Into the Uncanny Valley…

Scheduled to be released in the UK on 8th April, the quite frankly dreadfully-named Mars Needs Moms is dying on its animated feet in the US. Currently looking destined for ultra-flop immortality with the likes of The Adventures of Pluto Nash, many marketing and executive types must have been scratching their heads after it made only $6.9 million in its opening weekend. Four weeks later and the scratch must be turning to an unpleasant and nervously sweaty clawing as, despite its $200 million production and marketing budget, Disney’s new enfant terrible is sitting humiliated and embarrassed upon a paltry $34 million in worldwide gross. Where did it all go so wrong? Is it because all of the human characters look so creepy?

For features like The Lord of the RingsKing Kong and, most recently, Avatar, motion capture served its purpose rather well. For the former two, using the CGI whizzes at the Weta workshop to turn Andy Serkis’ gurn-intensive, knuckle-scraping performances from just a weird guy lumbering around a room into something really quite impressive and groundbreaking. But what all of these successes have in common is that the technology wasn’t used to animate humans, and this is where Mars Needs Moms’ problems start.

Robert Zemeckis’ previous efforts in the CGI mo-cap genre include The Polar Express (2004), Beowulf (2007) and A Christmas Carol (2009). With each he attempted to make his human characters look as realistic as possible and in each case (with ever decreasing box office takings) seems to have fallen headlong into the Uncanny Valley. What is this valley of which you speak, one may ask? Well the Uncanny Valley was a term first coined in 1970 by the roboticist Masahiro Mori to explain the eeriness that many people experience when in the presence of very realistic humanoid robots. The hypothesis more precisely states according to Wikipedia (that well known bastion of infallibly accurate information) that:

“As a robot is made more humanlike in its appearance and motion, the emotional response from a human being to the robot will become increasingly positive and empathic, until a point is reached beyond which the response quickly becomes that of strong revulsion. However, as the appearance and motion continue to become less distinguishable from a human being, the emotional response becomes positive once more and approaches human-to-human empathy levels.”

The valley is not exclusively for robots however. Ever since Pixar’s 1988 short Tin Toy featured a baby that can only be described as horrendously horrendous looking (see horrendously horrendous looking baby below), the hypothesis has gained wide acceptance in the world of CGI animation. As depicted in the graph above, the Uncanny Valley is a cold, dark place littered with costly mistakes ranging from aforementioned creepy humanoid robots, the virtual cast of the game Heavy Rain and now apparently the freaky, dead-eyed residents of Earth as envisioned in Mars Needs Moms.

There are many theories concerning why the valley comes about but the one that seems to hold the most water, at least for me, is that we view these imitations of humanity as being deformed in some way, and it doesn’t sit well with us. Our subconscious brain seems to consider them somewhat human and then sees every flaw as detracting from this and can even render us (evolutionarily speaking) somewhat solipsistically suspicious of them. The flaws are subsequently oddly repulsive and can neuter our emotional connections to them. For example, take two animated films released in the same year: The Incredibles and The Polar Express. The first has a stylised, cartoony art style and the second (as mentioned before) a look that aspires to photo realism. The family in The Incredibles may have looked distinctly inhuman with their odd proportions and obviously computer-rendered flat skin tones and clothing, but the characters themselves were emotive and expressive – when they represent human characteristics as well as they do, the mind often seeks to imbue them with human traits, even humanity itself. This can be extended to robots, animated robots (Wall-E) and (for the more twee and soppy amongst us, from which group I don’t feel I have sufficient grounds to exclude myself) even real-life things like teddy bears and puppies.

One common criticism of The Polar Express was that its characters’ eyes seemed somewhat devoid of emotion and occasionally their looks of sadness or wonder came off as oddly threatening. Initially, this seems to account for the uncanny factor but the trailer for Mars Needs Moms suggests a massive technological leap for eyes, all sparkly and emotive. It’s the skin textures, the tiny little movements, the mannerisms, the walking and the mouths (which seem to be the main challenge). Oh, the mouths – too flappy or too rigid or too toothy or sometimes all at once. They’re weird and people don’t seem to like them (just look at Jeff Bridges’ younger, pixellier self in Tron: Legacy). There is so much to think about, even without the technological constraints, that it may be years until we rid ourselves of the uncanny creeps inhabiting our less-inventive children’s animated features.

That said, comparing The Incredibles to The Polar Express seems a somewhat futile practice in this instance as visual differences are the least of the latter’s worries. The Incredibles succeeded both critically and financially because of a multitude of factors. These factors include its outstanding pedigree, writing and production values. The Polar Express failed on an artistic level to be anything more than average because its script was lacklustre, its art design way too obvious (although it did get the job done) and overall it just displayed a dearth of that something special that Disney/Pixar seem to bring to all of their films.

Zemeckis may protest that he is merely throwing himself at the ever-weakening wall of progress as he sacrifices his studio’s finances in pursuit of his possibly misguided dream of removing actors from the screen altogether. But now is not the time to be looking for photo realism in our animated characters, because it doesn’t seem to work. Will it ever be? Live action is live action and animation is animation. Waltz with Bashir’s soldiers were no less pitiful and sympathetic for their oddly creative art style and its atrocities no less dreadful. Toy Story wouldn’t have been improved in terms of human drama by depicting actual humans, photo realistic or not. There is no need to depict photo realistic humans when we have them all around us anyway. I’m a photo realistic human and it took minimal technological effort on my part to achieve. Just train a camera on me or any of the 6 billion humans around us and you have got one. Spend a few million pounds and you can stick me on Mars or any planet you care to, cram in some aliens and you have something approaching Mars Needs Moms without the rubbery faced sinister undertones that CGI ‘realism’ seems to bring.

Animation can be so much more creative than live action so why there is so much effort going into homogenising the two methods of film making? Finding Nemo’s visuals were groundbreakingly beautiful, Cloudy with a Chance of Meatballs went a bit mad but in a fantastically Lynchian way and things like A Scanner Darkly (pictured above) were technologically inventive – using live action as a starting basis rather than the goal for their technique of interpolated rotoscoping – without losing sight of the aesthetic value of the final product. Mars Needs Moms isn’t flopping all over the Box Office floor like a fish out of water because of the Uncanny Valley. The film is generally misguided and misjudged and this extends to its aspirations to photo realism. By all means, use photo realistic CGI in live action but that’s its place. Animators should be getting all creative on us, going a bit mad and more importantly working from better scripts. Graphic fidelity is not everything – it’s what you do with it that counts and if this means going a little bit off-piste then I’m all for it.