The ISDF, since its contemporary base reference was the United Nations, originally was designed with a white and grey scheme in mind. After some time wrestling with that color combination, we decided that something bolder was needed. Since we really couldn't do realistic camouflage patterns (doing this would have made the units blend into the terrain too well and hence, was a design limitation), we decided after some attempts, to go with the orange and black scheme. Since the ISDF was basically a covert United Nations army, we wanted them to look regal, like a police motorcycle, but flashy enough to look cool.

The Scions were a military group which were much more organic in nature. They had this ability to morph into different vehicles. This nature led us to design a much more organic texture scheme for them. We needed textures that would be capable of bending and contorting such that they would look good under any morph condition.

One thing we couldn't help but notice in the most recent screenshots is that the look of the Scions seems to have changed. What prompted this change and how did you arrive at the current look?

The Scion side was extremely challenging to resolve. After designing a texture scheme which could simply be described as a skeleton with a tight skin wrapped around it, a number of people thought they were too organic (and a little too gross looking). We tried two more treatments until we arrived at the current design, which now looks much more metallic and military. This actually works quite well as the morphing vs. hard metallic look provides a nice contrast.

What unit has changed the most from the original concept sketches? Were the changes made because of a change in the unit's function or because you just didn't like how it looked?

As was the case with original BZ, the Recycler has been the most redesigned unit in the game. The reason for this was that it was a unit that had to convert itself into a building. This being the case, it needed to look good both while driving around and while planted on terra firma. This presented a number of problems with constructing it so that the pieces could fold into themselves and deploy elegantly at the end. While it may have only undergone one drastic change since its conception, the change required a lot of forethought.

The other unit that was also a pain was the Scion Scavenger. Later in the project it was decided that the unit would have tracked physics instead of hovering physics. This involved a change. Then it had to directly convert into a building. Then it was too tall. Then it was too short... The list goes on and on. It's done now.

Can you comment about the issue of visibility and identifiability of a unit vs. fitting that unit into the world/environment?

I mentioned that we could not do a camouflage scheme due to game design limitations. I will now explain why . The beauty of camouflage is that it does just that--it camouflages--and in our game it does this exceedingly well. As a result, the game designers and myself thought that designing a camouflage scheme would hurt the gameplay since seeing units would be difficult. Since we want you to like this game and we don't want you to spend inordinate amounts of time finding your guys, we opted for a very forthright look to both the unit sides.

What is the thing that you most look forward to being able to do with game art that isn't technically feasible now?

While we've been able to do a lot of great stuff with BZ2, current gaming technology and scheduling permits only a limited amount of programming time dedicated to dynamic events--that is dynamically changing environments and animation. Soon I hope to be able to work with an engine which could take my animations and change them to fit a given situation, like being shot from the side versus the back, versus the front, etc. That way, the environment would seem unscripted and the player would really feel like he was changing part of the world dynamic. While we can do a certain amount of this, most noticeably in the particle effects, it will be great in future titles if we can implement it on a larger scale.