Skip to main content
Loading Explore guides…
Topology for 3D Modeling · Acedra
Home
Explore
Engineering
Topology for 3D Modeling
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
92 segments
Copy
Hide video
0:00
hey i'm jay and this is everything you'll ever need to know about topology
0:09
all right so we're gonna get super nerdy in this video all right so if you're into 3d modeling uh getting better at 3d modeling making characters that sort of thing stay tuned hopefully you find some useful info in this video if you're not really into those sort of things go into like a deep dive then this video is probably not going to be for you my goal is to make this video packed with all the useful information on topology address kind of the
0:35
most common questions misconceptions and mistakes that i've seen throughout my years modeling and teaching and trying to make a video that just encapsulates the whole everything you need to know about topology that i could send to people that you can send to people so i just want to be on my channel for all time an overview of what topology is how you can make it better and how it can help
0:56
you be a better modeler i've separated the video into chapters down below so you can skip around if you want to get info on a specific topic or refer to something in the future okay so first up what is topology seems like we should start there so let's define topology also referred to as edge flow topology is the term used for the arrangement of the verts edges and faces of a 3d model all of the verts of a 3d model are connected by edges and those edges are
1:27
like the scaffolding that make up the faces which is the actual shell of the model itself the faces of the model is what you see and the topology is like the web that holds everything together 3d artists starting out uh usually don't think enough about topology or on the other hand some artists think about it way too much kind of overthink it so we're going to try to describe what makes good topology what makes bad topology and try to talk
1:53
about the things you don't really need to worry about and the things that are best to focus on okay so why is topology important anyway like why do you need to care about it if you're a 3d modeler i think of three main reasons number one is the poly count the topology is the mesh so
2:12
it is what determines the poly count now i'm a character artist in the game industry that's like my day job so poly count is more important to me than other types of modeling let's say because it has to run at real time and with all the other assets in a game so it's something that i think about maybe a little bit more so than if you're modeling for a different reason but polycount's
2:35
important no matter what kind of thing your model is for and we'll talk a little bit more about that but i think number one is polycount you don't want to have a bunch of faces that aren't doing anything you know you don't want to be wasteful and you need to have enough to make your model look good so it's all about this balance which we'll talk more about too of of getting the model right to do what it needs to do and also fit under your budget for whatever you're doing
3:03
okay so number one poly count okay second reason why topology is really important is deformation so especially if you're making characters characters are meant to move right they're meant to perform so the topology is the actual mesh that's animating your final product that you make as a modeler so not only do you have like simple joints which are pretty straightforward but you have things like a face you know that can make all these like squishy squashy poses and so it's
3:30
really important that you build your mesh in a way that does everything that it needs to do which is like renders correctly it's nice and smooth and then can be manipulated and deformed by technical animators and animators to give a performance you know as a modeler that's kind of the ultimate is that you can make a mesh that when collaborating with others can come to life and give a performance so it's important that your topology can deform well in a believable way in a
3:58
naturalistic way to do what your model is meant to do i don't want to make it sound harder than it is we're going to talk about just a few of the key principles you need to keep in mind when you are building a mesh that's meant to animate and deform so we'll look at some examples a little bit further in the video again don't want to make it sound like it's you know super hard it's just a couple things you need to be keeping in mind and pay attention to
4:21
when you're building those meshes okay third reason maybe the most important reason it's how it looks the topology that you build is your mesh so how smooth it is uh how detailed it is whether or not the you know maybe a parts of a costume are detailed and so the modeling there is what will make it look realistic if you're making a game model it's whether or not the silhouette looks jagged or smooth you know how it's shading how it's rendering what is actually
4:50
3d what is texture so this is all going to be part of like a balance that you're considering as you build your model for whatever it's doing so having a good understanding of topology is going to help you to make better meshes that are the correct poly count and under budget can deform well and animate and look good when they're rendered nice and smooth okay now let's talk about a few of the most common ways that you will be making meshes thus making topology
5:19
generating topology and kind of the stages of how involved you have to be and how specific topology would need to be and for what reason okay so let's talk about making topology all right so there's two main ways to make topology that's procedurally and manually so let's first talk about procedurally so procedurally obviously automatic that's its strength it's doing a lot for you a lot of people may not be thinking of it this way but dynamesh and z remesher
