Body topology


#2555

HI Shark,
The Idea is to use the poles to change the direction of the geometry, i find it makes for quicker rigging and less geometry then normal. The extra polygons tend to get used in extra details. The models below are supposed to be of game-ish quality, they have some details that a Normal map would take care of. I still put full geometry resolution on the ears for instance, yet manage to stay under 7K. I’ I’m debating i’ve been able to get some of my students understanding the techniques and some of them are consistently hitting lower polycounts then mine. I debated lowering their grades for showing me up like that…

Below are two different models just to give an idea on two different resolution targets:



The back for instance has alot of muscles and many people neglect it. but just following the structure gives a nice almost sculpted look.

The has so many muscles and this method helps with solidifying the cheek bone

The front also is able to get a nice cheat as well.




#2556

I’m sorry to say but the modeling paradigm that you seem to use has already been outdated five years ago. Nowdays computers can easily manage far, far more complex meshes which allow better deformations.
Even some of the leading games like Uncharted already use 20 to 80 thousand polygons for the characters. The head I’m building blendshapes right now is about 11.000 polygons alone, the entire character is above 350.000 (granted, most of it is hard surface and not organic).

If you really reach modeling and rigging than I also have to say that you owe your students to take the time and effort to get up to date with current industry practices. This kind of stuff wouldn’t really get them hired even at smaller studios, not to mention big movie VFX or animated feature houses…


#2557

Hi Tamas,
I tend to concentrate on how or why they should or shouldn’t do something and this has seemed to serve my students well. Most the modelling I’ve had to concentrate on are game engines (on decent machines) and locally people are looking for mobile to console games. Low polymodeling is more in demand. I am a big fan of higher poly craft myself, but if you look at the edge loops you can see they mimic anatomy as well as possible at that polygon count. The poling here is used mostly for anatomy and not polygonal reduction and localization like what you posted. Localizing geometry to add detail seems to be your biggest concern as it does allow for more articulation to occur. I assure you this models topology can be turned to a higher poly mesh with more anatomical detail, with cuts and bevels. But the base form of the body deforms pretty well, since it was used in a training video. Except the forearm, it’s one area, I feel

about.

It would be nice to push the boundaries in models in the 350K range but at some of the locations I taught it would crash the computers
yes…I know
I know. Hence the concentration on lower poly’s and building a solid technique that allows for expanding into higher levels of detail easy and more importantly easily explainable to beginners.

The image you posted earlier, I do like the heavy use of localization in the face. I am curious to know, so I can stay up on industry standards, how did you build it out? I usually don’t have the luxury of a maquette so I start depending on what the final resolution is allowed to be I trace out muscles and combine them into one object before hitting a computer.


#2558

There are two separate issues here IMHO.

First, if you’re working for games you will almost certainly use normal maps, or if it’s an even less capable platform then you’ll rely on pre-painted shading in the textures. In both cases the focus should be the silhouettes and not the small surface details which will be washed out by the shading in the end. Most game studios prefer evenly spaced polygons for this and tend to cover large areas with very few edges - they want to focus the poly detail on the highly curved parts instead.

Second, anatomy shouldn’t be treated as static, one always have to keep movement in mind with characters. And there’s a lot of movement under the skin here, clavicles and scapulas and rib cages move around a lot. The way you place your edges based on that single pose can not accomodate the range of movement, the vast changes in shape that a real living creature goes through. This is mostly why every studio prefers to work with a lot of even geometry because they want to separate the anatomical structures and the flexible skin that covers them.

Localizing geometry to add detail seems to be your biggest concern as it does allow for more articulation to occur. I assure you this models topology can be turned to a higher poly mesh with more anatomical detail, with cuts and bevels. But the base form of the body deforms pretty well, since it was used in a training video. Except the forearm, it’s one area, I feel

about.

Localizing is actually quite rare, we tend to remove the edges from the cranium to keep the model a bit lighter because that area is going to be covered with hair anyway. It does make the surface a bit bumpy so we don’t use it on other areas as much.

The models are actually quite similar to NURBS patch models of the previous decade, with the exception that we can place the poles at irregular intervals and we don’t have to pay computational and other costs to keep the patch edges stitched together, there are no cracks and we can UV map the model in whatever way we want, too.

