Hi, take a look at this. This has boggled me for a while.
Delete these and then the middle one disappear too.
SubD delete bug.3dm (53.8 KB)
If I delete the top two first and then the bottom two, or visa versa then it works as expected.
Hi, take a look at this. This has boggled me for a while.
Delete these and then the middle one disappear too.
SubD delete bug.3dm (53.8 KB)
If I delete the top two first and then the bottom two, or visa versa then it works as expected.
This was a quick test to see how 50% soft edge would appear compared to the old fashioned custom built âfilletâ, on a build like this.
None smoothed version:
And here is the workflow of that last build:
SubD corner on round - step by step.3dm (195.1 KB)
@schultzeworks this might be of interest for you, a simple âhard surface SubD buildâ.
Feel free to adapt or use directly in teaching/video.
I started with a cylinder with 24 around faces (4x6) and then deleted the caps and the 3 quadrants I didnât want.
Inspired by this blender(?) quick tutorial, but didnât like the âvertical buildâ since I like my modelling to be none directional, so I changed the modelling process to be more accurate and equal.
Hey @Holo
Wow, very cool workflow. Thank you for tagging me so I saw this.
This the kind of transition that I see attempted all the time in NURBS and I never cared for it. Why? There were so many DANG steps and I always ended up âmodelling myself into a corner.â Meaning, there was no easy way to edit; just re-build the whole thing. So, I avoided these transitions / situations whenever possible.
But, your suggestion on using Sub-D is EXCELLENT. It is still editable! Amazing.
I also appreciate the step-by-step screen cap. Thanks again.
Yeah I donât see why this would happen or be the expected result.
RH-84837 SubD edge dissolve or delete: deletes too many things
My pleasure Dave!
And thanks for taking the time you spend on spreading good workflows!
Exactly, for me SubD is perfect for concept studies and designing, and then when the design is locked it can be rebuilt for accuracy.
You probably know all this, but just in case, and for anybody else who stumble upon this:
The painful part of SubD is figuring out the best patch layout and sticking to quads. It is the best modelling by far IMO.
And toggle between smooth and unsmoothed often. If it looks good unsmoothed then it will look good smoothed, and will be much easier to modify.
And use as few points as possible (like with curves) and use the same patch layout for different parts IF possible, it will make editing so much easier and faster if you can turn on points for all parts and move them together.
If the model is clean and as simple as possible (but no simpler) it will transfer to cycles and render fast too:
Or 7 seconds at 100 iterations to check the latest customer ideasâŚ
Oh thanks this was bugging me every once in a while
So every material is a different separate part?
Yes, that is faster for me both in the short and long run. Easy to select edges, extrude with gumball and snap to the other geometry. On builds like this it is important to remember that if it looks good enough it is. And then you can detail it up when needed.
Have you 3D printed such a car? Would be nice on a polyjet or one of those new mimaki ink printersâŚ
No, not yet, itâs on the todo-list⌠We have a Snapmaker 2.0 collecting dust here, so it should be good enough to test it out⌠One day when I donât have so many projects going on!
These started out as SubD evaluation/learning processes where each build has had a different focus from a technical point of view + it scratches my dormant industrial design itch since I now work with landscape architecture
(If you recall I even made my own mesh subdivider that keeps the initial vertice location and semi-inflate the mesh instead of shrinking it like CatmullClark does + some efficient modelling tools for fast building of control cage meshes, but I have not had time to adapt that to Rhinoâs SubD. Mostly since the SubD structure is a whole new animal)
Iâll look into the mimaki though! Thanks!