hey!
in my misadventures to and fro the UV rabbit hole, I've encountered a (for me) perplexing behaviour...
when unwrapping a (UV, #nopun ; ) sphere, the pole regions are separated into their own island DESPITE (at least intended) seam placement...
(see screenshots, hope that they illustrates clearly)
?
for the peeeps in the know (instructors and/ or students alike):
am I missing something obvious?
: )
Thanks!
it appears part of the problem is that instead of one vertex at the pole you have 36 or so...
if you remove doubles or weld the 36 down to one I believe you will achieve your desired result...
hope this helps
hmmm...
: )
thanks for chipping in me1958424 that gives me... fresh per perspective!
specifically, never thought of that (hiding in plain sight) "vert dilemma" although, my goal is NOT to have them merged or welded, quite the contrary...
the screenshot bellow will explain better than I ever could in words; the "default" (as in, unwrapped with no intervention whatsoever) is what I was after, only "manually" understanding what's going on "under the hood", with some "custom tweaking" (body quads square, etc)... the "spikes" (triangles) should be separate (each ending in a vert of "its own") while remaining attached to the body in a single unified island and, how I thought that is achievable is exactly by marking as seams all the edges ending in the pole; the puzzle is, why do the pole regions separate into island, without any seams (and not stay attached as expected and as they infact are in the "default") and how to potentially go about it achieving the desired result, but, most importantly understand the why...
I believe that I got it!
: P
(not sure what's the policy, mark as solved, remove thread altogether or keep it if someone may find it useful)?
What seems to have separated the spikes into islands is that I've treated the initial (now reproduced successfully) unwrap with 'follow active quad' (a tool completely new to me; didn't assume that only quads follow quads (am I on the right track now at least?? : D )and drop the tris)...
the funky unwrap (now acting as expected, as far as the overall integrity is concerned) but with heavy distortion as a tradeoff:
distortion addressed, at the cost of a single continuous map being broken down into three:
Although the initial problem is somewhat solved, I would still appreciate any tips and suggestion on this (both specific and more general) kind of solutions from the people with more UV experience...
: )
generally let the thread exist as you said for future reference to someone with same issues...
interesting way you are approaching it, still unsure of the advantage of leaving the verts unwelded at the poles...
but if it works as you want it to...