Pixar has $180 million to spare, and doesn’t mind having hundreds of people manipulating 400 yottabytes to create one film. Good for them. But that shouldn’t scare off the rest of us.

The barrier to creating great looking animation doesn’t seem to be time, money, or resources; it’s the misconception that those, in fact, are an obstacle. Video game developers constantly show us that there is the potential for great looking visuals without an excessive amount of polygons. The reason video game graphics don’t yet rival Hollywood CGI isn’t because there isn’t enough processing power for detail, but that there isn’t enough processing power for implying detail.

Simple depth of field can dramatically improve a low-detail image. By simulating the focus of a camera lens, not only is a more photographic look achieved, but strategic use of it can remove most of the ugly portions of an image. This is possible in Valve’s Source engine, and can look great, but it’s not practical for video games; about two frames of it per second can be rendered in real time. But movies don’t need to be rendered in real time – a three minute high-definition sequence from Final Cut can take up to half an hour to render on my laptop, but after that it plays back smoothly; Pixar, by contrast, produces films that would take a single supercomputer several million hours to render (that’s why they have a ton of supercomputers), and they too can be contained within a smoothly-playable video file. But by taking graphics which could, theoretically, be rendered in real time, then rendering it with realistic looking blur and smoothing effects at two frames per second, an independent animator wouldn’t need to have access to supercomputers with enormous hard drives to make a film.

Perhaps I was too quick to call Meet the Scout on par with The Incredibles or WALL-E. Jenni Chasteen‘s comment about the lighting design was spot-on – Pixar has people who know cinematography, and Valve isn’t nearly as experienced. I disagree that it’s not even close, though; while it’s not great for photorealism, “cartoony” CGI is very possible to do with just a video game engine and blur effects.

Valve’s promotional movies may not be Pixar-quality, now that I think about it, but the technique and technology has the potential to be. In the hands of talented filmmakers, it’s wonderful for compositing onto live action. Is extending it to a full animated film that much of a stretch? I doubt it, myself.

Perhaps I’m uneducated in the ancient Hollywood art of Unit Production Management, but it’s baffling to me that WALL-E had a budget of $180 million. Yes, it was a gigantic Hollywood production, but consider the fact that all of its visuals were made by pressing buttons and waiting for the images to appear. Well, it was more complicated than that sounds, but that’s basically what they did.

So what exactly cost so much money? Without access to Pixar’s financial records, I’ll take a few guesses. But the short answer is that they’re spending way more money than they need to.

I’d imagine, from my experience, that some of the the biggest expenses in filmmaking are film stock, set construction, stunt coordination, and other tangible materials. Pixar requires none of that. The only materials they really need to buy are computer parts, which are relatively cheap on their own and downright bargains when bought in bulk. Yes, their concept artists need clay and paint and such, but that hardly seems like it would cost more than a couple thousand, even for a gigantic team of the kind of overzealous madmen who spend more than a week in an art school (i.e. me).

So that brings us to visual effects, which is essentially 100% of what you see on the screen. Yes, WALL-E and Happy Feet had live action bits superimposed, but anyone with access to a green screen and a DVX can imitate those. The actual CGI, though, is achieved by making a bunch of 3D models, texturing them, arranging them in a world, moving them around, and then pointing a virtual “camera” at various parts of the world to get the images you need.

It may sound simple (save for the whole “talent” part in terms of modeling, texturing, and animating), but to create the impressive visuals, Pixar needed a very large amount of polygons. Add to that the calculations necessary for lighting, post-processing (focus, motion blur, etc.), and spitting that out into 24 images per second at 4000 pixels wide, and you have something that requires a lot of very expensive computers to eat a lot of expensive electricity for a very long time.

Except it actually is as simple as I made it sound in the paragraph before that. Take a look at this video:

Impressive visuals. If you had never heard of Team Fortress 2, and that video cut off after the first 30 seconds, wouldn’t you imagine that perhaps this was an upcoming Pixar film (albeit more violent than their previous efforts)?

Now, obviously, this is actually a promotional video for a game. Let’s take a look at the game’s graphics…

Obviously not quite as impressive, but you’d be hard pressed to say that the Scout model in the game isn’t the same one in the Pixar-quality promo video. In fact, the line between the two is very blurry — specifically the fact that the promo video has more blur effects based on motion and focus. Well, that’s because it’s absolutely true.

Meet the Scout, rendered in Team Fortress 2, contains a significantly smaller amount of polygons on screen at a time than a Pixar movie like, say, The Incredibles. Can you tell?

On close inspection, you can see that the Team Fortress 2 models are slightly more jagged than Mr. Incredible. However, it’s unlikely that you’d notice this unless you were looking for it in a freeze frame. In addition, Team Fortress 2′s models are actually far less complicated than those from many other video games. Take, for example, BioWare’s Mass Effect:

(Image from Gamespot)

There, slightly more complex models and the jaggedness is solved, but still significantly less intense than what Pixar does. And most blockbuster video games cost things that are considered “indie budgets” to Hollywood: Gears of War cost $10 million, $5 million less than the pretty much indie-produced film Slumdog Millionaire. And as you can see in that article, many games cost “crazy figures” like $30 million, which is still pocket change to Hollywood. Off-season movies like Taken cost somewhere around $25 million, for god’s sake. Now, if you were to take all the work on Gears of War related to getting the thing to run smoothly on an Xbox 360, and channel all that into putting the camera in the right places, pressing “render,” and letting three laptops run for 10 hours to give you a 2 hour movie, I’d imagine it would cost even less.

So now we’ve thoroughly debunked the myth that CGI is inherently expensive. What’s left? A-list voice talent? Yeah, I can see that being up there. And marketing costs a lot. Indie filmmakers would probably replace those two strategies “five of my friends” and “Twitter,” respectively, but I understand Hollywood’s obsession with such luxuries. That can’t account for all $180 million, though, can it?

Well, in addition to Pixar’s overkill in terms of hardware and technical complexity, they employ a gigantic amount of people, paying them salaries and such. That’s probably necessary because of the technical complexity of their animation.

In theory, a team of four passionate geeks who already have MacBook Pros and copies of Maya, Photoshop, and Final Cut could probably make a Pixar-quality film in their spare time. Perhaps it’s the massive, daunting complexity of Pixar films that discourage budding independent filmmakers from trying to compete using more efficient methods.

Or maybe they’re all developing video games instead.

Comments on this post are closed because nobody read the gigantic bold text at the top, and I’m tired of having to deal with this all the time. If you really must insist on explaining to me just how much of an idiot I am, you can scroll all the way up to the top of the page and click on the Contact tab to find out how to send me hate mail.