The approach to sound in a video game should be no different then the approach to graphics or any other simulated sensory element. Just as a video game should have mesmerizing graphics, so should it have realistic sounds to draw the gamer into the game environment. Whereas, graphics can be created from the eye, mind, and hand of an artist, sounds must be captured. In order to mimic reality, I apply considerable commitment to three important steps: capturing the source; processing, editing, and sculpting the sound into the essential sonic elements; reassembling the elements so they can be run within a game in real-time.
Overall success is gauged by how well these three things are accomplished.
The following section will focus on the first of these steps, utilizing
a typical recording session of a racecar in preparation for modeling an
engine set for a cross-faded sound system.
In order to get the specific vehicle I want to record, I always have to establish some contacts. I start with car associations and auto clubs, and from there I eventually (after running into many dead-ends) find the vehicle I want. This stage of the game can be a wearing exercise in itself, especially knowing that you've committed yourself to a contract and that 'giving up' is not an option. The upside is that I've had the pleasure of meeting some fascinating people and occasionally I get a ride in their cars.
Before I head out on the open road with a couple of DAT machines and a case full of microphones in hand, I usually have the recording session worked out on paper first. This usually eliminates the forgetfulness caused by typical early morning starts at the racetracks.
It is an advantage to know how the game's sound system works and the platform to which the sound will be implemented. For example, most console boxes have memory constraints so that the samples have to be short and this can be quite tricky for cross-faded engine sets. If the game has a larger memory budget that can accommodate separate in-car and exterior engine sets, and allows for stereo samples, then I go with stereo in-car samples so that the ambient qualities of the cockpit and chassis are captured.