Lair

Ha. This finally came out. I expected not to get quoted, as I made it pretty clear I was only on the tail end of the project. But there ends up being a blurb about the control/FOV issues.
This is probably the single feature I wasn’t able to get through that I regret the most, as I think it’d have had a massive impact on how people felt about the controls.
Basically, if you were turning, trying to follow an enemy as they flew around you, even if you had a *perfect-to-the-frame* reaction time, you’d overshoot the target (over-rotate), and have to turn backwards to correct your aim.
This wasn’t a subjective thing – if you were actually relying on your eyes, and not a built-in knowledge of where an enemy would be, the FOV was so narrow, and you turned so fast that you didn’t have enough time, given a standard human’s normal reaction time, and given the momentum of how the dragon slowed their turn, to hit a target. This was made worse by the fact that the game ran in the low-teens/high-single-digit FPS for the bulk of the tail end of development, which made things significantly worse.
So the fix was pretty simple – we added a “camera look-ahead”, where the camera would turn *in addition* to the dragon turning, by a tunable number of degrees, and it’d ease in/out of the dragon’s orientation over a tunable time. Think of it as the dragon pilot turning their head to “look-ahead” as the dragon turned.
It gave us a bit of wiggle room to keep the dragon turn rate independent of target acquisition, and then tune the damping of settling back to a neutral state independent of the dragon’s animation. The result was that if you had decent-but-human reactions, you could get the target dead center of the screen every single time, even from a maximum-speed turn.
It didn’t make it into the final game, which is, I think, a real shame. There are valid reasons why, of course, but I’d have made a different call than the one that was made.

Leave a Reply