5:47
are procedural and they're fast and that's the strength is you're giving up control for speed and this makes it quick to iterate that's why this is awesome and used all at the beginning so you can use it to make base meshes even more refined meshes with z-remesher something that's really good is with z-remesher as opposed to dynamesh is you can give it some control so dynamesh is just throwing topology at your meshes which again is fine for like building quick shapes but then
6:15
typically i'll be transitioning to z-remesher it gives better topology for the shape whereas dynamesh is totally agnostic of the shape so zero mesh will help support the shape you'll get better topology that way and you can control it a little bit you know you can have poly groups and say keep these groups and kind of control it in that way you can also kind of manipulate the mesh
6:37
because it's a little bit predictable in that it's looking for strong edges and things like that so still doing a lot for you but you have a little bit of control in some of the use cases of z remesher but it's not optimal that's the downside if there is one but it's used all the time also i'm using it for things like decimating you know with decimation master if i want to export high polys out for references or even for baking and also procedural
7:04
topology is used for lods in games that's something that we used to have to do by hand but now it's super common to just use some kind of a third party tool like simply gone or something like that and do your reductions of your game characters procedurally it's way faster since it's far away from camera you don't need as much control as you do on the lod0 the main model okay so that's procedural now let's talk a little bit about manually doing it
7:32
obviously manually meaning doing it by hand so typically that means drawing or creating a mesh on top of your source mesh this is also referred to as re-topology probably seen this like in maya you can use quadra so maya comes with a tool made for this purpose and then there are third-party software like topogun that are it's an entire software made for this practice so you get like extra bells and whistles uh using topogun but you have to pay you know separate costs it's a
8:01
separate piece of software so if you're serious about retopology you're doing a lot of work uh it makes sense but for the most part you know you don't need that i also know blender can do re topology but this is generally the practice you're drawing your polygons on top of some sort of live surface of your source mesh that you made and you probably made that source mesh with dynamesh and or z-remesher okay so why would you do this it gives you the most control you're
8:30
making polygon by polygon in some cases you know you're adding and removing and editing every edge so this means for production models it's essential still all series productions essentially do this unless you know for some reason they can be wasteful and and in some cases maybe you could do that but for games every polygon counts really that's becoming a little bit less the case but still you know it's not very common to procedurally generate meshes
8:59
that are going to be used in some kind of game production so the downside to this obviously time consuming the most time consuming way to make topology but gives you the most control so it's the most optimal mesh usually right it's the best for deformation it's just the right amount of polygons that supports your shape without being too wasteful so the silhouette is clean and it can animate and you're left with an optimal amount of polygons which you wouldn't be able to
9:25
as easily get with procedural modeling so those are the common ways of making meshes from auto generating it and just not caring about it that much because it's all about the shape to the far end where it is about the topology so if you're making a real-time mesh that's meant to deform and animate the best way is to do it by hand because it gives you the most control so you're going to make an optimal mesh it animates as good as it can and it's like as close to the
9:52
budget as possible and that's how you get the most efficient mesh that does as much as it can do and we'll do a little quick example of of what that would look like in practice okay so how do you determine a polycount that's a tricky one uh probably the most common question that i get and that's you know unfortunately impossible to answer the real answer is it depends which uh you know is
10:18
is kind of unsatisfactory uh and understandably so so let's talk a little bit about what it depends on and how you can better determine what a optimum polycount is for your particular model okay so the polycount how can you determine a polycount for your model it's very situational uh so there's not going to be a one size fits all but this is the overall strategy now in an efficient model you
10:47
know one that's built optimally and isn't super wasteful but one specific like game characters or efficient rendered models the topology is really serving three major important things and that's the deformation the silhouette and the detail of this model the deformation you know can it animate uh believably does it look good when it's moving around does it have enough loops to do that okay check silhouette is it jagged or is it smooth enough to be believable the textures
11:15
are going to take care of the inside details but does the model overall have enough to support the three-dimensional look of the character and in a way that isn't noticeably uh faceted okay and then the detail does the model or character have some costume elements or props or weapons or little details like glasses or jewelry or something that require in a disproportionate amount of polygons so that it looks believable okay rather than talk about numbers we're going
