There have always been two sides to the animation medium: On the one hand there’s the opportunity to exaggerate reality; to distort and caricature with the goal to create unseen worlds with their very own wicked logic. On the other hand side is the striving to come as close to reality as possible to create complex illusions or just to test the boundaries of what can be done. While appeal is a challenge in any case, there is a big pitfall on the path to realism: The Uncanny Valley.
What is it?
Generally, the appeal of a character rises the more human it looks, but there is a point shortly before complete realism, where the design becomes extremely repelling. This plunge in the appeal curve is called the Uncanny Valley. Motion, which usually helps to make something more life-like and appealing emphasizes the negative effect.
Why does it happen?
There are a couple of different theories why we react so negatively to almost-but-not-quite-real designs. A lot of them have to do with us reacting strongly (namely with fear) to abnormalities, because in a real human those little differences would indicate illness and anti-social strangeness. Consequently, it reveals that this robot or computer generated human is not really alive and destroys the illusion. And here is one of my favorite theories why motion makes it worse: Once our brain has detected that the Uncanny Valley character is not a living human, it sees it as a corpse. And a moving human corpse is super-creepy. This may be why zombies are so oddly fascinating.
Interesting side note: A research team from Princeton university found that out of a picture of a caricature, a photograph, and an Uncanny Valley monkey, our animal siblings disliked the Uncanny Valley version (you can find the article here). This indicates that it’s not a mechanism put in place by social norms, but that it indeed is an instinct.
But the human mind not only has the ability to detect if something is off with a super realistic human design. On the other side we have the ability to feel empathy for things that aren’t even remotely human-like.
Technology conferences offer an excellent example for both of these abilities. The newest humanoid robots are always fascinatingly creepy – especially the once with a fake human skin. But, have you ever seen one of these vacuum cleaner robots suddenly realizing that it runs out of battery? When it rushes to the charging station and gets stuck on a carpet fold causing it to spin the wheels like crazy you feel amused and almost sorry for that little guy. This is human empathy in full action: We interpret the rushing and the struggle to overcome the obstacle as panic – as if the little robot was a living creature.
This means in conclusion: We appreciate traces of human behavior in things that we don’t find to be a realistic being, but we are repelled by traces of strangeness in things that we are supposed to see as realistic beings.
How to avoid it?
How can you get around the Uncanny Valley? The answer is simple and logical: Stay away from very realistic designs! In most cases your film can benefit from a design that moves away from realism and into something that emphasizes the tone, theme or some other aspect of your film. And if you want to have realistic humans in your film, why wouldn’t you shoot them in live action? Keep in mind that the Uncanny Valley also applies to just borrowing realistic elements (a realistic eye on a cartoon character can be really, really strange) and motion (this is why motion capturing still has a long road ahead).
Now you might say “what about VFX films with humanoid characters?” Well, if you’re really sure that you cannot film real people with VFX elements on them, you should still try to step away from exactly copying a human. One great way to do this is to slightly change the proportions away from human anatomy. Take the Avatar aliens for example or Gollum from Lord of the Rings. In these cases realistic textures, and elaborated bone and muscle systems make them belong in a realistic environment – but the slight changes in anatomy (bigger eyes, longer limbs) kept them out of the Uncanny Valley. This way the control of appeal is back in the hands of designers. On top of that the bigger eyes are an excellent stage for the emotional performance.
Another way to cope with the Uncanny Valley is to consciously use it as an artistic device. Take Chris Landreth’s work for example: The Uncanny Valley designs are part of his style and perfectly fit the atmosphere of his surreal films.
An example where the Uncanny Valley is even part of the theme is Quantic Dream’s tech-demo Kara. Rather than ignoring the Uncanny Valley, the whole film addresses it and carefully progresses from there. But see for yourself:
Now, you might know that Quantic Dreams makes games like Heavy Rain and Beyond Two Souls, neither of which have overcome the Uncanny Valley (nor want you to be aware of it). I think they only work because we are used to motion-capturing stiffness in games. The style would be critically destroyed if it were a feature film. Plus the quality of the actors and the script is just so good it brings you to overlook the lack of realism. But this again only proves that you don’t need the characters to look realistic in the first place to make your audience be emotionally invested.
Of course another way to avoid the Uncanny Valley would be to actually overcome it. But even if we were able to do that – portraying every little detail of our world will always require an insane amount of work and computer power. So why not use our mind’s ability to see life in not so realistic designs? In the end it’s emotional believability and authenticity that will make a story work.
What do you think? Are you interested in movie and game technology exactly copying the real world? What are your favorite examples that work despite or even because of the Uncanny Valley? Please join the discussion in the comments below, we’d love to hear your thoughts!