Revitalizing a Classic

Darrin Klein,
Art Director,
Jumpgate Evolution

Darrin Klein, NetDevil - I started on Jumpgate Evolution a few months after Auto Assault shipped. I had been on [Auto Assault, or AA] for most of its production (3 years). At the time, Warmonger and LEGO were in their early stages. I was approached by my Art Director, Peter Grundy, about working on JGE rather than moving to one of the other projects. I saw it as an opportunity to take a stab at heading a team and gaining the further experience of creating the art from the ground up. I started by working closely with the programmers to hammer out things like geometry exporters, shaders, lighting, etc. Because of my work on AA I had a lot of experience with modeling/texturing and particle FX. That experience really helped me to determine what the art production pipeline would initially need for JGE.

Darrin - For Jumpgate Evolution my responsibilities are to oversee that the art team produces the visual content we need at the level of quality that we have set for the project. Seeing as the art team consists of three people, this has been very challenging. Because of the size of the team we had to really focus on methods used to develop graphics at the quality we wanted. We spent a lot of time on the first few assets to really make them polished all the while keeping in mind the techniques used to create them. Once we had some really great pieces made, the others came much quicker and of the same quality if not better. We knew that it would take more time with this approach but the end result has been well worth it.

Darrin - Initially we didn’t know where to start with graphics and art direction. We spent a lot of time considering different avenues. Cartoon, hyper-realistic, happy-pappy, dark. These were all directions we could explore. Finally we settled on a more realistic feel being mindful to keep it from being too gloomy. We achieved this by having sharp contrast, lighting and high saturation in the textures and skyboxes. In order to “tie” these elements together we implemented things like lens flares and random particle debris or clouds. All of this is kept at a pretty high color saturation to ensure that any given “scene” has a dramatic feel.

Darrin - Priority number one was “RUSH TO PLAYABLE”. As soon as we could fly around in what would become our “space” we would have a much better feel for what would need to go in it visually. The three of us as an art team have one goal, and that’s make it look awesome. We agreed early on that we wanted to come as close as we could to FPS visual quality in an MMO. All of that while maintaining a nice low min spec has been our biggest battle.

I feel that we have maintained that standard as a direct result of “iterate, iterate, iterate”. The JGE team as a whole has adopted the approach of taking the first sector (equivalent to a level or map) and reworking it till it’s awesome in both visuals and game play. Until then we don’t move to the next one.

Darrin - We started from the ground up with a new graphics engine developed specifically for JGE. While it is new, it isn’t technologically taxing. We always have min-spec looming over our heads and because of that we really only have the engine do what is absolutely necessary. We saved time by removing graphic effects that really aren’t apparent to most players. This allowed us to once again keep system specs low and spend more time on the quality of the art. A large factor in what was necessary for the graphics engine came from the time spent in concepting great pieces. With finished concepts in hand we could easily point our graphics programmer in the right direction. If we absolutely need to add something new we just have to make sure that it’s worth the payoff.

Darrin - In many aspects it was the other way around. The concept art really set the tone for what the new graphics engine needed to be capable of doing. After we had our first ship modeled and textured we put it in game and developed the shaders for it to make it look like the concept. For us it’s the mentality that if you start with great concept art and maintain that quality through the rest of the pipeline then you get great assets. And if great art doesn’t pop out the other end of the art assembly line then it can only be one other thing. Crap!

Darrin - A lot of our art production techniques have been developed out of the need to have as low a min spec as possible. For instance we’re using normal maps and detail maps much more extensively and wisely. We want a lot of detail but not the poly counts. Also we use little tricks like adjusting the overall specular power in the mip maps of our textures. This allowed us to give objects in the distance a sort of “shiny glint” with out having the programmers develop some sort of distance based texture bloom. A lot of our work comes down to “What can I do or figure out on my own without adding support time to other peoples work load”. It’s all born out of necessity… you have to make due with what you got.