11:48
to think about this as proportional all right since every model is going to be different but if you think of the proportions of what the overall polycount is it might look something like this deformation is uh priority maybe priority number one but doesn't take up a majority of the count because it's just are there enough loops on the joints that doesn't take a large amount it's just placed properly now silhouette is going to take a little bit more depending on your
12:19
character too maybe it's got a lot of hair maybe it's got round elements and armor or something and then detail could take a disproportionate amount so let's say your polygons are spread out like this well we have a budget so our overall poly count might come in high if you don't have a budget you have to set a budget per game per character everything is so different um my general suggestions to people is if you're making a character for a portfolio then i would do
12:46
it anywhere between 50 000 and 200 000 triangles i know that's a big range because um why not i mean every year it gets better and faster if you make a character between one and two hundred thousand triangles it should look pretty good and so that's generally what i would recommend to people if you're making a certain type of character that's for other games like you know an overwatch type
13:10
of thing a fortnight type of style then you don't need you know 200 000 would be wasteful really a lot of the poly count will come from the hair or how smooth everything looks and that's really for certain you know particular types of characters so you don't really always need that many so that's why i give such a broad range so it's kind of depending on the style of character that you're doing and you try to get it within that range and then it would be reasonable uh
13:36
in a portfolio in my opinion all right so that's my most general answer to what a poly count should be so you can kind of think of it as your overall poly count is like a point system that you're allocating to the most important parts of the model and that's what's going to determine where you start placing these points then you're going to have to start taking away to get our overall budget down now you don't want to take them away from deformation first like i say
14:00
that's a priority you want to be able to animate that's the whole point and the silhouette needs to still look pretty good so what are you going to take the most from it's going to be detail and you're going to take away topology in places that aren't focal points you know like below the belt and shoes and things like that it maybe they have some props on their back so it depends on the game and the camera but you're gonna start where the camera doesn't see it most often and
14:25
you're gonna start taking away tiny details that would be nice to have so when that happens then we start to get below the budget and now maybe our ratio looks like this but this is how you're going to finalize your poly count you're not just going to build your low poly and then it's done you're going to build your low poly in a reasonable way try to make as best predictions as you can but ultimately you're there's going to be like a cutting phase at the end when you're really
14:51
trying to get specific so with experience comes kind of a general understanding of how high a mesh will be as you're making it but you're almost always going to come in over and that's really what you want because it's easier to remove than to add so it's even good to come in a little over and bake it because the textures will line up to high poly a lot better and then you cut it to get it down below budget whether you're making an object or a
15:20
character that's meant to deform or you're making a character for a game or a cinematic or something you're going to make different decisions on your topology so let's talk about the difference between making a character for offline rendering cinematics movies things like that and then games so we'll continue on with games for a little bit since we're already talking about it in games it's
15:40
typically going to be lower obviously and when it's actually rendered in the engine with the gpu the entire model is thought of vertices and triangles so that's why in games we measure the poly polycount in triangles that's also why it's totally normal to use triangles when you make your game characters now you don't want to triangulate your game character that would make it really difficult to uv map to select edge loops and edit the model so we still are typically
16:10
modeling in mostly quads but because it's going to be triangles anyway it's really common that you'll use triangles to end edges to make things cheaper because obviously that's the name of the game when it comes to making real-time art but also you'll sometimes add a triangle yourself so that you know exactly how a polygon will be split because if you leave it up to the renderer it might rotate the edge in a direction that you don't necessarily want so sometimes you are adding
16:37
triangles to make sure you're getting the right shape that you want now with models for offline rendering cinematics and all that stuff it's typical that the mesh will subdivide on render or at least what is being rigged and what's being skinned and animated and modeled is not the final polycount version it'll be subdivided okay so it's really important that the mesh is quadded out because when you subdivide a quad you get four polygons and it's very predictable so if you're
17:09
