Hi everyone I’m wondering how I can improve the subD geometry when I reach these kinds of edge loops. I’ve always felt a bit lost when I reach these kinds of semicircular loops and how I would ‘rebuild’ them.
Would like to learn from you all. Thanks in advance for the help! (I’m on Rhino 8). Original rhino model attached in the middle of the message. Thanks!
I think you’ll learn subD topology better if you avoid using creases until you understand how to sharpen edges by adding edge loops near the areas you want to refine.
By adding two loops (drawn in blue) on either side of the one on the red line, you’ll create a sharper crease the closer they are to the red line, the sharper the edge will be.
I also think you’ll learn more by modeling closer to the final silhouette of your object, rather than starting with a very low-poly mesh and relying on smoothing to define the shape SubD.3dm (133.3 KB)
Thanks Gijs! Yes that’s really helpful! It’s what I needed! My main question now is workflow, how to get there. In this case I have a drawn closed curve that is accurate, but how did you go from the curve to the subD you sent me? How did you get these transverse edge loops in such rotated directions? I feel quite uncomfortable going from a curve to such a ‘non rectangular’ subD, especially with rotated edge loops.
Thanks Tommy! Yes that makes sense, and is helpful. I can really see that in your model. I wonder what workflow approach did you use to arrive at your model: the edge loops are kind of complex and splitting in multiple directions (see green circle annotation). Also some are 3-sided, four sided, rhombus-shaped, and some curve (in smooth shade) and kink (in flat shade) at very precise angles. This is wholly different from how I learnt subD as mostly ‘rectangular quads’ and blobby rectangular primitives, and more complex like nurbs.
I wonder how you got from the model I sent to the model you sent. Did you loft it? Or Sweep 2 then add edge loops and move those? How do I conceptualize it?
Might also be worth watching some blender topology tutorials. Not the same as Rhino but it still conveys the idea that your subd topology is important. Subd feels like a ball of clay you can have absolute freedom with but the reality is that it’s still a modelling tool that requires planning and good constraining of topology. It’s very rare for me to have a face with more than 4 sides.
First, I rotate your model so that it’s flat along the world C-plane; it’s usually easier to model that way, and then rotate the finished model. Modeling mostly with rectangular quads is a good approach when you’re first learning SubD, but it’s not an absolute rule. Contrary to what you may have heard, strictly modeling with quads can limit your ability to handle complex shapes. It’s perfectly fine to use triangles, n-gons, and poles in your model, as long as you understand how to use them properly. N-gons are fine on flat areas and can still produce clean results on curved surfaces. One thing to remember about n-gons is that they are divided into quads during smoothing—the same goes for triangles. Check out the post by odd_enough on twitter for great uses on n–gons on non planar surfaces.
All quads is a good approach to modeling in general, but you can break that rule especially when making hard surface models.
Here a quick and dirty video on how i modeled it, i dont usally do subd modeling in rhino. I usally would use blender.
I’m wondering how I can improve the subD geometry when I reach these kinds of edge loops. I’ve always felt a bit lost when I reach these kinds of semicircular loops and how I would ‘rebuild’ them.
