Poles and Loops


Hi everybody,

I hope that this thread can become sticky, because I’ll be talking about important issues concerning organic modeling. The information I’ll be putting in this thread is discussed at great lengths too over at www.subdivisionmodeling.com.

I’ll try to condense all that information and I’ll try to make it more readable for noobs. Especially noobs! I’ll try to use Some Artists images as less as possible , so I’ll be modeling too from time to time.

The main goal is that when this thread has matured enough, I’m am willing to put it in a tutorial format to publish it over at wiki.blender.org (from Noob to Pro perhaps?)

At any time if you guys think I’m just waisting your time, let me know, I don’t want to be pretentious because I’m no master modeler or something (far from that even).

I’m aware that there is a modeling thread already, but in my opinion the information is not dense enough and it is mostly unstructured, hit and miss stuff etc.

I’ll try to do my best for you guys, and hopefully this thread could help you become better than @ndy, Endi, Robertt, Martin Krol or Stephen Stahlberg :eyebrowlift:

My point of departure will be boxmodeling. Although many say that Blender is one of the best program (if not the best) for the poly by poly method (vertex by vertex). But I never could master this technique. On the other hand, I think that Blender is one of the best for box modeling. And although Blender lacks N-Gons, I think (and that is my stupid opinion) that this is rather a strong point for box modeling.

Why are triangles problematic? Should we learn to hate the triangle? The answer is: yes!
Although even the great Stephen Stahlberg once wrote that a triangle shouldn’t pose a problem, I disagree.


The first cube got a clean subsurf, the second one is the same cube with some of its faces triangulated. The subsurfed result is a mess.

So the lesson here is: unless you have a very (VERY) good reason, you should try to eliminate triangles.

If I had a euro cent for the every time someone asks how to eliminate a triangle, I’ll be a millionaire!

There are a few simple methods to achieve this:



The Solutions:


Select the edge until the start of the triangle==> Xkey, delete edgeloop.


Cut through here ==> then you are left with all those tris ==> convert them to quads: ALT+JKey

For cutting one may use the knife tool (KKey) or the way I like it: select the edges to be cut and hit WKey, Subdivide.


Now cut like shown above. The resulting triangles can be joined into a quad. If Alt+Jkey fails to convert it into a quad, just delete the internal edge, then select the edges of the ‘hole’ and hit Fkey to fill it.


This is basically the same procedure as above…

There are more methods to eliminate tris, but those are wacky… so I won’t go into those.

What method of eliminating tris is completely up to you and depending on the situation.

I guess that whenever you could introduce 1 or 3 tris adjacent to the offending triangle, you can convert them into quads.

Next I’ll be introducing the concept of poles and loops.

So what are poles? That is just a fancy word for a vertex (vert) that hasn’t 4 edges connected to it. The verts … excuse me , poles, that we are interested in are those who has 3 edges connect to it and those who has 5 edges connected to it. Of course make a 20 edged pole (pole of a UV sphere for example), but for the time being, just assume that it is right thing to avoid 6 or more edged poles.

I would like to call a 5 edged pole an E pole:


And a 3 edged pole a N pole:


Maybe this will make it clearer:


OK, now why would I mention poles? What are they used for? Why are they important for organic modeling?

Poles creates flows, they control and or conserve density of the mesh. It is almost impossible to create some nice model without any poles at all.

For example: this is a all quads poleless head:


This was my skill level years back: just subdivide a cube away and then with proportional edit, pushing and pulling verts. In this case I was left with a polycount of 1750 faces and an ugly model.

By mastering poles and loops you are able to build light models that are very detailed and suitable for animation.

Next I’ll be talking about loops.

my vote for a sticky. the tips u gave me (back when i was working on my head model) i haven’t stopped using and i never will.

Thanks Jason! I’m happy that I was, I am and I’ll be of some help. :wink:

A very important concept in modeling are loops. Now, there are 2 kinds of loops which are edge loops and face loops.

An edge loop is a line on the mesh that crosses exactly two edges. An edge loop stops at what I like to call terminators which is: hole, a pole and a triangle.





A face loop is very much related to edge loops, in fact I might go back to the images above to show the face loops. A face loop are adjacent faces in which the verts in those face are shared by only two faces. HUH?!?


In contrast, here you ave 3 face loops:


I believe that it should be clear now for the reader what a pole is and what a loop is. Next I’ll be talking about the different methods to create loops.

Loops can coexist. So in this way you don’t have to be afraid that there might be some dire consequences that loops you create might disrupt loops that are already present:


As you can see, loops can touch each other, intersect or be a part of each outer loop, which is cool. So this means that you can form your eyes loop, mouth loop and pull a nose loop that intersect the mouth (or mouth-nose tip loop) without a hassle. In the example above I showed only circular extrusion loops, but this property holds its ground for every kind of loop.

Thanks for the sticky :yes:

Loops are important when modeling in particular when you are doing organic modeling. An organic form should have smooth and sharp features alike and this all can be achieved with proper placement of loops.

From now on I will focus on face loops rather than edge loops. An edge loop exists because it is ‘drawn’ on a face loop. When explaining topology, it is far easier to comprehend the situation when you are looking at face loops.

When you create a loop, you are creating poles too. Poles will play a big role soon in organic modeling because they control the flow of the loops.

For now, there are 4 easy ways to create a loop:

  1. extrusion (EKey)
  2. spin quad/ spin edge (CTRL + EKey)
  3. rip (VKey)
  4. Knife tool (WKey) or related to it: Subdivide (WKey)


After you invoke the extrude command, you may pull a ‘limb’ out the mesh, or confirming the command right away, leaving a face loop on the mesh.



If you have a grid/ mesh and you perform a spin edge somewhere, 2 N poles and 2 E poles will be produced. It seems that in normal circumstances the number of E-poles is equal to the number of N-poles.


As you can see, after a spin, you are left with 2 loops like railroad tracks opposing each other. Notice that loops bend at N-poles. I’ll be explaining this effect extensively in the following posts.

The cool thing about the spin edge command is that you can bend the loops anyway you like. One major drawback is that each spin edge spawns a new loop.


As you can see that if you intent to use spin edge to create loops, you will get this rather nasty side effect.

But you can use spin edge in situations to eliminate poles (just reverse the procedure above by spinning the edge in the opposite direction) or to correct edge flow, which I will explain in a later stadium.

Remember the simple extrusion that leads to a closed loop? Well, you can make a closed loop too with spin edge and judge for yourself how much it differs from the extrusion method:

Here you have a closed edge loop, but with side loops bordering its corners.


What is mesh ripping? With this tool (VKey), you rip the mesh open by pulling at a vertex. Afterwards you should fill the hole it produces Select verts/ edges ==> FKey).


The mesh above was ripped open at the upper N-pole. So, like spin edge, rip mesh produces a pair of N-poles and E-poles. Like I said before, the direction of the face loop is determent by the N-pole.

Creating C-loops:

The rip mesh tool is very flexible in conjuction with the knife tool. Depending on the method, you are left with a single C-loop, or with a mirror pair of C-loops.

Creating a single loop:


And after the cut:


This was a very minimalistic loop because I ripped at only one vertex.

If you want to make a much wider loop, you must rip all the vertices in a row. You need to use the knife tool to obtain the face loop. Depending of how you cut the mesh, the results will vary.

To obtain a single broader face loop:


Here I’ll cut before I fill the hole, else there will be triangles in the corners


And then after filling the holes and cutting (and smoothing) you are left with one C-loop (by the way: I assume that C-loop means C shaped loop and not closed loop).


There is a number of ways to make a number of wacky face loops/ edge loops with this method, but i suggest to keep it simple because simplicity and predictability is the name of the game here.

Lets try to make a a closed loop like in a extrusion:

After filling the quads, you are left with a standard extrusion topology.


The knife tool is the topology tool of choice. It has one drawback tough: if you want to cut around a model, you can’t rotate the model and then keep cutting. That’s why I use subdivide often. It works best in “edge” mode. Select the edges which you want to cut through. Using subdivide you can cut around a model. I think the knife tool is well known and it don’t need further explanation. There is a pitfall though in using the knife tool, but I want to discuss this further when I get more in depth with poles.

I’m sure that there are more wacky methods to produces loops, but the above method should cover most situations.