doing animation on something that's made with quads and you subdivide it it'll look more smooth triangles don't subdivide very well which in game's case is fine because it's not ever going to be subdivided but in the case of offline rendering it's not going to be great it's going to start to make the mesh messy it's going to add like pinches and stuff and because you're not having this perfect square that's subdivided then things like your displacement map aren't going to move the
17:38
vertices in a predictable way that you're going to want so it just causes problems so if triangles are used on a mesh for offline rendering they just hide them you know put them in the bottom of the feed or between the fingers or behind the ears or whatever something like that you definitely don't want to have triangles on all the squishy parts that move around so that's why we measure
17:57
triangles for game characters whereas for offline rendering and you know like how high you know how many polygons is your high poly model that that type of stuff we just typically are counting polygons which uh are usually quads so let's talk about high poly models that is the beginning stage of almost everything most of the time you're going to spend as a model as making high poly
18:20
models one of the things about high poly models is really the poly count doesn't matter what you need from a high poly model is just enough polygons to hold the detail that you need for that specific model that you're making and no more and it needs to be able to run in your machine in a reasonable you know amount of time so that you can sculpt on it and that's it so the number doesn't matter it
18:43
could be 5 million it could be 10 million you know doesn't matter just make it as high as you need don't go higher because it's just wasteful it's going to make everything slower so if you don't need tiny little pores then you don't need it and if you have a jacket or something you don't need you know 20 million polygons high poly models are made for the shape and for sculpting and that's it it doesn't have anything to do with deformation because this is
19:09
about high poly models this is about a source mesh that you will derive production meshes and textures from one of the most common ways to make final meshes that become these high poly models is to use zebra mesher so we talked about manually doing topology for production ready meshes when optimal poly count is important well in high poly models optimal poly count isn't important what's important is the shape and the sculpting right so z-remesher is made for this purpose it makes
19:37
geometry quickly that takes into account the shape and does its best to evenly distribute polygons on it okay last thing about high poly models is subdivisions use subdivisions none of this super high resolution dynamesh stuff you want to use subdivisions so that you have the most control it follows the principles that we talked about before which is having topology that reinforces the form that you're after helps enable you to create clean polished forms and surfaces
20:08
more easily without the resistance of having just that generic topology being able to subdivide lets you get up and start to create details and at any time you can drop all the way down and work on primary forms subdivisions can also serve as a proxy in zbrush so the entire scene moves faster too so those are the main differences in topology for high poly models now we can look at more production models all right now let's look at some examples
20:36
all right so some of you may know i recently published my base meshes that i made for my own work so that people that wanted to use those or just buy a base mesh can get them and so because of that i've got my own base meshes my own topology that i spent a long time on so i thought let's use it as an example for this so i'll just briefly touch on a few of the principles that we discussed and i'll show you how i implemented them and then i'll show you a couple i guess unique-ish
21:09
things that i ended up doing to my bass mesh and i can explain the reasons why that might give you some idea on something that you might want to do in your model so important off the top to try and convey what i wanted to accomplish with this base mesh is that it would be
21:26
enough to render in real time and look pretty good but then also i can subdivide it and sculpt on it it's my preferred way of working on organic characters if i can i'd like to make the high poly model or the subdivision level one the actual real-time mesh that way they're connected so i'm only doing the topology once so it's kind of serving both purposes as the render mesh and the base mesh so when i can do that i will and that's what i'm doing with this this will be what
21:56
i subdivide and then work on so because of that you know down here i'm making the polygons pretty evenly spaced like a grid now this helps with sculpting because you know just like pixels in a canvas you want them to be similar size so like if i sculpt veins and wrinkles then i have a similar resolution over the whole hand as i subdivide so same is true with the face though i can't make it all even because just how faces are pretty complicated you can see like because
22:30
of deformation it makes more sense for me to have more loops around the eye and the mouth so we're starting to get rectangles you know rather than like square polygons which is okay but it's just something that is worth noting i think that's fine but that's why i have some more faces than i need but in general we're getting to a space now where real-time meshes don't have to be you know super
22:55
cheap so i still don't think this is very high but this definitely could render and you know if i showed you without any lighting whatsoever it still looks pretty smooth you know for the most part the head looks smooth which is an important part the arms look pretty smooth uh the fingers look okay so it doesn't look stand outish you know it doesn't look like really polygonal so that's
23:18
