Edge Bevel

Wow, thank you guys, this is fantastic. I'm going to need a minute to digest all of this as I am just getting refreshed. You are all so generous with your time and your talents. This community is easily the number one best feature of C3D! Thanks again, I will likely be pestering you on a case by case basis if (likely), my brain fails to comprehend any of the above tutorials. 😁
 
There are different methods to achieve such a form, and each of the shown examples certainly has it's merits. But only Misoversaturaded cared about a quad-mesh, and that last method was probably the fastest. In some of the other examples I saw some areas that certainly pinch, especially with a reflecting material.

Now I didn't have the time to do a mini-tutorial about my simple approach, which took me a bit more then 5 minutes and has a polycount of 412 (of course it has a subdiv modifier); just boxmodeling as usual, cube, make the down poly less wide, create an edge in the middle, catmull the upper part, connect the dots with the down part, catmull, bridge after pushing the polys into a circle (there's a script for that I couldn't find anymore) and create the "hole", putting in some holding edges, and that's that (sorry, no minitut because I couldn't even write the first sentence here without being interrupted twice and that was more than 5 hours ago). I know, it's not 100 % the same form as rick intendend but it's something I will use later on myself for a seal ring (don't know if that's the correct word, just checked with deepl and that gives me the same result with the alternatives "signet ring" and "sealing ring" (somehow the latter ones make more sense).

The beauty of a subdiv modeler like Cheetah is that you can keep the polycount quite low, therefore can change or refine as much as you want, and if it still doesn't seem round enough, you just increase the subdivision. And most of the time it's quite fast as soon as you understand the form (that said, spline modeling is more accurate).
 

Attachments

  • ring.jpg
    ring.jpg
    143 KB · Views: 209
There are different methods to achieve such a form, and each of the shown examples certainly has it's merits. But only Misoversaturaded cared about a quad-mesh, and that last method was probably the fastest. In some of the other examples I saw some areas that certainly pinch, especially with a reflecting material.

Now I didn't have the time to do a mini-tutorial about my simple approach, which took me a bit more then 5 minutes and has a polycount of 412 (of course it has a subdiv modifier); just boxmodeling as usual, cube, make the down poly less wide, create an edge in the middle, catmull the upper part, connect the dots with the down part, catmull, bridge after pushing the polys into a circle (there's a script for that I couldn't find anymore) and create the "hole", putting in some holding edges, and that's that (sorry, no minitut because I couldn't even write the first sentence here without being interrupted twice and that was more than 5 hours ago). I know, it's not 100 % the same form as rick intendend but it's something I will use later on myself for a seal ring (don't know if that's the correct word, just checked with deepl and that gives me the same result with the alternatives "signet ring" and "sealing ring" (somehow the latter ones make more sense).

The beauty of a subdiv modeler like Cheetah is that you can keep the polycount quite low, therefore can change or refine as much as you want, and if it still doesn't seem round enough, you just increase the subdivision. And most of the time it's quite fast as soon as you understand the form (that said, spline modeling is more accurate).
Well, this is very inspiring. My biggest problem is that I use Booleans like candy, as geometrically that is how my mind sees things... as some form of added or subtracted material components. I realize this creates overly complex surfaces and areas that nullify the use of certain functions, so I'm hoping I can find some YouTube or Vimeo tutorials that show the proper way of pursuing 3D construction. Thanks very much for your help.
 
Joel has in the tutorial section here started some threads about subdiv-modeling. Youtube and Vimeo are full of modeling videos, but almost always for other apps than Cheetah. Sometimes the makers try to be software agnostic (although not always that successful).

I would avoid Blender stuff, though, but you can translate for example Maya or 3ds max tutorials to Cheetah which hasn't all the bells and whistles of those programs but allows you to learn to properly subdiv model. Everything you really need is there. There are also some vids about topology, i. e. the proper way how to tackle certain problems (how to avoid pinching for example, what to do with poles and so on). Anyway, Cheetah is a good tool to learn the ropes because it's not cluttered with functions and tools, and there isn't much to cheat.

Actually it's natural to start with booleans; most of us did the same. I certainly did and produced messy geo like everybody else til I learned to avoid them. Later on, when you really know what you're doing, you will use them again where necessary or helpful (they aren't bad) when you learn how to produce clean geo with them or for pure poly-models that don't use subdiv at all.

It's kind of hard to get into proper subdiv-modeling, to really wrap your mind around it. But it's very rewarding, because you can be lightning fast (not in the beginning, though). And as soon as you are used to it, it starts to get quite simple.

The trick is to exercise as much as possible and never stop to learn (I still watch tutorials on a regularly basis, sometimes even for beginners, because there is often something someone else does in a different way or you re-learn something that you have forgotten, sometimes just to see how it's done in a program I don't really know.).
 
Back
Top