how to add a half cone please?

Many, many years ago, I helped beta test Bryce2, which had just added booleans. As everyone was learning how to use them, we wound up competing to see who could use booleans with the Bryce objects (Sphere, cone, cube, ring. Just the very simplest of primitives) to model a toilet. The beta-sharks came up with some amazing solutions.

Even though the software has been extinct for half a decade, Bryce still has the strongest boolean engine I know of. While most software I've used starts to choke when you put a boolean inside another boolean, in Bryce, I've regularly gone a dozen live layers deep (Bryce has no polygon modeling tools).
Good ol´Bryce: 2005: https://www.cheetah3d.com/forum/showpost.php?p=1348&postcount=11,https://www.cheetah3d.com/forum/showpost.php?p=7611&postcount=9
:p

Cheers
Frank
 
* Conceiving a Boolean hierarchy with 7 layers (as MoneyT mentions about Bryce™) seems to be beyond my intelligence. But then, again, I am not a student of Benoit Mandelbrot and fractal geometry :frown: :confused:

Yeah, booleans within Bryce™ were their own beast.

Unlike C3D, they worked by grouping. Every primitive had a flag you could set: positive, negative, or intersect. Once you set the flags, you grouped those items, and they all interacted with each other exclusively, live - in no particular order. This means that if you had two items, one positive and one intersect, the rendering basically drew the first object - because the intersect object was entirely contained by the positive object.

The important part was that once you created a group, that group also had a boolean flag: positive, negative, or intersect. Which made building complex shapes easy, and then applying both of those complex shapes as boolean objects within their own group. And so on, and so on. Each boolean surface also remembered what texture had been applied to each component - on both sides of the new boolean created surface.

Since I was mostly fascinated by building architecture in this naturalistic landscape software, my shapes became very complex. Especially when I figured out that building an interior by subtracting the negative space from a solid shape instead of trying to assemble complex walls made from primitives led to very cool stuff.

You just sort of fell into complicated booleans, closed their groups and forgot how complex they were.

I think Bryce™ also had an advantage because, for the most part, it used procedural textures on procedural shapes instead of keeping track of polygons. A long as you stuck to primitives, booleans were never translated into polygon meshes (as far as I could tell), so there were very few structural conflicts and glitches when rendered, even when animated. It really was amazing.
 
* I suspect this is a typical example where the "language" of a specific application starts to dominate your conceptual 3D imagination. Sort of a linguistic relativity / Sapir-Whorfism, be it strong or weak.
* Questionable / discredited as this relativity may be by now, programming tools are much, much, much simpler than natural language. I am perfectly aware that working in C3D (as opposed to, eg, Rhino) requires a totally different mindset. Even in C3D, starting with poly-based entities requires a different approach as compared to thinking in splines and creators.

* Every few hundred years there are geniuses, Shakespeare, Mozart, Picasso at al, who just blow away any of these conceptual walls. These are just minor warps in cerebral space-time and should not be takes seriously :rolleyes:
 
Don’t knock booleans. In general they work very well (now). They used to be horrible in C3D but today they’re great.

That said, the easy solution to the original question is a cone with sections or a cone with inner extrude on the tip followed by a ring cut and some cleanup.
 
Don’t knock booleans

I don't knock them and Helmut gives in one of his posts a great example where they are not just helpful but necessary.

What I tried to say (not very successful it seems) is just that a beginner shouldn't use them very much.

I told about my own experience but know of other people who really got into a dead end modeling wise because they are very tempting in the first stages of learning 3d. But they don't bring you much further.

And yes, Cheetah's booleans are quiet good, the resulting geometry is ok, but usually not a good way to create clean geometry.

Also, there are no hard and fast rules for 3D modelling. Arch Viz, stage design, engineering, human modelling, scientific concepts, 3D printing (and a few more) are far too different to be comparable.

I don't agree fully. In my opinion it has more to do with the techniques you use to model (for example subdivision), hard edge and organic types. You can create a hard edged mesh with several methods, and it doesn't seem to me, that it is very important for what you'll need it in the end. Of course, all this different types of 3d usage have their special problems and necessities. For gaming you still need relatively low poly models with normal maps, 3d printing is special, for arch viz you often have to deal with imported CAD-Models (and in the worst case have to recreate them). And so on.

But you use the same techniques for modeling in the end, have the same materials and problems with animating. So for example with human modeling it's not the modeling techniques that are difficult to learn (it's the same as with other organic forms), but the knowledge of anatomy. This is the main reason why I could never model a real photorealistic human being.