good you can see little jaggies here if you really look but for the most part um this is my kind of minimum for uh most things now i could also use this for some kind of pre-rendered i mean i could easily just smooth this once and get four times as many and now all of a sudden this is going to be totally fine more than i need so depending on my purpose i can do that so i want it to be as versatile as possible okay so other than the amount of polygons and the size you can see
23:46
the general flow here is pretty easy right this is all horizontal because she can bend back and forth same with the elbow you can see it's already getting a little bit of the bend in there now on the joints that bend like this that just bend in one direction pretty easy way to think of it is just you just want to make it just like a grid just like a bendy straw you know you want to have
24:14
some edges and as it bends you want to be able to have more edges so that when it does bend it still looks smooth if you didn't have that many edges then all of a sudden when it's starting to bend you know it starts to look sharper and that's not what you want when you see these facets right here even now it's getting a little pointed on the elbow but i think that's fine it is an elbow i even have an extra loop here to surround the elbow bone so that when she does a full extension
24:43
hopefully this shape can still be supported the chest you may notice isn't as round as it could be that's fine with me because i did make this base mesh universal this is the same topology for my male shape as my female shape that we're looking at right here and i don't plan on ever doing a character that is 100 topless or anything like that that would be a custom model you know the fact that she would be wearing some kind of costumer clothing
25:09
i could always rip the arms off if she was fully covered so i'm not even using this body mesh other than maybe a shape to make clothing on top of so that's not really a problem at all all right so now i'm just going to point out a few you know somewhat unique things that i did to my bass mesh and i can give you a couple reasons why i did that first we can just start with the head is if you notice i have some stars here stars are vertices that have five edges coming from them
25:35
uh typically it's indicative of a direction change it's just a lot of polygons and so there's like you know the flow is kind of turning so you do need to have stars for the most part if you're going to have you know a mesh that has quads this has quads everywhere game meshes are going to be triangles right so that's not a problem if you use triangles but because this mesh is meant to be versatile when i'm sculpting on it i don't want triangles and if i'm doing subdivisions on
26:04
render for displacement maps i don't really want triangles so why not do quads it's also easier to make selections and uv map and stuff like that but here you can see these are all direction changes here's some more stars these are all direction changes uh without using triangles so back to her neck so yeah you can see some stars here now this is because i have this head split like this
26:26
i wanted to be able to have the head with enough room to have neck holes and stuff if i wanted to detach the head and still have a little bit of the chest so i prefer doing it this way rather than having concentric circles go down lower and lower and lower and also the shoulder right here has its own lines so that she can raise her arm right and then this gets to compress and then this stretches out right here on the underside so i'm getting the lines here for deformation
26:55
i'm getting the topology here that's gonna flow from her chest over to the top of her deltoid again when she lifts her arm up hopefully it keeps that shape that way you can keep the bump here and you can even see on the underside i actually just used it to reinforce the armpit shape so as she moves around hopefully this will still be able to deform the shoulders are a big point of deformation and stretching in characters so really a lot goes on the technical animator being able to
27:24
have like supporting bones and good skin waiting and stuff like that but this is just a way that i reinforced it with some extra geo and kept the flow going without using triangles so you can see here if you pay attention that you can see like some more stars star right here start right here this isn't necessarily a star but it's a direction change so let's take a look at some of the stars
27:44
on the face so with these stars or poles uh a vert with five edges it can cause some tension in the mesh visibly actually when you're smoothing it you know it'll just feel sharp so it's not going to be as smooth as like this area right here when it's subdivided it's going to be super smooth not a problem at all but you know stars like this could and when you're sculpting it it's not as easy to get the exact shape that you want so when making faces it's best practice to
28:16
strategically place where you're putting these stars so i'll explain why i chose where i chose this is just a direction change not really a star but you can see like i wanted a forehead with lines like this so that it can compress so she can like raise her eyebrows and furrow them so it's best if the forehead is a grid in my opinion and then so i just have this you know flow come around so that it can meet up with the eyes which are concentric circles concentric rings which is
28:46
good same with the mouth the mouth and the eye are both very similar in that they are rings going outward so it can do like the squashing and stretching and stuff and it can look around the nose is really just about holding the shape there's some deformation here right here you can see the flow come around so instead of the concentric rings from the mouth we end here and now we're coming up over the nose so this is like if you raise your cheeks and stuff you can create
