Motion Capture: Classical Animation's Uncanny Double

Reading through a mailing list for character animators that I subscribe
to, I am struck by the disdain and revulsion that so many animators
clearly feel for the technique known as motion capture, even jokingly
referring to it as "Satan's Rotoscope" or "Mo-Crap."

Motion Capture is a new technology that radically changes the means by
which animation may be created. By attaching sensors or tracking points
to key parts of the body, an actor's physical performance may be
translated into the digital realm, where the "captured" motions may be
applied to almost any character model. This technology has been widely
used in videogames, and less widely in commercials such as the Shell
Dancing Oil Pumps and television shows such as "Donkey Kong Country,"
"The Real Adventures of Johnny Quest," and the recent
"steve.oedekerk.com."

What is it about motion capture that traditional animators so dislike?
It is lifeless, they insist, and the results are inferior to traditional
keyframed animation. While there are definite economic and artistic
concerns regarding the value of motion capture, I sense that there is
also a more basic reaction at work. Personally, I have often found
motion captured animation to be disturbingly uncanny in its reproduction
of the captured subject's movements, particularly as the reproduction
increases in faithfulness. And I suspect that this uncanniness might be
close to the experience of other animators, and a possible source of
their dislike for motion capture.

The uncanny is a useful tool for thinking about this disturbance, but
what is meant by it? Freud defines the uncanny as "that class of the
terrifying which leads back to something long known to us, once very
familiar," a specific instance of the return of the repressed. Often it
arises where the boundary between the living and nonliving is blurred,
where the subject "doubts whether an apparently animate being is really
alive; or conversely, whether a lifeless object might not be in fact
animate." Why is this disturbing? Because we have repressed the idea of
our own mortality, but the uncanny haunts us with the spectre of death.
The appearance of life in the nonliving threatens us with the
possibility of the intrusion of death into life. But of course this
emotional, psychic threat is not "real;" it is quickly repressed by the
act of intellectual examination.

Many animators contest the idea that motion capture can even be called
animation. While the process itself is more akin to the performance of
puppetry than the step-by-step creation of key poses of traditional
animation, the end result must certainly be classified as animation. To
animate is "to give life to; to give motion to" (Websters), and motion
capture is a process that invests immobile figures with motion and (I
would assert) life. This grafting of life onto the inanimate object
always carries the potential for the uncanny moment. Every animated
figure is always already undead, carrying with it this threat of
disturbing the boundary between living and nonliving, but this threat is
well repressed in traditional animation by the codes and conventions
that contain it.

Animation is codified in two major ways: a stylization of imagery, even
to the point of iconification; and a stylization of motion, particularly
a certain kind of exaggeration and violence of action that places it
beyond the realm of the real. This is typically described as the quality
that makes animation "alive" but it is really a hyper-aliveness, an
intentional exaggeration that marks the animation as animation, as
cartoon. Thus the presence of the author is made evident, not only in
the fact that the images are clearly drawn, or created, by some outside
source, but also that their movement is impelled by that same author.

While several different persons may contribute to an animated piece,
there is nonetheless a distinct "authorship" at work. This authorial
voice makes itself heard in the way the characters are drawn, the way
they move, and the way the events of the story are structured. The films
of Disney, Warner Brothers cartoons, even Saturday morning tripe all
have a very clear authorial voice. Animation does not need to be
exaggerated to be stylized; take the example of Disney's Snow White, who
is animated as a "realistic" character, in contrast to the dwarves. Her
motions are nonetheless stylized and bespeak an authorial intent. This
is not to suggest that traditional animation cannot evoke uncanny
moments in us. It certainly can happen, as that possibility is inherent
in all animation.

[…]

Perhaps, once we have stared at it long enough, once codes and
conventions have solidified around it, motion captured animation will
not as easily evoke the uncanny in us. The more post-processing and
smoothing that is performed on the data, the less it resembles its
source and the more stylized it becomes. Perhaps this is means by which
we will attempt to master and repress the life in death that the motion
captured figure represents.

Read more about motion capture at these sites…

http://www.geocities.com/Hollywood/1234/mcap.html
Download an mpeg of a mo-cap animated "sexy robot."

http://www.ascension-tech.com/maximize.htm
A white paper on the economic benefits of motion capture technology.
Thorough discussion of various technolgies used.

http://www.cgw.com/cgw/Archives/Magic/11/11story1.html
Discusses the relative mertis of motion capture vs. keyframe animation
for the production of animated TV series.

http://www.cinenet.net/users/rickmay/CGCHAR/cg-charfaq.txt
The CG-Char FAQ has a glossary of terms, including entries on motion
capture and rotoscoping.

http://www.cgw.com/cgw/Archives/1996/10/10story1.html
Article on Medialab, one of the pioneers in bringing motion capture
technology into television production.

http://www.cnn.com/SHOWBIZ/9706/17/motion.capture/index.html
A typical mass-media article heralding "New Hollywood technology -
almost instant animation!" Psychoanalysis aside, this is why animators
really hate motion capture.