It would be nice to push the boundaries in models in the 350K range but at some of the locations I taught it would crash the computers
yes…I know
I know.

There’s quite a range between 7,000 and 350,000 and I’m sure that there’s room for various compromises. I merely pointed out that staying that low isn’t a good idea nowadays. Also, your models are clearly subdivided which isn’t really possible in games.

The image you posted earlier, I do like the heavy use of localization in the face. I am curious to know, so I can stay up on industry standards, how did you build it out?

I’d like to note that I merely try to keep up myself, in no way do I consider our stuff to be at the front. We don’t have schedules and budgets to be the leaders in anything, we just try to learn and follow as best as we can.

Now the modeling method is simple - we start with concept sculpts for most organic stuff where an artist defines the forms and shapes, and then the modelers come in and build the geometry on top of that sculpt. Maya has Nex, 3ds max has Wrapit and the Graphite tools, there’s Topogun, 3D Coat, even Zbrush.
This way we can concentrate on the mesh itself as the artistic decisions have already been made, and the approach also makes it easy to completely rework any area without loosing the shape.

I see no reason not to start with Zbrush 3D concept sculpts any more, it just makes sense in every possible way. No wonder everyone’s doing it, even for hard surface with the new brushes in R4.


#2559

Tamas, as usual I love your response. To be honest, it was never quite clear to me before what your workflow was, so I was under the impression that you were using some kind of subdivision method and adding detail with complex edge flow.

It’s really nice to know your opinion of “best practices” in modeling. I’m going to take a look into resurfacing tools. Currently, I think that my best bet is learning good edge flow and modeling techniques. That way, when it comes time for me to use the sculpting method, I will have the skills necessary to create an accurate base mesh.

Do you agree with that?


#2560

Nice post Laa-yosh.

It’s so true. It’s hard to find a reason not to start in a 3d sculpting program anymore and then take the sculpt into topogun or something like that. And Normal maps and Displacement maps lend themselves much better to more a more even and squared out distribution of polys.


#2561

Tamas,

Thanks! I actually do appreciate the time that you’ve put into it, the explanation of your workflow is great and I may be able to use your example when I try to revamp the curriculum, it’s struggle to make some schools add and drop things from their curriculum. The first model I posted was supposed to be a blend a lot of low poly and high poly detail. it was built to meet a specific contract. That’s one of the reasons why it goes from low frequency to higher frequency areas and not strictly silhouette driven or denser. I am curious though when you say 7,000 polygons do you mean quads or tris? I’ve been given low poly counts and always in tris. So 7,000 tris would be 3500 quads.

Most the developers around me are working on android and IOS devices, so I’ve been catering more into that market (with the processors speed growing so quickly, I think 10K should be the norm soon any thoughts?). Also I’m wondering what your preference is in the 3D coat vs. Topogun war (avoiding NEX since plugins are harder to get admins to commit too). I am looking at getting one of them into the curriculum. The upper limit for the computers have average is 30,000 per scene (4 to 8 gigs of ram average)…

I’m really curious as to how to twist in more detail in without Zbrush seeing that one location doesn’t have any sculpting based programs. Hence the tracing of higher poly geometry to think about the topology. if you don’t mind… should I focus more on the high details or base anatomy? I’ve finally been given a chance to rewrite a modeling curriculum so anything anyone thinks I should consider from anyone would be helpful.

As for not being at the front, I think your a bit modest. Through the summer I think I showed the trailer Digic worked on at least five times.


#2562

Heh, thanks. Always keep in mind that it’s just one possible workflow, one that fits our needs and constrains the most.

I’m going to take a look into resurfacing tools. Currently, I think that my best bet is learning good edge flow and modeling techniques. That way, when it comes time for me to use the sculpting method, I will have the skills necessary to create an accurate base mesh.

Depends on what you want to do…

If you want to create nice stills, it’s mostly irrelevant. You can nowadays easily render multi-million polygon models in nearly every app, so a straight export from Zbrush will work. If there’s any small detail left out you can add it with normal mapping.

If you’re looking for a character modeling position it’s useful to be good at both sculpting and resurfacing, especially in the game industry - although they have somewhat different goals and methods. Like re-using basemeshes is far more common, in order to keep UVs and bone weighting.

