I’m curious concerning the staff’s expertise, the software program you employ for various facets of the artwork, and the way you are taking one thing from a live-action TV present and adapt it for a sport. What’s that course of and pipeline like?
Ryan McMahon: We use Unreal Engine 5 because the engine, so numerous the rendering is finished by that. We use Maya for a lot of the 3D modeling. Our idea artists use some Blender mixed with Photoshop, and the character staff makes use of ZBrush to sculpt the fashions. All of the atmosphere and character artists use Substance 3D Painter for texturing. We additionally use some Substance 3D Designer for the tile maps and environments.
We’ve got a fabric library in Substance 3D Painter that’s used throughout most of the characters and environments for metals, non-metals, leather-based, and related supplies throughout the totally different heroes.
For the likenesses, it’s all accomplished by hand in ZBrush. If a photograph scan is accessible, we use it, however we don’t have a lot photo-scan knowledge for the actors. Lots of it’s hand-sculpted, which is a large problem. Lots of effort and time goes into getting the likenesses proper, with some back-and-forth with the actors as nicely. The artists use PureRef to construct giant reference boards of all of the characters from totally different angles. They collect as many photographs of the actors’ faces from totally different angles as they’ll.
For the environments, we use numerous procedural content material era for map set dressing. We’ve got a superb PCG workflow, a lot of the design will be accomplished by the sport designers whereas constructing a map. The artists then do fairly a minimal set-dressing go on the finish, dressing over the terrain. We additionally use some Houdini for procedural-generation work the place we will.
Our animation workflow may be very attention-grabbing. We use a method known as VAMP, which is principally a vertex-animation materials pipeline. We rig and animate the characters usually, then use Unreal Engine to bake all their animations down right into a texture on a fabric. On Age of Darkness, we used a method known as VAT, which is similar sort of factor however by the feel pipeline. Basically, it’s a texture by which the UVs run within the materials, and it offsets all of the vertices to play an animation relatively than requiring a rigged skeleton on the character.
One problem we had on Age of Darkness was that we didn’t have animation mixing as a result of it was only a texture, so it snapped between animations. With the VAMP model, we now have animation-blending help. We do that as a result of it permits us to render many extra items on display screen at a decrease value. We don’t need to account for the rigs and bones on the characters as nicely.