How does the realroad technology handle separate graphical and driving meshes? For example, if the physical mesh (the one that is driven on) is not rendered, but a low poly graphical mesh is (but not driven on), will rubber still accumulate on the low poly mesh if both have the requirements for the rubber buildup to occur?
RealRoad requires the interaction of cars, i.e. drivable RaceSurfaces. If you're driving on the high poly mesh, you're not interacting with the low poly mesh and vice versa, whether it is visible or not. Not very ideal ... Both pMotor and gMotor can handle a good amount of polygons though. On our recent tracks, RaceSurface tricounts have more than doubled compared to our early rF2 tracks (which weren't very optimized), but even those were already tripled from the higher quality rF1 tracks. Good object, texture and material optimization is a much more limiting factor in terms of smoothness (which is much more important than the FPS numbers people keep staring at). How many tris is your road mesh?
That's what I thought would be the case unfortunately, just seeking clarification as I hadn't seen it brought up in any documentation. However, I thought could try and get away with less work, as in AC (which I was looking to bring the track to as well) the use of two meshes is more common. While still very much a WIP, it's looking like it'll end up at around 200-300k tris.
If it's a long track, 200-300k tris should still be alright ... usually means the scene isn't packed and the layout isn't snaking around itself. Heaviest RaceSurface I've seen is 150k tris for a ~5km track. That was overkill, and the scene wasn't optimized in any way, but it ran reasonably well on a powerful CPU.