So there are a few principles in modeling that stay the same, like to understand a form to really be able to recreate it in a believable way.

Bryce still has the strongest boolean engine I know of

Well booleans have gone a long way since. There is mesh fusion (modo) and similar plugins for maya or 3ds max, and of course live booleans in zbrush. Those produce in the end geometry that is or can be subdivided (only a bit dense).

And stacking of several forms, added, cut or intersected, to get the end result is probably similar to the layers mentioned. There are possibilities to change, thicken or lessen, the edges, too. And you can group this stuff, see the results live, and in the end create a normal mesh out of it.

Stuff like this is on another level than the usual booleans; it's a modeling technique of its own.

That said, in Cheetah booleans are very mighty and well done. Don't get me wrong here.
 
Bryce’s Booleans were an exploit of Ray tracing rather than modeling so it’s not a fair comparison.

With booleans I’d say they are great for beginners and experts (now) and will tend to produce very clean geometry (now). Being against booleans because they used to suck is a bit like the way expert illustrator users poo-pooed freehand’s booleans when they first came out. Now no one would ever bother doing such operations by hand.

It certainly helps to know what you are doing when using booleans but almost any approach to avoiding booleans is less exact, more painful, and less obvious.
 
Being against booleans because they used to suck is a bit like the way expert illustrator users poo-pooed freehand’s booleans when they first came out.

Pod,

Nobody (except you) mentioned that booleans used 'to suck' in older versions of Cheetah. I don't even know them. My first contact with boolean operations was long before I ever heard of Cheetah. They didn't suck there.

And nobody is against booleans. They can be helpful or even necessary.

The only point I tried to make is this: They are very tempting for a beginner to use as a jack-of-all-trades, to create everything with them, because it's nearer to the way we think of creating things. And that is a dead end.

You and me have obviously different conceptions of clean geometry. It has not to be all quads in my opinion, but that is usually helpfull. Clean geometry is usually without ngons. And as the resulting forms of booleans are endless in their possibilities, I'm sure you agree with me, that often, not always, it's a bit difficult to bevel the edges. In reality most things don't have such hard edges.

As soon as you combine booleans with subdivision you get into trouble with most forms without cleaning them up a little bit. Yes, it's the ngons.

So for a beginner, in my opinion, it is important not to overuse booleans. They are a tool, but not THE tool. The perception could be otherwise. And I got the feeling (and that of course can be wrong) that it was in this special case.

And, again for a beginner, it's not that difficult to avoid them. You don't run that easily into situations where they are the only answer to create something when you start to learn modeling. But they seem to offer a cheap way out. Probably you agree that the 'cheap way out' is not the best way to learn something.

So it's just a piece of advice I give to beginners, and I mean people who really are starting out in 3d. If you learn modeling, there is no way around patch modeling or box modeling (preferable both), if you want to get any good at it. Never use booleans to avoid this methods. By the way, one of Cheetah's modeler's strongest points is it's subdiv ability. Sad, not to use it.
 
* Let us not argue, folks. :smile: :smile: :smile:
* Every NEW tool sucks because:
a it is new and has a few bugs.
b it is new and our minds take a while to "rewire" to cope with the paradigm shifts in conceptualising / synthesising stuff.
* As to Booleans "sucking" I got to a NSFW page dealing with Henry Boole´s oral preferences.

* I do agree with Hasdrubal that Booleans are grossly overused by newbees. I never understood, why.
* Booleans are NOT a simple and intuitive tool which reflect daily 3D experience. With trivial exceptions (hole => subtraction, merge => addition) there are limited equivalents. Sooner (or later) most serious 3D folks discover that a quick (and somewhat dirty / untidy) method may cost more time than it saves.

* When letters and writing were invented (shortly before my lengthy career in kindergarten) users were quite cheesed off. The first alphabet had a measly 3 letters, clay tablets at that time were limited to 140 characters, lacked a spell checker and, in any case, nobody was literate.
* BUMS! SAD! Totally useless WANK!

* Today, advanced reading and writing is a skillset reserved to the elitist and educated upper classes. For the rest there is Fox & Friends, NFL, NRA, dimwitted action movies, TWITTER and whacking your balls with a big stick.
 

Exactly, the old version has greatly improved.


OK - so just for the record: you can still have tris with
Boolean operations you still need to avoid conflicts with non-coplanar quads+nGons.

Cheers
Frank
 

Attachments

  • TriangulatedBooleans.png
    TriangulatedBooleans.png
    21.5 KB · Views: 240
Back
Top