For the past week I’ve been debating on how I should handle all of the stone blocks that are used extensively throughout the temple (for the walls, floors, columns…etc). Part of this process is to insure that I am mindful of the actual placement of these elements. I’ve been comparing the Heidelberg photographs to Ranke’s site illustrations. From what I can see they seem very accurate.
I’ve started placing proxy bricks (basic polygonal cubes) in the Pronaos and ramp.
I’m testing a few approaches to see which one might hold up better in VR. Initialy I thought I would break the various elements up into sections, then using either Mudbox or Zbrush, I would detail the proxy cubes and extract a normal map back onto a more simplistic type of geometry. I tested this on the ramp, which kind of works. However, because this is just a bump map if someone where to view this on a shear angle, the illusion of detail would flatten. Just a side note, once I get into the Unreal Engine I might revisit this approach with displacement maps….or possibly vector displacement maps.
My hope in using this approach is to keep the interactivity within the virtual world crisp and speedy. I wasn’t particularly happy with the initial results, although untextured (lacking diffuse colour, specular maps….etc) I could tell that there was a fidelity issue, particularly from extreme angles.
My second approach is to texture 2-3 groupings of individual low resolution blocks then hand place each one (similar to how I placed the proxy cubes to begin with).
UPDATE: The known floor tiles (according to Ranke) have been placed, as of June 8th. I’ll need to fill in the gaps with a similar patterning. But I’m moving onto the interior walls first.