10.09.2008

Up Close and Too Personal Part 2

Picking up where I left off two weeks ago, I'd like to continue with a little more on camera's and camera problems. In the last camera topic, I discussed the problem with camera's being in too close, close to the point where the game became frustratingly difficult to play.

The counterpoint of course, is too far out. But what about the perfect camera height? Why is it difficult for the players view to simply be placed in a good position perpetually.

The problem is twofold, but intertwined. The problem is that while a camera angle from an ideal height may be easy to find, it is another problem all together to support this camera in the game world. To set up an example, in movies a script is concrete and a characters actions determined beforehand. Regardless of how many times the director shoots the scene, a camera only needs to follow its prescribed course to keep with the center of action. However, in a game, script is based upon the players actions. A so called "rail-camera" (so called due to the idea that the camera is based on a set of rails that follows a prescribed course) therefore must have previously predicted every action the player makes or the character will vanish off camera, often in the middle of something critically important, leaving the player with no image of what is happening.

There are several solutions to this. First: The developer may confine the player to a static path and program the camera's course to a fixed rail system. this limits the players interactivity however, reducing a free experience to a series of set paths. While this may be the easiest camera set up, it often leaves players with a less rewarding game, often making them feel as if they are simply following a scripted routine.

Another option is the "fixed camera". A fixed camera employs a static unmoving view angle that jumps perspectives as the player walks from one scene to another, like a security system cycling through camera's to monitor someone. On one hand, this system allows great freedom of exploration to great effect, often why it is used in point and click adventure games such as LucasArts dearly departed Curse of Monkey Island series (may it rise from the grave, zombie-like, but smelling better) or the early Resident Evil games. Of course, nothing is free, and this system has several drawbacks, control being primary. Fixed camera games often suffer from difficult control issues raised by the question of whether or not a character should move relative to themselves or the perspective. Without going into detail here, both systems can be awkward and difficult.

A "free camera" is the most often employed camera system, and although good developers will often mic and match cameras to striking effect, most games gravitate towards a camera that will follow the player like an inquisitive paparazzi group. The advantage is that the player almost always has full view of his character, allowing for the most free exploration and a feeling that the player is in control. The big drawback: Camera clipping. Camera clipping is one problem that has yet to be eradicated. It can be fixed, but as it requires time and therefore money, publishers and developers tend to slip this under 'bugs' and pass the savings on to themselves while handing you a less polished game.

Clipping occurs in one of many situations, but the most common is when either an object comes in between the camera and the player or the camera is forced into a space smaller then its distance from the character. In the case of an object coming in between the camera and character, the problem is obvious. You cannot see the character, simply a close up view of a random object. Or worse, an inside view, where the camera's physical properties 'clip' and the camera slides inside an object, resulting in flickering close ups of polygons (shapes used to build objects in games or computer effects) and untextured (or blank) surfaces. This problem also occurs in small surfaces. A classic example is to walk a character into a corner, then turn him around. Center the camera behind him, and it will either go through the wall, or give you a very close clipped view of your characters back.

So how do we solve this problem? There's no real easy answer. Some games give marginal camera control to the player, but often, this only allows the player to further play with an already slightly broken camera, doing as much harm as good. The only real solution is for developers to knuckle down and spend more time at it. Successful games have done this such as Super Mario Galaxy, a stellar example of a carefully coded camera system. Doubtless weeks of work went into the system, but oh does it pay off. Almost (note: almost, nothings perfect) never do you feel like you're out of control, which is a fair testament given Galaxy's unique "3.5D" style (YouTube it to see for yourself).

So, there is no magic bullet for cameras. Just more work, but better results.


Games I'm currently playing: Crysis (PC), Command & Conquer 3: Kanes Wrath (PC), Megaman 9 (Wii), DDR Universe 2 (360), Final Fantasy IV (DS)

No comments: