is this case, when beveling an edge, blender creates an n-gon that destroys existing faces, that imo could be without being destroyed to achieve the geo i want. in the picture above (right side) you can see the 3 faces which the edge bevel shouldn’t/doesn’t need to affect (is there an option or way to preserve those faces automatically which i don’t know about? i tried the settings in the bevel options window but didn’t find anything that achieved the result i want).
manually, i can achieve the geometry i need but it looks simple enough for blender to just create the “new face” n-gon instead of destroying existing faces with it’s useless n-gon, don’t you agree?
maybe blender can achieve that “new face” by itself through some other function? in which case, can anybody tell me how?
this sounds like a simple operation which i’d expect to find in the edge bevel options window. the loop slide already follows the existing existing edges to create vertices.
those are way faster hacks than the way i fixed the problem. just shows how unexperienced i am modelling. still, the issue remains that it’s a hack - it seems someone’s reworking the bevel for blender so maybe these hacks won’t be necessary in the future.
Another Method…
4 Edges to connect too, so 4 segments in the Bevel.
Dont forget to M > Merge by Distance to remove the overlapped Vertices. And the lower face in the corner needs to be cut in half to make 2 quads.
There may need to be another edge loop (parallel to the drop to the lower section) on your curved example.
the issue is the destroying of the existing faces that make up the sphere.
when you make new faces and connect them from the old vertices to the ones created by the bevel it will create shading issues cause the new faces no longer makes a spherical shape.
look at the geo
and the shading near that vertex i pointed out.
it’s very faint, but there’s a squared shading.
if you put the mat cap light over that reconstructed area you’ll see a big shift in the shape of the light.
thanks for the link though. been watching some videos on shading and dealing with it.
A common thing in car modeling is to duplicate the model, subdivide as far as your computer can handle, then shrinkwrap the original onto that super smoothed version.
There is also the cast modifier - which will force the mash back into a sphere shape. Use vertex groups to skip your inset.
There is something similar have seen in GeoNodes - that keeps everything to an exact radius.
Mesh Menu > Transform > To sphere is another edit mode tool.
It may be a matter of adding something like 5 loops in there between that bump and the inset.
Also - does it matter? Will that spot be right there big and obvious, or will that sphere be 1 of 1000 zipping around. I remember this being a big aha !! moment for me years ago while following a tutorial. The classic “stressing out on tiny details that make no difference once rendered”. This is something else to keep in mind. I have no idea where you are going with this, but take care not to over do it.
That’s the reason why a professional is hired and payed for this kind of work …
…because this is the Holy Grail of modelling… and those question came up almost daily…
You might just look into an other professional orientated forum using every 3D app you have heard of (or not)… with almost 200 posts in this thread… and of course dozen of similar questions in others… ( because users are unable to search if they are focused too much on ther own problems…) :
It’s not only about finding the form but also the best distribution and the normals…
Shrinkwrapping will not a magic wand and probably will not help you in such cases.
Issues with shading coming not from geometry position imperfection, i.e. the fact its not ideal sphere (despite the fact about its polycount anyway low and its not a “perfect sphere”) but from the way how normals are interpolated and how the shading (how lights are look at the surface) works.
For having ability to create that kind of pretty ugly (compared to classic subd) topology and have good shading without baking normal maps (from ideal subd/dynameshed topology) you should use custom normals.
thanks for the videos ill check em out.
regarding shrinkwrap, i dont think it was suggested as being a magic wand - just something that might help fix the geo, like many things in blender one uses to try and fix stuff automatically.
it’s made liek this//
…