29:12
these shapes so you have to have these polygons and able to make some of these shapes if you're going to do blend shapes and things like that so it's important that you have the edge flow echo the facial shapes or the direction a face can slide around on top of that it's extra geo here that's maintaining the shape of the head looking good while it's rendered in all angles so it's serving dual purposes so where i place these poles right here and right here and right here and right
29:37
here now i put them here because i heard a good tip a while ago that if you have to put a pole somewhere on a face it's good to put it in an area that's that doesn't deform that much or already is somewhat hard you know by that i mean like a bone or something so here we have this temple doesn't move that much either i think it's a perfect place for a direction change this just trying to find a space in between like like i said like here's a section that deforms
30:07
right and then here is a section that deforms this is like what would make the crow's feet and like if you're doing a big smile and then we have the jaw which can open so this area right here can swing open so it's a good place to put it because you have these different sections that are kind of moving in different ways in their in their major directions but then also i kind of think of
30:30
a place right here is kind of a good spot because not it's really just floating around i'm not doing a lot of crazy detail right there so for me it made sense to put it right there right here i'm thinking maybe kind of like the orbital bone again like we have the squashy eyelids around the eyes and then we have like a cheekbone and it's coming in different directions so i feel like this is a
30:53
fine place to put it if there's a little bit of a corner on it i think that's okay most of this stuff by the way not that perceivable this is just the super like geeky stuff that you'll get into once you're once you've made several base meshes and you're just trying to get the perfect mesh something i think is probably worth pointing out and calling out here is this kind of structure here i wouldn't do this if i was doing a mesh purely for sculpting purposes
31:19
and i wouldn't do this if it was for offline rendering only i really only added this here for facial animation at this resolution if i were to use this as a real-time mesh and that's to get the squash of the brow so that we can move the brows in this direction and get this to to kind of compress and make a wrinkle right here other than that i would have just like had the lines come up you know that would be the simplest way but i put it there again you can see some stars or
31:48
some poles and those i feel like are the perfect place to hide these because there's going to be eyebrows right here so you can almost never see them when they're placed right here i've seen them placed in other places if they're too high you know might be a problem if they're up here so i always try to put them right here if i can same with the ones down here again like the nose isn't moving too much it's kind of like a a solid structure right there so that's why i put those
32:12
there as well so you can see the the line flowing underneath the structure of the neck uh the ear the topology if you if you notice some poles again and what that is kind of doing is making a kind of a section here so the ears it's like its own island and that's how i model it so that i can have topology that supports the structure of the ear the ear doesn't animate so it's purely for the form of it so really what i'm trying to do is stitch together these sections
32:42
that i know have to be a certain way you know and then i'm getting like even flow in between and then i'm trying to stitch them together and finding ways to like route edge flow if i need to if i have some extra edges or something but the ear kind of a good example of of edge flow that's purely for the model and the structure of it and not for deformation at all all right so this is my base mesh and uh some of the reasons why i did some of the things i did now let's take a look at
33:08
a couple of examples of meshes that i think are not too good okay so here is the demo head from zbrush this comes with zbrush it comes with a couple base mesh tools that you can use to get started so i thought we could look in here and see if we could pick some things that could be better overall it's not too bad i don't think um the main things that stand out to me are the the neck i mean this one kind of stands out uh obviously they put the poles in different places
33:39
which is fine but there's like a the direction of the flow you know is is not supporting the shape that much it's being stretched out here to conform to where they want you know essentially to conform to this part probably so i think that could be better it's actually kind of going against the the flow of the tendon of the neck and the muscles so i would do that a little bit differently you can
34:01
see we do have some concentric rings here around the mouth which is good to see this is the best way to do a mouth i might go a little bit more but this is uh this is not so bad uh then we have this star here that actually has like one two three four five six seven edges i think that's way too much stress on this one vert so you can even see like the flow how it takes a sharp turn and the difference between this face and this face you know it's really big so
34:30
i think i would probably add some more topology here and change the flow there's even a change in direction here on the bridge of the nose i assume to support the bridge of the nose but i think we could still do that by having like the edge rings going over the nose it would also help with animation and deformation like if you're doing uh snarls raising the nose squashing that kind of thing like intense frowns and stuff right now we don't really have topology that would reinforce