Poles (in this, N-poles and E-poles) are created when you produce a loop. Everywhere I go I hear poles being bad mouthed. One example is that every seems to say to avoid E-poles because it causes pinching when you animate the region. I’m still waiting for the evidence…
Of course a 20 edged pole or something will pose a problem because the topology and geometry in the vicinity depends to much on that single pole.

E-poles and N-poles comes in pair, unless you delete one of course. The most important function of poles is to steer the flow of the loops. There are ways to move poles around to give your topology that distinctive organic look.

Poles are needed too to mold and form your model. For example, if you take a regular flat grid and you extrude a few faces out, it will look like a cube sticking out of a plane right? The E-pole is at the base of the grid, 4 edges are still part of the flat plane, and that extra 5th edge is used to elevate out of the plane.

To be a good organic modeler, it is imperative to know how to move poles around and how to unpole them. Unpoling are methods to make poles become regular 4-edged vertices’s again.

It is still all very theoretical, I know. But some very basic skills should be explained first before we get into practical examples. Next I’ll be explaining how to move poles around, and why it is important to position poles elsewhere.

We’re at the heart of organic modeling. This topic is so vast because of all the wacky possibilities there are to move the poles on the grid.

There are a few situations that I discovered which makes a model looks ‘organic’ (instead of hard or mechanic). I’ve seen many noobish models (especially in lizard and head models) that looks just weird. I believe that if you take heed to the following points, you can tweak an old noobish model into something professional looking. The goal is to make the loops blend in the mesh.

Before I go explaining how to move poles around, the following are the situations that makes a model look more organic:

  1. Use ‘C-loops’ as much as possible! A C-loop is nothing more than half a circular loop of course. So, most of the time this loop will start and end at a terminator of course. Most of the time a terminator will be a hole.

A C-loop (Duh!)


This is just a 5 minute lizard head model. But which do you think looks more organic?

This one?


Or this one?


Both models has roughly the same poly count (around 400), But the mesh of the second one is more evenly distributed. By just adding some C-loops here and there IMO the second one looks better, more organic with flows that blend smoothly.

  1. Try to avoid too much E-poles that are in direct contact to each other. Example:

Nice, but it doesn’t look organic… rather more mechanic

This is becoming to look better:

As for everything else, be the judge of your own particular situation. I guess, when you are modeling a mouth, I don’t think it should be really necessary to move the poles away. But hey…

  1. Spiral loops

This will really blend circular topology in your mesh. You can use spiral loops for the eye sockets. A model can look just fine with plain old circular loops, but spiral loops is the solution when you think that your model is in danger of looking mechanic. I guess that if you model a fantasy creature with a lot of bodily perturbations, spiral loops will surely help you out.


One thing to keep in mind when you get into edgeloop is the Spiral Loop. Once you break the extrude Loop you’ll going to get the Spiral effect and when you do, don’t try to rewire it so it goes in a circle because it won’t work.

  1. Diagonal topology. The modeling of muscular features used to be hard. But once you start using diagonal topology, you’ll start modeling muscles with ease.

For example: that big diagonal muscle across the neck (sternocleidusmastoid)


Excellent stuff! There’s a lot to learn from SomeArtist (to bad he isn’t active on the sub-d forums anymore.)

It’s great to see that we have such a magnificent resource native to BA. A summary is necessary. After all, the thread at subdivisionmodeling.com is over 12 pages long…

I completely forgot about the knife tool. I’ve been avoiding it since I first started Blender. I would always get a bunch of triangles in places that I didn’t want them (and didn’t know how to remove at that time). Subsurf modeling was a big confusing jumble at that time, but after reading through some of SomeArtist’s studies, things suddenly started to click.

Thank you and please continue to provide this great resource to the community!

More! please.

Moving poles around sounds like black magic, but in fact it is really easy. Sometimes it seems that there is no logic in this, but after this thread it will be clear that there are a few basic techniques that can be used in different combinations and situation. I leave the experimentation to you afterwards.

Starting point:


First the rip command (VKey):

A rip normally produces 2 E-poles. In this we have to rip the mesh (/vertex) at a specific location so you are not introducing an extra E-pole. The logical place to do that is where the N-pole is located at. But not only you’ll shift the E-pole, but the N-pole will come along for the ride. But this technique can only be done when there are no ‘fill loops’ present, in other word, the E-pole and N-pole are in direct contact.


As you can see, E-poles that once was on the same edge loop, are now on different edge loops.