Or you could try to specialize, but large VFX studios also prefer industry experience, it’s rare to hire beginners.

All in all I think every good organic character has to start in Zbrush or Mudbox (not counting the basemesh you start the sculpt from, but then again I’m not a fan of Zspheres).

So I guess the point is, there are no best practices, just a toolbox that you can search for the best fitting solution…


#2563

Yeah, that I know of… :wink: Feel free to use anything and let me know if I can help with anything else.

I am curious though when you say 7,000 polygons do you mean quads or tris? I’ve been given low poly counts and always in tris. So 7,000 tris would be 3500 quads.

Well, as I’ve probably already mentioned, it depends - game artists always count triangles, but others count polygons so one quad is still only one polygon.

Most the developers around me are working on android and IOS devices, so I’ve been catering more into that market

Yeah, that’s definitely different and you do need to optimize down to triangle levels. Is normal mapping available on those platforms? It really is an important aspect of both the visuals and the content creation.

Also I’m wondering what your preference is in the 3D coat vs. Topogun war (avoiding NEX since plugins are harder to get admins to commit too).

I’ve never really tried 3D Coat because it didn’t look interesting at the time I’ve been using Topogun (happily). I’ve only moved to Wrapit (3ds max tool) because it reduced the number of apps I had to keep track of - Zbrush and Maya were already enough of a burden and I still need to get up to date with Mudbox.
Still, 3D Coat has a lot more functionality as it’s also a sculpting and texture painting app with some innovative new tools, mainly because it’s privately owned.

I am looking at getting one of them into the curriculum. The upper limit for the computers have average is 30,000 per scene (4 to 8 gigs of ram average)…

That sounds very, very strange. 30,000 polys is a very low number, the 5-year old consoles can routinely push at least a million triangles per frame. But they render most of them at least 2-3 times for all the special rendering passes so they can actually process about a hundred million polygons per second easily.

I’m really curious as to how to twist in more detail in without Zbrush seeing that one location doesn’t have any sculpting based programs. Hence the tracing of higher poly geometry to think about the topology. if you don’t mind… should I focus more on the high details or base anatomy? I’ve finally been given a chance to rewrite a modeling curriculum so anything anyone thinks I should consider from anyone would be helpful.

It’s kinda confusing here… Typical workflow for games is to start with a high poly and bake at least the normal maps to the low poly from that. But many studios go as far as to texture the high res and this way they can tweak the ingame lowpoly as much as they want to without losing existing work. Zbrush (or Mudbox) is pretty much inevitable for most pipelines.
The only case where you don’t need such a sculpting app is if you can’t even use normal mapping but so far that’s only been the Nintendo DS and mobile phone game territory. However an iPhone 4 or most Android devices should be powerful enough for that, too.

Still, sculpting is so much more natural and easier to use for character work that it’s a must have in any school IMHO. Technology keeps on advancing and within 5 years we may not even have game consoles, just a docking station for our phones so that it can connect with the TV and the wireless controllers…

As for not being at the front, I think your a bit modest. Through the summer I think I showed the trailer Digic worked on at least five times.

Well, thanks :slight_smile: but I think we still have a long way to go with our characters and facial animation…


#2564

try to spread out some of your geometry to make it look a bit better when smoothed.


#2565

Hi guys this will be my first post to this thread. Just want to say I think that all of you are totally amazing. I would love to have some c&c. I’m really having difficulty with the anatomy and topology.


#2566

Hey mate,

there are a lot of ways you can go around things, depending on the purpose of the model, your skill level in modeling, and maybe also the skill level of your rigger.
enlighten us some more on these few areas, and our comments can be much more specific and suit your needs :slight_smile:

anyhow, this is what i would say right off the bat (and im assuming you want to do some generic animation with this):

-model your characters with arms at about 40-45angle, it will make it easier for the rigger to hit the right poses and not struggle too much with the shoulder region.
-model the hand slightly cupped. the reasoning behing these two comments has to do with rigging. - if a joint has a 90deg. of freedom on one axis, it doesnt make sense to model it at 0 deg and then worry about creating correctives for nice deformations, IF you can fix it by modeling it a little bent already and avoid correctives at all. This goes for the fingers mainly, especially for the thumb. always think about the movement of the part. how often do you lift your thumb so that it creates a plane with the rest of the hand? probably not that often :wink: more natural is it going ‘down’ from the palm. if you think about stuff like this, it’s gonna help you enromously in the next stages.

