When we're looking for new tracks, we want interesting natural and man-made landmarks and we want versatility." As level designer at Codemasters' racing studios, Glenn McDonald's window on any given project rarely coincides with that of the programmers. "We're a racing studio," explains Glenn. "So we start working on tracks even before some projects are officially started. That way we have content ready to insert into the new game sraight away."
When sitting down to plan the tracks of a new racing title the world is Codemasters' oyster. "Essentially there are a limitless number of possibilities," laughs Glenn. "For Dirt 2 we had in excess of 100 locations mentioned in the brainstorming session. We tried to draw on our own experiences and based areas on places we've been."
To whittle Dirt 2's track suggestions down, the team used Google and Google Earth to research the areas mentioned in the brainstorming session. Expect Dirt 3 to feature even more: "For future projects we're certainly striving to expand the number of environments we use," Glenn stresses, unprompted.
With the shortlist finally decided upon, it was time to rack up the air miles. "We always either sent a research team from the studio or we'd get an outsourcer to go and take photographs, notes and diagrams. We preferred sending team members, though, as freelancers couldn't always capture everything we need.
"Obviously places such as Utah and London are quite expansive. We had very specific roads and areas in mind before we left. Sometimes, though, the areas weren't as good as we'd hoped so it was then up to the team to take stock and find somewhere else nearby."
First and foremost, the goals of these trips were to find ideal tracks. In reality perfection is unlikely, so the team had to learn the art of blending reality with fiction. "If we found particularly interesting landmarks and landscapes within an environment we used our 'game license' to work them into a track. We were always on the lookout for incidental things: a good sequence of corners, nice camber, interesting undulation, buildings, waterfalls... What we always tried to avoid, however, were the wacky tracks often seen in competitors' games."
While the designers were busy scouting for natural dynamism, the artists needed to gather materials. "They took hundreds of texture shots. They mounted an SLR camera on a massive pole and held it over the road, then used those shots to create accurate textures back in the studio." It's easy to see why the team prefers not to rely on outsourced materials for these research trips: the jobs of the designers and the artists are utterly dependent on one another.
When the designers returned to the office they constructed 2D plans of the proposed tracks. As Glenn explains, it was much easier to modify a track at this stage. "We looked for problematic corner sequences - a crest before a tight corner which would launch players into the environment, for instance. It's easy to iron these issues out at the 2D planning stage rather than when the track's been built."
Most track details were fleshed out at this stage to give the artists plenty of time to construct new assets. "On its own a straight is boring," offers Glenn. "It doesn't require any input from the user. However, there are certain things you can do to improve it. Height change is a good example. If you've got a crest it adds a blind element to mask what's further on, and when you come over this crest the car will feel lighter. You could also add a pinch-point - an archway would force people to thread through a narrow gap at great speed." Knowing this, an artist could begin designing that element long before the track shape was finalised. These decisions are usually left to the sole designer: Codies encourage designers to take ownership of a number of tracks for healthy internal competition.