Thank you very much taz, I missed it completely. I knew it had to be there somewhere. Now the definition not only doesn't require scripting but it looks exactly like the rhino's built in method. http://grasshopper3d.googlegroups.com/web/curvatureanalysis2.jpg
On Nov 9, 7:57 pm, taz <[EMAIL PROTECTED]> wrote: > viscose, > > I'm not sure if this is what you're getting at, but if you have points > projected on a surface you can still use the <Surface Closest Point> > component. > > This distance will be zero (approximately) and it will give you the > surface UV coordinates. > > Does that help? > > taz > > On Nov 9, 5:50 am, visose <[EMAIL PROTECTED]> wrote: > > > In fact, driving geometry is much simpler than coloring a mesh. Since > > i think (maybe someone can help me here) there's no way of getting the > > uv coordinates of a point in 3d space on a projected surface, you > > need to do some scripting to order the sample points that are closer > > to each vertex of the mesh and create a parallel list of curvature > > numbers that coincide with the number of vertices and is in the right > > position. > > > I guess there's no way of finding the equivalent uv of a point in 3d > > space, even if it was projected on a surface, because once it's in > > world coordinates there's probably some rounding errors that makes it > > not exactly on the surface. Nevertheless, could the project component > > have an uv ouptut? > > > On Nov 9, 10:28 am, fraguada <[EMAIL PROTECTED]> wrote: > > > > Definitely...it just depends the logic structure of the geometry. > > > Just like some parameter is going into coloring a mesh, that parameter > > > can drive the differentiation of the geometry. Simple example would > > > be getting the surface frames and placing a circle on them with a > > > radius that is based on the curvature. You could even extrude that > > > circle based on the same curvature parameter. It is just a matter of > > > translating the value from curvature to something that makes sense for > > > the geometry. > > > > luis > > > > On Nov 9, 9:17 am, oompa_l <[EMAIL PROTECTED]> wrote: > > > > > Hi > > > > > just wondering if you could use curvature analysis not to colour a > > > > surface but to generate geometry. I am sorry if this is super > > > > elemental - let's say I have a surface and I want to "enclose" areas > > > > by a new curve at the boundary between different degrees of curvature > > > > - instead of differences in colour. I would want the curve to lie on > > > > the surface being analysed. I presume that the gradation from one > > > > colour to the next is constructed by the user/designer but I am unsure > > > > what controls these "shifts". Better put, could you create threshold > > > > values that triggers different conditions - if there is a really > > > > "steep" area it might be cordoned off by this new curve while in an > > > > area with minimal slope something else happens... > > > > > sorry again, and thanks for any advice. > > > > o