In this way you can make a circular loop spiral.

In the method above the pole moves outward. In the following method the pole moves inward:



In the example above the poles shifted to the right. To shift the pole down, merge the other verts.

Spin edge:

After a spin edge you are left with 2 loops that are joined together. The possibilities to move poles around in this situation is numerous, but I’ll try to explain as much as possible.



n the image above there you can see those two loops that are joined at the hip. To seperate them you can spin the green of blue edges. Don’t spin the red edge, otherwise you are just reverting to the original situation (no loops) or if you spin twice both loops will flow in a oppisite direction.

The green loops are easy. When you spin it, the loop will seperate and drift diagonaly away. You may keep repeating untill the loop is at the end of the mesh. Or you can unpole it or collapse it to get rid of it.
Maybe it is not apperant in which direction you should spin this quad/ edge. Just visualize that you have to spin it so the edge will allign to the horizontal adges outside the loop. So in this scenario you should spin the upper right green edge once counterclockwise and the result is this:


The Blue edge should be spun counterclockwise. The end effect is that the loop drops one position/ row. But in between the spin quad action will look like a mess. Delete the lone vertex. What this realy is, is two quads sharing two edges. By deleting this vertex, and filling the hole up, you are converting these 2 quads into one quad.


Ofcourse after smoothing and stuff the result will be:

Indeed you might have noticed only a vertical shift. You can shift it verticaly too by spinning another edge. Like I said before, the combinations are numerous and you should experiment.

Maybe an easier way is like the second example is to merge verts (Alt+MKey)




did u finish the neck part? is there more to come with that or is that it?

No, it was just a illustration to prove my point. Don’t worry, in a few days I’ll be going down and dirty with modeling. Believe me, when I’m finished, you guys will be the one writing modeling tutorials and winning CGTALK front page awards :wink:

noice. thanks.


This is exactly what I’ve been looking for. Keep them coming! And thanks for the link to subdivisionmodeling.com. Excellent site.

This one is very tricky because it has to do with a fundamental flaw of the knife tool.


After this I’ll smooth it a little and select the center edge loop. The center edge loop is not really a continuous loop because it terminates at each N-pole.


Lets pull this edge out to form a nice vein or something

So it looks like a nice method to add detail or loops to a model right?
First of all remember that I’ve said how face loops got bent? They bend at a N-pole! All face loops layered after that one will get bent by the N-pole!

Let’s take a closer look at the cut mesh:

Observe how the face loop would get bent at the N-poles. Indeed there is a problem here: the location on where the face loop would bend alternates on both sides of the highlighted edge loop. This will result in multiple face loops running ammock on the surface of your model. When you try to use the edge loop cut you’ll see that the edge loop cut will snap unexpectedly on various location on the mesh.
Notice also that the bending side alternate at a moment that the cut changes direction on the clock. So if you keep cutting in a C-loop or closed loop (like a normal extrude) or keep spiraling with the cut tool, there won’t be any side effects. But as soon you make a S-like cut then you’ll have this problem.
So this explains a) Why nobody never experience problem while cutting simple loops (like eye loops, nose loops) early on when box modeling and b) that’s why you mesh gets messy if you want to cut details on your surface.

Here I’ll colour the face loops that are produced by this cut operation:

Things go wrong at the moment the the N-pole bends the loop on the other side of highlighted edge loop? Well, let us identify those N-poles that causes the problem.


Maybe the following image will make the problem clearer:


The method to flip it is to merge (ALT+MKey) this 3 vertices’s. Then you are left with 2 triangles that can be merged into a quad (Alt+JKey), and voila!




And the result is only one face loop. To pick up where the problems started: Let’s make a vein. Now we have to perform an edge loop cut.


Compare this image with the first image. They almost look the same, but this edge loop is truly continious. It doesn’t bump against poles along the way.

Now compare these two images and judge for yourself what looks better:

http://img266.imageshack.us/img266/3851/11diagonalloopwm9.png http://img405.imageshack.us/img405/9607/11diagonalgoodloopyx0.png

I reworked the jawline in the second image and the flow looks better now.

I guess there are a lot of creative ways to move poles around, but this should do for now…

Next I’ll be talking about removing poles.

