Team Fortress 2 - Art Direction
For those of you who don’t follow the gaming scene, one of the biggest games of the season is Team Fortress 2, which comes as part of The Orange Box by legendary developer Valve. The game is receiving a lot of well-deserved press over its art direction which brings a fresh, cartoony style to a genre which generally opts for a more drab, uninspired motif. The result delivers not only delicious eye candy, but enhances both gameplay and visual longevity as well.
In terms of gameplay, the distinct character forms allow players to rapidly categorize opponents and prioritize their targets (hint: shoot the doctor). All too often the figure designs for character classes, while different enough when standing next to each other, are harder to discern in poor lighting, at a distance, or any number of other battlefield conditions. Some may argue that character classes should be hard to discern as it’s more realistic, and that is a very fair argument if realism is the goal. But in the case of TF2, the emphasis is more on fun and balanced multi-player game mechanics, not to mention light-hearted fun/mayhem in general, and the visual stylings support that admirably.
Going with highly stylized graphics is always something of a win if you expect that your game will be played for may years, because their visuals age better and are cheaper to produce than say, photorealism. Take World of Warcraft vs. Everquest 2 for instance.
WoW looked great on day one and still does, despite the fact that it’s graphics engine is starting to show it’s age a little. EQ2, on the other hand, looked like a textbook example of the Uncanny Valley problem since day one. And now, with so many ‘next-gen’ photorealistic characters running around, EQ2 looks even more dated. Games such as WoW, Psychonauts or Rayman 2 continue to look good in part because they stand alone. That is, they aren’t constantly being compared against reality as the newer games which get us closer to the mark.
They are also cheaper to make because of the detail required. Cartoons have large areas of uniform colour, or at worst, gradient colour and highlights. This is much easier to produce than highly detailed textures necessary to bring photorealism to a character’s mesh. In photorealistic CGI, the Uncanny Valley is often the result of the character being “too perfect”. Artists must go though a lot of effort to add all of the imperfections that make the world what it is. In mimicing traditional cel animation, perfection of colour is exactly what you’re shooting for. It still takes a lot of effort to make it look right, but the overall work required is less for stylized/cartoony than for photorealistic.
Here is a video from Valve that talks about some of the above elements, plus a whole lot more. It’s quite technical, but very interesting: