UV Painting Project and Hellboy build
- Thread starter ZooHead
- Start date
ZooHead
0
If your looking at the fluting, it's an older version of the cylinder.
I wondered if anyone would notice and say something, my money was on you.
I wondered if anyone would notice and say something, my money was on you.
I'm not sure if that's a good thing.
And I guess those "teeth" come from "making editable" a bit too early (at least, on my ipad I haven't seen them). Hard to repair on a dense mesh without subdiv.
This is why I don't use booleans very often: all the cleaning up to get an all-quad-model (which I probably would do in such a case). But there are no rules in poly-modeling (no subdiv), and for subdiv-modeling it's more guidelines than a rule you have to follow.
Even while I advocate all-quad-modeling, there are a few exceptions (very view) where you end up with a tri you can't get rid off or (even more seldom) with a ngon that looks perfectly well but as soon as you cut it up, you'd have to remodel half the mesh because of serious pinching or so (in such situations, though, I sometimes go back to an early version of the model to do the whole stuff anew, this time trying to avoid the ngon-situation. With tris I usually can live).
Even while I advocate all-quad-modeling, there are a few exceptions (very view) where you end up with a tri you can't get rid off or (even more seldom) with a ngon that looks perfectly well but as soon as you cut it up, you'd have to remodel half the mesh because of serious pinching or so (in such situations, though, I sometimes go back to an early version of the model to do the whole stuff anew, this time trying to avoid the ngon-situation. With tris I usually can live).
ZooHead
0
Long ago I realized any flat area can have tris or ngons without consequence when box modeling.
(All tris and ngons are turned into quads under the magic of Catmull and Clark.)
Even so I always try to make all quads if for no other reason than better loop and ring selection.
But this is the first time I've tried this technique and I learn
more if I push the envelope to the breaking point.
(All tris and ngons are turned into quads under the magic of Catmull and Clark.)
Even so I always try to make all quads if for no other reason than better loop and ring selection.
But this is the first time I've tried this technique and I learn
more if I push the envelope to the breaking point.
Subdividing once I sometimes use when I have a part of a model that needs denser geometry than the rest - it's a cheat, not always that clean, but most of the time you spare a lot of time with that. But even on flat surfaces I get rid of ngons (because you never know if there isn't that slight difference between some verts).
With "this technique" do you mean not cleaning (much) up after a boolean?
With "this technique" do you mean not cleaning (much) up after a boolean?
ZooHead
0
I'd like to see you do that.
My approach is to try anything I can think of and Booleans
always seem to be a time saver but often with complications.
The problem is making the perfect shape of the Capsule in quads is non trivial.
Maybe you could use a capsule as a guide for point placement
but it's already more complicated that way.
My approach is to try anything I can think of and Booleans
always seem to be a time saver but often with complications.
The problem is making the perfect shape of the Capsule in quads is non trivial.
Maybe you could use a capsule as a guide for point placement
but it's already more complicated that way.