Cineversity

In the real world a Triangle has 3 sides and a Quadrangle 4. (even with my limited brain capacity I know this to be true :/  )

In the polygon world this has come into question.  If you have a triangle shaped polygon surrounded by other polygons and there is an edge/vertex terminating in the middle of one of the triangle edges,  is the triangle really a quadrangle with two edges that are just 180° apart.  If you move that vertex even a fraction the polygon becomes a quadrangle.

I have attached a jpg to illustrate.

This question comes into play when using hypernurbs.  We always want to avoid triangles AND we don’t want an edge that terminates in the middle of another edge, potentially creating an N-Gon.

The real question is “how does C4D look at this triangle with 4 vertices?  Is it a triangle or a quadrangle?

I hope that wasn’t too confusing.

Thanks

TAK

Hi TAK,

It is a “Quad”, not a triangle, it is the number of points that counts. Points are the main information in C4D.
If you use always HyperNURBS in such a case it should be OK. The resulting situation is critical, not so much the underlaying information.

Well, case-depented of course, and any general answer would be just not OK here. If you can avoid such situations (180º), I is certainly preferable.
To my knowledge, when it is processed for rendering, anything will turn into triangles anyway. So why to use Quads then in the first place? Because animated surfaces change the way a quad could be set into triangles to follow a smoother result, bend in one or the other way. If triangles used from the start this is not possible anymore. The “Help Content” has a chapter about degenerated polygons, please have a look into that, if you like to dig deeper.

Let me know what you do, still, animation etc. and I might be able to add information.

All the best

Sassi

This was really just a question for overall modeling work and curiosity.  I was watching a tutorial on YouTube and the instruction was to use the knife tool to add that extra vertex.  And then the tutorial said to move that new vertex so that you now had a quadrangle.  I just wanted to know if that extra step of moving the vertex was actually necessary.

Like you said it is case dependent.

I have found that triangles in the midst of a flat surface really don’t cause any problems.  Nor do N-Gons.  It is really only at corners and edges, or of course curved surfaces.  So I will sometimes disrupt clean geometry on the flat surface to make sure I have quadrangles at the edges.

Thanks Doc… I knew you would have the answer.

TAK

Hi TAK,

There is a lot of information all over the place, with a wide variety of quality, as is everything on the web. I don’t know what you have seen and I normally avoid source that I don’t know at all. (...please no link, thanks ;o)

In HyperNURBS the idea is “practically” like pulling on the points, and triangles work typically not that well there. Typically, which means that there are exceptions.
The ideal is a four-sided polygon with equal length, and worse case is defined as a triangle with two sides very long and a very short edge. The majority of the created polygons will be between these two extremes.  N-Gons are great for modeling, predominately as “step-between”. I try to get n-gons solved before I start texturing, as there is no “UV N-Gon”. It works, but I’m personal not really happy with such a solution. But admittedly teh developers have created a very stable solution, even with n-gons, so it is more a personal perspective.

The project target is certainly here a big part in the decision process, hence my reluctance to put a sloppy general rule here. However, if the flow of the final mesh is harmonic and very balanced, it’s easier to think about it as a nice mesh.

I have edited my message above, so it might seem repetitive, but the polygon that needs to be avoided is a “degenerated” polygon. If you keep that in mind, most cases should work. But then again, if a HyperNURBS is used, the generated information counts for the surface to be rendered. You can even select the degenerated polygons with an option in the “Main Menu>Edit>Project Info>>Attribute Manger>Structure”.

All the best

Sassi

It was actually a very good tutorial on model building.  I was just curious because the guy would always move that fourth vertex just a little to create the Quad.

I understand that everything is case by case and the final result is really the determining factor of how simple or complex a model needs to be.

Thanks

T

Hey TAK,

I never would comment on a tutorial, especially not in public. There is enough “cyber-bulling” on the web already, and sometimes it is just covered as “serious” critique. I do not play that game on the web, that is not my style.

I did not put into question what you have seen. Your reply indicates that you understood it in that way. Sorry. I should not even discuss things based on other peoples work, I will think about future questions based on 3rd party tutorials, perhaps I should just ignore it.

It might be the case that an animation moves that specific “point” just a little bit back and you end up with a degenerated polygon, case-depentend. As we all have no idea what will happen in the future to an model, I suggest to keep things safe—and I know that is not always possible.

Enjoy your explorations

Sassi