now for the topology itself:
-the abs are one area that you can do many ways, but yours feels kinda messy. if you wanna trace skeletal structures/muscles, do it right. - the ribcage could go lower, the abs loop doesn’t have to create a round U shape at the bottom…

-the breasts could defly use less polys to keep the model somewhat uniform in sense of detail. it won’t change the shape or very little, and will present a cleaner model.

-you could use more loops/definition at the elbow, for better deformations later on

-i see some weird pole on the inside of the forearm, just before the wrist, that doesn’t need to be there :slight_smile:

-you have a weird quad at the pelvis region. i understand you wanted to trace the external oblique or the pelvis, but you could get rid of it or end that loop better or even continue that loop down the leg.

-the back of the knee doesn’t need this amount of detail. if you want to stress the tendons, you can do a split down the edges. i gave my best to illustrate down in the pic ^^

anyhow, that’s my initial 2 cents.
always think about the movement of the body. when you bend over, there are wrinkles forming on your lower abs. how would you want the rigger to handle that right, if the topology doesn’t allow for it?
it’s this thinking ahead, that makes you a better modeller.

aaaaand im out. :slight_smile:

hope this is of use.

P.S. for some reason the CGSociety attachments and such doesn’t work for a change, so here’s a link ^^ http://postimage.org/image/ak0qxh20x/


#2567

nemeru - Thanx for a very helpful comment. i have spent some time reworking my model and ill be sharing my update as time goes by. my skill level is something only other people will be able to tell. Although I am a modeller by heart, my original mesh i made was to be used only for sculpting and then retopo in.

i have a long way to go with this model and i would like to eventually be able to impress someone i look up to soon(well at least that would be awesome if i can pull that off).

So here my update as off yet.


#2568

Haya wonderful people of CGS. have desided to do a whole retopo of my character’s body.

I wish to have some Crit or even paintovers from you guys.I’m kinda running dry on inspiration atm.

im having issues with the legs atm and im running dry of inspiration. please help.


#2569

Well, barring the pole above the knee, and the fact that the knee itself could do with an edge down the middle, it looks pretty ok to me.


#2570

You need to avoid triangle near area joint . If character for animation/cinematic and ll be with tourbosmooth over- need to avoid triangles everywhere. If it ingame character -in most of area truants is okay :slight_smile: also topology depends from which part of character will be animated


#2571

A model I’ve been working on recently. I started with a sculpt in ZBrush to figure out the proportions of the model, then I used Topogun to build different iterations of a low poly version and have been testing them with a walkcycle animation in Maya as I’ve been building a rig for it.

I mainly work as an animator/rigger but I trying to get better at modeling and studying animal anatomy too at the moment so that I can build complete creature rigs from scratch.

This mesh seems pretty good to me now as it’s deforming pretty well when animated, I’m interested to see if there are any notes on it though if anyone has some, I plan to bring it back to ZBrush next and start sculpting details and painting textures. My goal with this eventually is to have a good quality realistic creature that I can animate and render for a portfolio piece.

Cheers,
Brian


#2572

Hey guys, so I wanna chip in with a question:

do you have an idea about a resource, be it from movie, some making of, or personal experience, about good working topology (trying to avoid the word ‘correct’ since i don’t believe there is just one :slight_smile: ) for a muscle simulation rig? the model in question is a photorealistic male, and im just super-curious as to how people tackle this… The only nice working reference i found is the still-best-looking-muscle-rig-on-youtube by Adom Yip here

if you know of any such resource, im always on the lookout for another point of view on topology for photorealistic creatures/humans/quadrupeds, etc…

thanks in advance :slight_smile:


#2573

Hi nemeru!

Can this be a reference?
http://renpoo.cgsociety.org/gallery/835696/


#2574

hey renpoo!

thanks for your gallery, it definitely is another point of view!
i’ll give it a closer look when i get home from work.

could still use some behind the scenes of something tho :slight_smile: im curious how they deal with this in feature studios…