YOU ARE HERE: LAT HomeCollections

Game Design

Making Waves More Art Than Science in TransWorld Surf'

Physics helps power the interactive ocean, but animation does much of the work of making the surf look good and varying the water's action.


At Neversoft, the game development studio in Woodland Hills responsible for the "Tony Hawk" skateboarding series, aspiring developers must take a test, one that includes the question of how to make water.

The answer does not involve smashing two hydrogen atoms with an oxygen atom. Nor does it invoke the Navier-Stokes equation for simulating water turbulence.

The correct reply, as it turns out, has more to do with art than science.

A great deal of hard-core physics is required in the creation of water for video games. But because consoles don't have enough horsepower to simulate the fluid dynamics of anything bigger than a glass of milk, developers must find clever shortcuts to lull players into their virtual water worlds, especially when players are required to interact with the water rather than merely gaze at it.

Steve Rotenberg, 29, director of research and development at Angel Studios in Carlsbad, talks about how his team created an interactive ocean for "TransWorld Surf," a title published by Infogrames Inc.

Question: So what does it take?

Answer: There are four major things about water that the human eye will pick up. One is the physics of the movement of the water, or how it animates. Another is the reflections off the water's surface, the refraction, and what I would call subsurface scattering of light, or the diffusion of light. Those are the three visual components. Then there's the fourth--the physics of the water itself.

If you have those four things, you can do convincing water. But to do those four things properly requires more computing power than a game system is going to have. So the trick is to fake it, or capture the essence while preserving controllability for the designer. It's a big balancing act.

Q: Where do you start?

A: My approach on a lot of things is to study the real thing first. If you want me to do an ocean, I'll start with how the ocean really works. Then I'll go on to find out what the human eye perceives.

But I don't commit myself to the real thing. Techies like myself have a tendency toward doing things the most technically interesting way, or toward making it super-realistic. But you have to know where to cut the corners and know how to preserve different aspects and discard others.

There isn't a universally correct way to do water, because every game is going to use water differently. The important thing is to keep in mind the ultimate goal. Usually, we make sure that fun is the No. 1 goal.

Q: How do the goals in your game differ from other water-based games, and how do those differences govern your design decisions?

A: Take "Wave Race," "Splashdown" and our game. They're all solving different problems. With "Splashdown," the water is basically flat with ripples. It also channels through rivers. It's the same with "Wave Race," where you're creating a track with water of varying turbulence with obstacles scattered throughout. Our game recreates an ocean with waves. Waves curl over. It behaves differently. So we devote a different set of resources to make that happen.

Q: Where does physics come in?

A: Some of our initial experiments involved a more physical approach. We generated ripples on a computer based on fluid dynamics. As the game went on, though, we moved away from physics to more procedural animation, meaning a technique where you use algorithms to drive animation rather than physics.

Proper physics has two drawbacks. One is that it requires a lot of computing power. Two, it tends to be more difficult to control. Game designers want precise control over how the wave breaks, how fast it is, how large it is, how many there are. It's easier for them to just direct that, as opposed to coaxing the physics simulation to behave the way they want.

So, for example, we thought of modeling the bottom of the ocean, and having the wave shapes be based on the shape of the ocean floor. But for the designer to change the wave, he'd have to change the shape of the ocean floor.

Physics certainly has its place. It's particularly cool for handling more unpredictable situations. But if you want very precise controls over something, physics isn't always the best choice.

Q: How much real-world physics did you incorporate into "TransWorld Surf"?

A: Close to one-third to half of the [central processing power] is allocated to the ocean, in terms of physics and rendering. That makes sense for a surfing game. But very little physics of the ocean is left in the game. How the surfer responds to the water is physical, and that remains. But the wave is more precisely controlled through animation.

People have the impression that it's a lot more physical than it is. In fact, it's more physically inspired than physically based.

Q: There have been many stabs at surfing games. Nearly all have wiped out. Why is it so hard to make a fun surfing game?

A: Two reasons. One is the challenges involved in getting it to look good. A wave is a very dynamic thing. It's constantly changing. That adds technical difficulty. The second is game play. We went out of our way to put as much stuff in the world as we could.

If you were just doing one wave after another, it would get tiresome. We knew we had a challenge in making the levels feel different, giving players something at every level that they hadn't seen before. Previous surfing games hadn't put as much effort into this. So they tended to just have one wave after another. That's not very exciting.

We also put a lot of effort into the feel of turning and performing tricks. Other surfing games just steer and turn. We tried to put more subtleties into the controls.


Alex Pham covers the video game industry. She can be reached at alex.pham@latimes .com.

Los Angeles Times Articles