Ever heard of Beowulf? The recent movie, not the story. How about Avatar? You probably saw that one. Most moviegoers did.

There are two big movie trends right now that seem impossible to escape. One is so-called “realistic CGI,” in which the animators try to achieve a photorealistic look with purely synthetic images. (Photorealistic images have already been achieved in movies by pointing real cameras at real things, but let’s not digress too far into that.) The other big trend is 3D, which requires the viewers to wear large plastic goggles. Hollywood asked you if you wanted these things in your movies by making them, and you responded by opening your wallet.

The royal “you,” of course. I’m not pointing fingers.

There have been some oddly specific objections to these movies, more than just the backlash you’d expect from a burgeoning trend. The complaints range from mild annoyance to actual physical pain. People think that the CGI characters look weird, waxy, and lifeless. Some have complained about getting headaches while watching the 3D presentations.

The Uncanny Valley

It turns out that human biology itself might be resisting the CGI graphics. The recent buzz term for the unshakeable “offness” of realistic computer images is “uncanny valley.” Our brains are fully prepared to accept images of real actors, photographed in the flesh. We’re also prepared to accept animation, so long as it’s on the cartoonier side. It’s when realism and animation start to encroach on each other’s territory that the trouble begins. That point of convergence—too realistic to accept as a cartoonified representation, never realistic enough to accept as the genuine article—is called the uncanny valley.

Psychologists have done some recent research on macaques, which might shed some light as to why this happens.

Princeton University researchers presented images of real monkey faces, unrealistic animated faces and realistic animated faces to five monkey subjects and recorded how long they gazed at each. Similar to the human response to objects in the uncanny valley, the monkeys avoided looking at the most realistic animated faces. The scientists, who published their results in the Proceedings of the National Academy of Sciences USA, speculate that realistic animations might resemble sickly or diseased animals because they lack subtle cues of health such as normal skin texture and hue—and that an aversion to such sights may have evolved to keep us healthy. [Scientific American]

While this is a compelling explanation for why CGI characters never quite seem to look right, I doubt it will deter further attempts by movie animators.

H3daches

Then there’s the issue of 3D, which presents yet another psychological puzzle.

Recent web-discussions on the issue of ‘Avatar h3dache’ seem to agree that the problem lies in presenting the viewer with a fairly rich 3D environment, but no opportunity to choose to focus on a part of the scene that was filmed (or rendered) ‘blurred’, i.e. out-of-focus foreground elements such as leaves. Some of us seem to be fighting Avatars determination to make these choices for us, and getting our cognitive perception in a twist in the process.

In his determination to avoid criticisms of traditional ‘jack-in-the-box’ leveraging of 3D (wherein a director will engineer a shot so that things deliberately swing out at the viewer) Cameron seems to have compromised by shooting as much of the movie as possible with a very limited depth-of-field, in order to accentuate the 3D illusion.

…

Examining unfocused areas of the frame in Avatar is literally quite a headache, and counterintuitive to our enjoyment of the ‘baked and locked’ 3D planes that we are being presented with. Knowing that depth-of-field is all he has to play with if he’s not going to shoot rocks directly at us, Cameron doesn’t hold back - he relentlessly racks focus in scene after scene.

So the trick to avoiding a headache when watching this movie is to be obedient, and concentrate on the parts of the shot that the focus tells you are ‘important’. Once I understood this at the preview screenings last week, my headache began to clear up, but I was conscious too of the effort of having to ‘zip over’ to the next point of rapid-focus in order to keep up and preserve the 3D illusion. [Shadowlocked]

To simplify, your eyeballs are focusing on an unmoving 2D surface—the movie screen—which is lying to them, telling them to focus and refocus on distances that aren’t there. The actual surface isn’t going anywhere. Like realistic CGI, nothing in nature has prepared our eyes to accept this.

I humbly submit that the audience shouldn’t have to “be obedient”—in other words, politely ignore the movie’s problems. After all, if 3D has any possible benefit to the art of moviemaking, you’d think it would be giving the audience multiple planes of action to look at. While Cameron’s 3D renderings have a deep, rounded appearance, the movie itself—the images on the screen—stick to just one plane of action at a time. It moves no differently than a 2D movie. When Avatar goes from one plane of action to another, it doesn’t do so by composing in multiple planes and letting the viewer decide. It drags the viewer, through rack focusing.*

Texture aside, Avatar isn’t terribly 3D at all. In classic movies, there is a longstanding tradition of “deep focus” shots that show everything clearly, far away and all the way up to the foreground. This style of camerawork gave filmmakers many planes of action to play with. In The Rules of the Game, characters bicker like children in the foreground while other characters sneak around them in the background. It’s played for laughs, and it works. In Citizen Kane, the boy plays outside the window while his parents debate his future inside the house. These movies might not “pop” at you with the special glasses, but they are composed in 3D. Multiple planes of action, playing upon one another. That’s real 3D cinema—movies that use depth in the action.

The irony is that Avatar might cause even more headaches if it were photographed in this way. With eyes skating all over the place and constantly refocusing on depths that aren’t there, the presence of actual stuff to look at might just drive them haywire. Just give them one surface to look at. Be confident that the viewer’s brain is smart enough to bring the movie into the third dimension on its own, without the aid of big plastic glasses. So long as the movie is good enough, of course.

(*Rack focusing is when there are at least two things happening on the screen: one near the camera, and one further away from it. The camera focuses on one, and then quickly refocuses onto the other. You see it happen every other second in Paul Greengrass movies.)

This entry was posted on Monday, April 5th, 2010 at 7:40 pm and is filed under Criticism/Critical Thinking, Film. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.