It always surprises me when I come across a die hard gamer who doesn’t know what the Uncanny Valley is. They could be forgiven for thinking that it’s a misty ravine described in a science-fantasy novel, filled with 4 headed monsters and dragons, creeping through thick, black swamps and reeds. It’s not, but the reality is equally as macabre. So just what is the Uncanny Valley?

The Uncanny Valley is a theory by roboticist Masahiro Mori which he developed in 1970. He came up with a curve that compared the realism of a humanoid robot (or CG videogame character) to a healthy human being. The Uncanny Valley is the point at which the curve dips as the artificial humanoid seems almost human, but subtle nuances cause our mind to clearly identify it as false. This contradictive curve, compounded even further when the being moves, is as relevant as ever with the evolution of our beloved videogame technology.

With the dawn of a new console, or with every new video card release, the unit of measurement used to justify paying launch prices for new hardware has always been ‘how real the games look!’ As time goes on, and even more technologically advanced systems rack up debt on our Mastercards, we realize how ridiculous we were for crying the virtues and life-like realism of Perfect Dark on the Nintendo 64. Until we see the next generation, and lie to ourselves all over again. The current generation is no exception, so when will we really get to see life-like characters that don’t look like plastic toys that have been photographed? If Masahiro Mori’s hypothesis is to be believed, will we ever?

Recently I stumbled across a demonstration video that floored me. It was a press release movie from Image Metrics. The woman talking in the clip was describing a new animation technique that would revolutionise the characters in our games. I kept waiting for them to show a sample of what this chick was squaking on about, but just as my bored, heavy eyes were beginning to close and my head descended toward my desk, it became apparent that the woman on screen was in fact computer generated. Instantly my eyes burst open and I quickly emailed the video to everyone. But as I watched the video over and over again in demonstration to those now gathered around my desk, the easier I began to pick the faults. Before long, I was the one pointing out the disappointingly obvious flaws in the animation and textures. My mind simply would not let me believe that this animation was a filmed actress.

Realistic still images and sketches have been around for centuries. There are renaissance paintings that you would swear are real photographs. The issue, it would seem, is the animation. What is it about their characters that the animators can’t get right? Eyes always look glazed over, mouths and teeth seem to move in slow motion and never quite in sync with the words they’re articulating. Anyone who’s studied anatomy would tell us there are an incredible amount of muscles and millions of nerves that go into making you pull stupid faces while trying to do the Mile High Achievement on Call of Duty 4. So it’s understandable that they can’t pull-off what nature has achieved, right? Well in my opinion, yes…and no.

So far, they’ve come pretty damn close. Films like Beowulf and The Polar Express, both rendered entirely in CG, are created to look as real as possible. And while each character’s rubbery motor-skills makes it glaringly obvious that you’re not watching stunt actors in a sound studio, the techniques are still very impressive. So it’s just a matter of time, right? Surely over the course of the next few years, animators will have more advanced techniques and years of experience under their belt that will enable them to animate a character so it doesn’t look like latex pulled over two fists.

French studio Quantic Dream, developer’s of the upcoming action adventure title, Heavy Rain, claim that “there is no uncanny valley any more, not in real-time”. While I think they’re kidding themselves, and it sounds like they’ve been getting PR tips from Peter Molyneux, they’ve come the closest of any game to date. The interesting thing about their claim is that they qualify it with the words ‘real-time’. Videogame animation has always been separated into two distinct categories, pre-rendered and real-time. Real-time animation and CG characters rendered on-the-fly have always been a couple of steps behind pre-rendered animation. It’s a sacrifice I’m sure we’re all willing to make for the sake of allowing our customized characters to interact with the game world in real-time. Gone are the days where  the Master Chief would go into a cut-scene holding a Plasma Pistol only to be holding his trusted Battle Rifle later on. It would have been more noticeable if we weren’t all picturing ourselves as Cortana’s silent cartographer, but the distraction just isn’t there with guys like Marcus and Dom.

Are we chasing an unattainable dream? The closer the animators come to controlling and manipulating human motion and emotion, the less we have to use our imagination and the more cynical we become. Those of us smart enough to still recognize the value in reading a book can tell us that the human imagination is a wonderful and powerful tool. Never have I called out the realism of a chapter, claiming that it wasn’t real enough. My brain creates a picture in my head, and it’s far more real than anything my eyes can interpret. The same can be said for motion pictures. We assume that the characters on screen are real, so our minds just put that idea aside and focus on other things.

Over the past few years, the techniques used to create realistic in-game characters has taken priority, and gameplay mechanics and enjoyment seem to have taken a back-seat. I can’t help but feel that this period in videogame history will be looked back upon with a sense of regret. We were the pioneers, trekking the  Uncanny Valley needing to be crossed between the cartoon-looking characters that our minds accepted, and the film-realistic 3d models of our future. Or maybe nothing will change, and our minds will slowly learn to start filling in the blanks again.