Talk:3dm2sculpt
Ok...seems there's work left to be done. Made some stuff in Rhino for testing.
A bent cube - the "I can't possibly screw this up" example
Teapot - a more complicated shape, the classic object for any 3d test
- Sculpt Map | Screenshot in Rhino | Screenshot in SL (the one on the right) (The one on the left is a torus, but it's lost its hole and split in half.)
Good luck though. Elle Pollack 20:06, 7 May 2007 (PDT)
Oh, thanks for the Bug Report. Seems like i should get a Demo of Rhino, too, to do more testing. So far, my Converter can only handle a single Object (i will change that, but each Object will result in another Sculptmap). Another Limit : the Object needs perfect UV Coordinates.--Cindy Crabgrass 21:13, 7 May 2007 (PDT)
- I think the UV mapping could be throwing me off. What the heck is meant by "perfect" UV Coordinages? The above models all have a spherical map applied to them, which is what I had started to think was what was needed based on how the sculpt prim is based off a sphere. Then again, I've seen a lot of stuff in the last coupple days that seems to break that rule, like the heightmaps people have been doing in paint programs. I may have to pester Qarl some more. Elle Pollack 22:09, 7 May 2007 (PDT)
I did some more testing with Rhino Files and found something : most Rhino Objects are made of more than one Surface. The twisted Cube (i made one for testing) should have 6. My Previewer showed exactly one nice and twisted surface. Thats a Problem... can I create a Sculptie from multiple Surfaces ? More Research needed :)
Let me quote Michael Gibson : I should mention that this algorithm only works for certain conditions - your object should be made up of just one single surface, not multiple surfaces (like a cube has 6 surfaces in it normally, this won't work. You have to create a special cube that is made up of just one surface instead). And also the surface should not be trimmed, like it should not have been run through booleans or those types of processes. That's because you want the UV space to be completely covered by the triangulation which is only the case for untrimmed surfaces.
I'd love to get my paws on the source, I want to make a Managed C++ interface for it (bring it into .Net). I want to add it to the SculptPreview mod I'm working on. -- Strife Onizuka 13:04, 9 May 2007 (PDT)
The latest version (besides the download being slightly corrupt, nothing a little hex editing doesn't fix) crashes on most (if not all) of the OpenNURBS sample files, [1]. Not to mention that it sometimes exports things in a crazy fashion (example: v3_SphericalSpiral.3dm). I wish it would output at 128x128. -- Strife Onizuka 15:16, 9 May 2007 (PDT)
Ouch... well, you asked for the source, i guess all you need is this : here it's based on the OpenNURBS example_read, and has the same Dependencies. The source is a real Mess... but you asked for it :P I guess you will rewrite it, its quite simple. --Cindy Crabgrass 15:24, 9 May 2007 (PDT)
- You make it sound worse then it really is (or I just have a high tolerance for bad coding). It's allot simpler then I had expected. -- Strife Onizuka 19:27, 9 May 2007 (PDT)
Yes the OpenNURBS Library does all the hard work... i wished i could do the same to .obj Files, that would be a lot more useful, but the more i learn about it, the harder it seems. Rebuilding the Topology of a random Mesh to a Sculptie - someone will do it sooner or later.--Cindy Crabgrass 19:56, 9 May 2007 (PDT)
- Converting an arbitrary mesh to a sculptie would be a hack, it would require an alg willing to hidden polygons. Not to mention having to refactor the vertices. Then comes the mess of having to convert textures because of the change of UV cords. It wouldn't just be hard work, it would be a nightmare that would result in imperfect results (as a programmer and a bit of a perfectionist, that turns me off). I do not envy anyone taking on the task. -- Strife Onizuka 06:35, 10 May 2007 (PDT)
Just for Fun, i improved my old obj2sculpt a bit, now it 'resamples' from the UV coords to the Sculptie. The results are not exact, because the interpolation within a Triangle (the Barycentric Stuff) is still missing. At least, it finds the correct Triangle in the input Mesh - so it works OK on Huge .obj Files. The real Pain is unwrapping the Mesh to a closed UV Map...