So, I'm modeling a cave using Recap Photo to generate high poly-count meshes with numerous large texture maps. Working in chunks, a key step is to establish a common scale and coordinate system to bring each new (overlapping) mesh into Maya, scaled and aligned to the previous (even though each the meshes isn't brought into the same scene until way down the road after they've been processed). I'm told by Autodesk support that in order to satisfy the needs of "other software" (I'm thinking it's AutoCad, which I believe is where most folks making use of Recap Photo do their work with applications in architecture), fbx meshes coming out of Recap Photo are scaled up 100 X, also rotated -90, -90, 0. This causes a two-fold problem in Maya, thanks for any ideas how to work around.

The workflow for propagating scale and coordinate system calls on me to select 4-5 common points of interest in the overlapping region of two meshes, the first one viewed from a particular camera in Maya, the second one viewed from the same camera over Autodesk 360 in Recap Photo. With the component editor, I select the vertice closest to the common feature, use a MEL script to return translation, copy/paste one at a time between Maya and Recap Photo, output at max quality in Recap Photo, download the fbx, and bring it into Maya just to see if the alignment routine worked. Since I know these fbx meshes are scaled up and rotated, I know I'll need to scale the new mesh 2 back down and unrotate it to see this alignment. The fbx mesh inside the parent node and sibling to each the cameras reflects a scale of 100, I set it back to 1 and change rotation from -90, -90, 0 back to 0, 0, 0 — nice!, the alignment is dead on. Problem is by not working at the parent node level, the cameras don't follow along, which I'll need downstream using them for cleaning up textures with projection mapping. If I instead rescale mesh 2's parent node from 1 to .01 and rotate from 0, 0, 0 to 90, 90, 0, I don't see the alignment, I can't even see the mesh, have to believe it's there, but must be beyond the clipping plane or something. I don't understand the unpredicted result.

A tip may be that Autodesk support also advised me that Maya doesn't like meshes scaled way big, perhaps the downscaling and rotation trick isn't working at the parent group node for related reasons. Might there be another way to first downscale and unrotate each new mesh in Mudbox, 3ds, or Softimage (these came with Maya Entertainment Creation Suite), such that the cameras follow, the new fbx allowing two adjoining meshes to share the same scale and coordinate system?

I'm a total newb to CG, thanks for some handholding, this issue seems ridiculously simple, and yet it's a significant barrier if I can't generate properly scaled and positioned meshes at the outset.

Best,
Benjy