yelocloud opened this issue on Sep 15, 2007 ยท 315 posts
bagginsbill posted Wed, 26 September 2007 at 8:23 AM
Just to strengthen the understanding of the UV issue, I saved the Poser Hi-res sphere as an obj file. Then I filtered it to find the UV coordinates for the equator (where V = .5). Here are the 36 UV texture values for the 36 points around the equator.
vt 0.142536 0.5
vt 0.166258 0.5
vt 0.189971 0.5
vt 0.213677 0.5
vt 0.237374 0.5
vt 0.261064 0.5
vt 0.284748 0.5
vt 0.308427 0.5
vt 0.332103 0.5
vt 0.355776 0.5
vt 0.379448 0.5
vt 0.403121 0.5
vt 0.426797 0.5
vt 0.450476 0.5
vt 0.47416 0.5
vt 0.49785 0.5
vt 0.521547 0.5
vt 0.545253 0.5
vt 0.568966 0.5
vt 0.592688 0.5
vt 0.616419 0.5
vt 0.640157 0.5
vt 0.663904 0.5
vt 0.687656 0.5
vt 0.711415 0.5
vt 0.735178 0.5
vt 0.758944 0.5
vt 0.782712 0.5
vt 0.806479 0.5
vt 0.830245 0.5
vt 0.854008 0.5
vt 0.877767 0.5
vt 0.90152 0.5
vt 0.925266 0.5
vt 0.949005 0.5
vt 0.972735 0.5
Notice that there is no U = 0 or U = 1 as there should be. The smallest U is .142536 and the largest U is .972735. So the first 14.2536% of the image is never shown, and the last 2.7265% of the image is never shown.
Taking the difference, we see the the U span is .830199 units wide. I used .83333 after much visual experimentation, but now that I've done the math, the correct U_Scale would be -.830199. I'll have to do some experiments to find the correct U_Offset to go with that.
Renderosity forum reply notifications are wonky. If I read a follow-up in a thread, but I don't myself reply, then notifications no longer happen AT ALL on that thread. So if I seem to be ignoring a question, that's why. (Updated September 23, 2019)