Excellent! Keep rolling :slight_smile:

At first I held back since I didn’t want to post in a thread where a master is speaking and didn’t want to disturb him. Now, I can’t hold on any longer and I must say: a-m-a-z-i-n-g. Thanks so much for the work, it is quite an excellent read, with good imagery and explanations. One day I really want to get into organic modelling and I know where to start now.

Please keep it up!

a master is speaking
Woa WOa :eek:, I’m as good as the next noob. All talk and no action lol. I’m just taking another approach to modeling. Most tutorials are not explanatory enough. They never give the “why” or “how come”, only the “how to”. That’s why I think why tutorials as the Joan of Arc tutorial is flawed. In analogy, is like painting by numbers. Yes, anyone can paint a master piece… .by numbers. But to really produce your own stuff, we have to analyze this whole thing extensively.

By the way, it is a forum… so don’t hold back any comment, critique or corrections, most of the stuff I’ve written was made up as I went along. :smiley:

OK then, to kick off again, I call removing the poles “unpolling”. It’s a term that is made up by SomeArtist, and since he’s the first the have the guts to give it a name, I will use this term too.

Unpolling can be very tricky, in fact some times it is damn hard to remove a pole all together. Sometimes unpolling is very similar to moving poles around. I want to start with one of the most hated kind of poles: the 6 edged pole.

A six edged pole can be created by merging opposing verts together. You still have an all quad mesh.

http://img236.imageshack.us/img236/4126/12mergeto6jl2.png http://img61.imageshack.us/img61/6400/13opposingmp0.png

Anyway, the method to eliminate this 6 pole is by cutting the highlighted edges to begin with (stay away from the N pole).



http://img481.imageshack.us/img481/6827/14mergeoa8.png http://img264.imageshack.us/img264/792/15separate6ic1.png

When modeling you could end up with this 6 pole situation:


This especially happens often when working in mirror mode. If you don’t want such a situation in the first place, be sure to add a loop before adding flows. But if you’re stranded in this situation anyway, this is the way to break the 6 poles appart:

Cut it like this:




And E-poles can be removed like this:


After the cuts, you are left with 3 pairs of triangles that can be merged into 3 quads. The result looks a little funky, but if you eliminate another ADJECENT E pole, you are left with a C-loop


For this method to work, you need to delete all filling edge loops (XKey==>delete Edge Loop) so that the E pole and the N pole are in direct contact. It is possible to delete it if there are more edge loops in between the N-loop and E-loop, but with this method the E-pole will travel in the direction of the N-pole until they meet. But meanwhile, your mesh will get denser.

I wish I could tell you more about unpolling. It is not that easy to remove a pole.
I think the best mindset is this: ‘Provoke a triangle’.
What I mean by that is: when you cut a regular grid and stop somewhere in the middle, it produces 3 triangles. Hmm… gray matter working… 2 triangles can be joined into a quad right? So … 2 times 3 triangles is equal to 3 quads + poles. So… what if I provoke triangles by cutting just one edge of the pole? You’ll get a bunch of triangles. So can I cut somewhere to make the numbers of triangles even? The result is: All triangles can be converted to quads and the pole has moved.

For example:


Also, try to envision your target topology over your old topology and figure out which cuts will get you closer to your target. I’ve seen SomeArtist’s video clip where he tweaks the flow on a nose and it looks like magic, but ultimately he is just provoking triangles. He doesn’t even use the cut tool, just select edges then subdivide then join triangles.

Next, I’ll be introducing the T-loop.

This is a T-loop and this is a way to construct T-loops:


What’s so special about the T-loop is that the N-pole is not in direct contact with the E-pole. And what are T-loops good for may you ask?

Well, maybe you want to add more density locally on your model without messing your topology. Most people want to cut just one line in the middle of the mesh, but they keep getting a bunch of triangles… so with this method you can terminate a cut and still keep everything all quad.

It is handy to make folds on your model. But what I like best of it is, that it is an excellent way to add diagonal topology on your model. The topic of diagonal topology will be discussed in a later stadium.

Generally speaking, if you have that urge that you have to have a triangle somewhere, you should use the N-pole, since you can shape it in a triangular fashion, but is still a quad. And that is what makes T-loops so powerful: The triangle like quad is aligned on the grid.

Next I’ll be talking about the diamond loop.