The Pole

[SSwayne]

From what I've gathered, from the better tutorials I've come across, as well as somewhere on someartist's blog (or his thread on Wings3D form), it is recommended to spin an edge or two of a newly extruded area, which results in the area being much more "blended" with the rest of the mesh, pretty much what you've accomplished in the last image of your last post. I've started to do it in my modeling adventures, and it's certainly improved my results.

Like you said, you get some side-effect loops happening when you spin edges, so for noobs like me it can add to the confusion... but I try to look at fixing the trash loops as the opportunity to practice all these techniques.

[Toontje]

Don't worry, side effect loops can easily be eliminated by unpoling it. SomeArtist showed how that is done. It is shown too in the first page of this thread I think. Just cut across both sides of the E-pole and merging some tris afterwards. So, keep modeling and keep unpoling

[King Mango]

I don't understand why "E" and "N" have been used to designate these poles. I don't see any resemblance... Am I supposed to be able to see an "E" and an "N" when I look at them? Why not "5" and "3?" It seems there is some way of perceiving these poles that is eluding me because I am still finding the correction methods confusing.

[Magnas]

An E-pole (extruded) is one you get when you extrude a face while the N-pole has 3 edges, like the capital N. |\|

[johnfnoo]

The N stands for nose pole, it is a little confusing because he has the n's have 3 edges concept in his explaination, but they actually look like a Y. Here is the entry: http://www.subdivisionmodeling.com/f...read.php?t=907

[King Mango]

ah ty. I kept trying to figure out the flows based on their resemblance to the letters and was failing miserably lol.. ty very much Magnas and johnfoo.
I forgot the reference to extrude in that explanation because of the numbered "E" diagram. That little diagram made me think I should be able to see an "E" shape formed by the edges and that completely obliterated the "extrude" reference from taking seat.

[Toontje]

Originally Posted by King Mango

I kept trying to figure out the flows based on their resemblance to the letters and was failing miserably lol.

AH! That is a strong argument. Now I see that indeed it can be confusing. The best thing is to name them by their number of connected edges (3-pole, 4-pole etc). But the thing is that we used those letters so much that I wonder if it would be prudent to turn back. But if you name it by their numbers, I doubt that it would bring confusion. It's kinda like when their discovered that it was the electrons that flows, so electrons flows from - to +, but the current flows from + to -.

By the way, N-pole to signify a 'Nose' thing is wrong also, because poles normaly come in pairs, also at a nose!

I have some more post coming up about edge flows and such. I'm far from finished. Keep tuned guys.

[SSwayne]

I have also seen the "E" pole referred to as the star-pole or star-connection, as in -> *

@Toontje: Looking forward to your posts...

[Joe]

Originally Posted by Toontje

By the way, N-pole to signify a 'Nose' thing is wrong also, because poles normaly come in pairs, also at a nose!

Not just "normally"... If you are working with a closed, all-quad mesh, odd-number poles always come in pairs.

You can even go one step further: Let's call 3- and 5-poles and 3- and 5-sided faces "defects". If you have a closed mesh that has defects but is otherwise all quads and 4-poles, you will always have an even number of defects.

This implies:

• Any operation that adds a defect in an area that has none always adds another defect at the same time.
• You can't remove a defect by itself -- you can only move it around.
• The only way to eliminate a defect is to move it next to another defect of the same type (pole or face) and eliminate both at once.
• One way to eliminate a defect on a symmetric mesh is to move it to the line of symmetry, then merge it with its twin.
• Another way to eliminate a defect on a symmetric mesh is to start a new defect on the line of symmetry and move it to the defect you are trying to eliminate, then eliminate them together.

I have found these points extremely useful when cleaning up my meshes.

Note: E-poles are as poorly named as N-poles: While it is true that you will get 5-poles around the base of an extrusion, for each 5-pole at the base you get a 3-pole at the tip of the extrusion (thus preserving the even number of defects). Since extrusion makes equal numbers of both kinds of poles, maybe we should call them all "E-poles"?

[Toontje]

@Joe: that is very true. Awareness of how these pole play out on the 'canvas' is of paramount importance to good modeling.

I seem to be reiterating myself from time to time. I want to keep stressing that for organic modeling it is important to blend the loops. And talking about pairs, there is nothing better than to separate those pairs, and preferably as far from each other as possible.

Before you say: "again another diagram? ". Be patient, I'm going somewhere with this. I can go model a muscular monster intuitively, but I wanted to lay it out as clear as possible before we can go crazy.




I began with a circular loop. I moved the circled N-pole E-pole twins form each other. Observe that all poles are still in the same circular loop. Normally E-poles and N-poles are joined together by that extra diagonal edge of the E-pole (green), but break that bond, and your topology blends better. There is also an extra loop here (yellow: the N-pole bends it both ways here). You can see that the separated E-pole causes nice diagonal flows.

I don't think that modeling a face should be difficult by now. The human face/ head has a relative simple topology (ears not included). Fantasy creatures (see Golem's head topology for example), muscular bodies and such are a lot harder.

If no sudden discoveries comes in between, I'll be talking about limbs, how the loops should flow around the root of the limb to connect to the muscular feature of the main body.