is no need for it to be a stable throughout the timeline or even to “make sense”. Sometimes a polygon dataset is ambiguous, the software gets confused, resulting in glitches. Letting go of sculptural metaphors, we can see that confusing the software is not necessarily a bad thing.
There is a lot of knowledge embedded in computer software, but it is fruitful for a 3D animator to try and begin in poverty. One way of achieving this is to take a back-to-basics approach to the software, which means avoiding modelling presets as well as complex lighting and shading algorithms. To this end my research has used Colour as Light and Geometry as Shadow - strategies sometimes enabled by custom tools such as Plein air Still life and Shadow Play. These strategies are a way of minimising the number of decisions (about the essential attributes of a thing) that have been made on my behalf and they allow me to explore the accidental surface appearance of things. Like Monet, these strategies allow me to see everyday things as abstract areas of colour and to (almost) forget what it is that I’m looking at. By bringing the viewport display closer (in appearance) to a final render, these strategies also encourage me to appreciate the screen image, i.e. to pay attention to it and to not assume it is merely indicative of a final rendered image. They encourage me to notice when small local changes alter the image as a whole, and to be attentively present at the moment when pixels on a screen become a recognisable object.
The world of perception
Continued advances toward the achievement of photorealism in 3D computer graphics suggest that the way something appears to a camera can be convincingly simulated using computer algorithms, but if we examine our own perceptual experience of things we find that it is too ambiguous and elastic to be captured by formal mathematical models, no matter how complex. Although (or perhaps because) it is “going against the grain”, an examination of perceptual experience is a potentially rich site of enquiry for 3D animators.
For CG researchers aiming to simulate reality, things exist in the world “out there”, separate from our experience of them. Teapots, cars, billiard balls, scarves, trees, snow, even skin: CG seeks general models for convincingly (and objectively) describing these things and uses photography as its benchmark. But when we examine our own experience of things, we find that things are never entirely separate from who and what we are.
One obvious reason for this is that what constitutes a thing (as experienced) depends on the types of distinctions that we, as biological entities, are able to make. As philosopher and biologist Francisco Varela explains, “there are many ways the world is – indeed even many different worlds of experience – depending on the structure of the being involved and the kinds of distinctions it is able to make” (Varela, Thompson & Rosch, 1993, p. 9).
In accordance with this insight, I have noticed that, although they overlap, Ginger’s world is different from my own. Watching her sniff the air, it’s obvious that her world is populated by things that don’t exist in my world, and the reverse is also true. As well as being based on my biology, the things that populate my world are the result of my cultural background and personal history; they are “a result of a mixture of sensory experiences, emotional responses, memories, prejudices and the like” (Willis, 2001, p. 1). As individuals, cultural groups, or whole species, we can't simply project our own meanings onto things (we can’t just decide to see all dogs as unicorns for example). But equally, we are not simply material entities which respond to things in a mechanical, causal fashion. We exist between these two extremes. Just as making can be conceived as a type of conversation, perception also has a conversational structure: i.e. perceptual experience can be conceived as a conversation between ourselves and other things.
Our relation to things is not simple – it is entangled from the start. But standard approaches to 3D animation make it easy to overlook this fact. Along with its emphasis on human agency (promoting the idea that, for the human user, 3D software is a means to an end), obscuring the role of the body as well as personal experience in perception are ways that 3D promotes a simplistic model of our relations with things.
As well as providing the impetus to disrupt standard practices, returning to the world of perception helps us to notice the complexity of our relation with things, including the mundane things that constitute our local environment.
Mapping process to outcome
When working with paint, pencil or charcoal, the artist’s intuitive movements leave marks which sometimes remain visible in a finished work as multiple overlapping brushstrokes, graphite lines or smudges of charcoal. In a digital medium, there is no longer an indexical link between the artist’s actions and the traces that they leave. For the viewer of a 3D image or animation, unconscious or intuitive decisions, movements and marks made by a user are often difficult discern. This research has developed several strategies and custom tools that capture something of a user’s intuitive actions. The most successful of these are Modelling as Animation and Plein air-Still life, which record iterations of a model while I work.
Using strategies and tools developed in my research, a finished 3D animation can embody something of a user’s intuitive decisions made throughout production.