34:57
shapes like that it looks like it's built more for the static shape than it is for if it deformed even right here putting this star right here isn't so good because we want to have this shape right here of the nasolabial fold so that when the face does this you can get that crease you can get that compression of the forms of the cheek and everything you know i don't know is this helping but you know what i'm saying so i don't think that's a great place for a star and i wouldn't
35:24
want to do a vert that has so many edges anyways definitely want to loosen that up and just get the the flow a little bit nicer like this and probably just cross the nose uh the ears okay doesn't have a lot of the internal structure supported probably do that in sculpt though but yeah so that's this mesh then let's take a look at one that i think is worse all right so this is the julie mesh
35:45
from zbrush also comes free i think this has subdivision so it's probably you know you probably wouldn't want to start with this one i mean you could always dynamesh it but you can see here how the ear is just totally botched you can tell that the topology was just made simply and the ear was tried to was like attempted to be built out of that so in this kind of case you'd have to subdivide a lot in zbrush to get enough topology in here to make the ear shapes and you'd have to
36:13
fight the topology a bit to like tuck it and make those convex forms so by not having the topology like reinforce the shape be built custom for an ear it's gonna be a lot more difficult to sculpt and model and obviously you lose it completely unless you have a very high subdivision level so i think that's no bueno and then we can see the mouth here has some problems first right but but also we we don't have those concentric rings to make like a kiss face or like a big smile
36:43
right here the edge flow just comes off this way and it would really wouldn't support like you know this kind of cheek shape you wouldn't be able to get the puffy cheeks of a smile making like a kissy face would be almost impossible so this is an example of topology that deformation would not be good it definitely was not made with deformation in mind and it looks like it was made probably on for high poly models it was probably made so that you
37:09
could sculpt a static shape but i still think you could make topology for this purpose a lot better if the structure of the mesh supported the shape it was trying to be more like in the ears case and like in the noses case you know by not having the nostril like reinforced with edge loops then you have to sculpt everything and you're going to be fighting against this grid like shape rather than an actual like circular shape that's reinforced with the topology itself all right so
37:38
there's a couple examples of topology i think are not so good and while we're on the subject i want to recommend a couple books if you want to get more deep into topology for animation and of faces there's the book stop staring by jason osipa which i've recommended before i've read that book and i like it a lot not only does it talk about good techniques and principles for edge flow for facial
38:01
animation but also it talks about modeling a face uh modeling a bass mesh for a face so i recommend that book to any character artist starting out or someone that's interested in making characters for facial animation another book that you can get which is a somewhat famous uh ebook is this one the art of moving points by brian tyndall so in this book it's a you know it's a digital book like
38:24
i say but um it's even got some videos let's see if we'll see a video here oh here we go so here's a topology laid out you can see these um sections the concentric rings that we're talking about and they describe the sections and what they do and why you'd want to build them that way so great reference this book is really good was recommended to me by an animator and you can see uh because
38:45
it's a digital book there's even some videos which i can't play right now but trust me there's videos all right i want to take a moment to thank this video sponsor skillshare skillshare is an awesome online learning resource with tons of videos on all different kinds of subjects so if you're interested in learning stuff and you've got an interest in anything that you want to know more about skillshare is a great place to check it out i watch all kinds of different videos that are
39:12
you know that are somewhat related to my interest somewhat not i'm just curious and i like to uh look into stuff one of those things is this logo design with aaron draplin i've seen his other videos on the site and even though i don't you know think i'm a graphic designer i love hearing about graphic design and his passion i also love the way the videos are made but there's a ton of different subjects on skillshare that you could check out for yourself so really you
39:38
could type in anything that you want and and see if there's something that you're interested in to learn more about the first thousand people to click the link down below in my description get a one month free trial to skillshare so you can check this out for yourself take a few classes and learn something for free oh cinematic color grading dope so that is as much about topology as i could think to put in this video for now so thank you for
40:03
watching up to this point i hope you learned something something that you could apply to your own models i'll have an art related video coming out pretty soon so stay tuned for that but thank you for watching and go out there and make some cool models until next time peace out you
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat