Thats where the
per seand
buffer polystory comes in. The process worked, but where I
d push-me-pull-you`d the model at the left-right join, despite having creased the edges to resist shrinkage, the polys did get knocked out of line.
I think it would probably be better to model the whole model up to the highest safe level you can in Z-B, trying to get as much detail into those future seam areas as possible; and then perform the dissection, preserving a two-poly deep buffer zone band of polys belonging to the discarded parts of the model. If youre breaking the model up into just a left and a right part, then the left part gets a band of the right part
s polys; and the right part gets a band of the left part`s polys.
Then you just make sure you dont model too close to these buffer polys when you
re manipulating the separate parts tools.
In Photoshop you can superimpose the generated normal maps ontop of eachother and erase away the unnecessary buffer bits.
The advantage of having buffer polys is that even if the bumps and valleys dont perfectly match up, the worst you
ll get is a slightly foggy, blurred area of the composite normal map, where the lower normal map layer buffer poly section shows through the upper layers erased section. You won
t ever get a crisp mismatch seam.
Good luck,
R