View Full Version : NURBS,MURMS,Sub-d Question
06 June 2002, 08:30 AM
Okay I just want to get this cleared up in my head.
NURBS are quad patches with CV points and isoparms. NURBS can only have four sides and in the case of a triangle one edge of the quad is collapsed to a point but it's still a side, so a triangle still has four sides with one side that is really small. A NURBS patch can have a unlimited amount of CV points on any edge or isoparm. Also if you join two NURBS patches together they become a single quad patch. Joining patches is tricky because the edges need to line up. In the case of a NURBS cylinder there will always be a seam because a NURBS patch always has to have four sides and the seam needs to carefuly blended to be invisible. This is what I understand to be a true NURBS or Non Uniform Rational Bezier Splines.
NURBS patch is not the same as Patch modeling right?
A Patch or Patch modeling would be a Bezier Patch which can consists of three or four edges each edge is a Bezier Spline and they're used to make a patch. Each edge can only have two CV points and each point can have only two control handels that control the curve's shape. You can attach two patches together by welding the verts but then you have two patches attached together unlike NURBS where to two become one.
NURMS (lightwave) or MeshSmooth (max) is a poly mesh subdivided to smooth it, and has nothing to do with NURBS or Patch Modeling.
Box Modeling can also be called Sub D modeling, but Sub D modeling isn't necessarly Box Modeling? Or is Sub D modeling NURMS/Meshsmooth and box modeling just for poly meshes?
Now I'm a max user and this is how I understand it, is about right for all 3d software?
Is a NURBS in Max diffrent from a NURBS in Maya or Soft Image.
Is MURMS in lightwave diffrent from Meshmooth NURMS in max?
How about bezier splines are they basicly the same across the board?
06 June 2002, 05:10 PM
I won't speak about NURBS since I don't really use then and don't want to pass on any mis-info by mistake... but NURMS are actually from max when you mesh smooth... lightwave used to call it's subd objects metaNURBS... but now they just call them subpatches or subd objects... which is a lot better since lw has never supported nurbs and that old title was very misleading...
as for box modeling that's more *how you build your models as opposed to *what you're making them out of... you can, and I think most people do, use box modeling when modeling subd objects... but you can also go point by point if that's what you like
06 June 2002, 08:33 PM
Sub-D simply means subdivision. When someone refers to Subedivision modeling it simply means that whatever modeling technique they are using is intened to be a low res, and then sub-D'd to a high res mesh.
as for nurbs, nurms, hypernurbs, Ubernurbs and the like. the best way to understand them is by what theya re called.
Non Uniform Rational B-spline (NURBS)
Nun uniform Rational Mesh Smooth (NURMS)
HYPER Non Uniform Rational B-spline (Hypernurbs)
Uber Nun Uniform Rational B-Spline (Ubernurbs)
These are just some of the various names you'll find lying around, some people also call Nurbs Non Uniform Rational Bezier Spline, which is a litle misleading, ad they use a certain smoothing algorithm used in B-splines, where as a bezier has handles and lets you control each points interpolation and direction.
NURMS, HyperNurbs, UberNurbs, and Mesh smooth, are all various smoothing techniques, most of which I assume use the Catmull algorithm to smothe out the model through subdivisions.
ANyone that wants to add on or correct antythig feel free, this is just my understanding and little tidbits I have learned here and there.
06 June 2002, 08:53 PM
Okay I looked up Catmull and got this. (http://symbolcraft.com/pjl/graphics/subdivision/)
Kinda interesting. It's called Catmull-Clark and it uses the algorithm in "Recursively generated B-spline surfaces on arbitrary topological surfaces" by E. Catmull and J. Clark (Computer-Aided Design 10(6):350-355, November 1978).
06 June 2002, 08:58 PM
Thats the one. There are other algorithms, but this is the one used by most, in their own vaiations of course.
01 January 2006, 08:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.