Does topology even matter ?

I have been working a lot with organic models lately, modelling aliens, animals, and humans

And i found that sometimes destroying the topology from nice quads in rows, into triangles, makes it able to add detail in a certian part, without affecting anythign else, especially with animals legs its useful

I am a pretty advanced blender artist in modelling, so i wanted to know the opion of other advanced artists
Is it worth the extra time having nice topo, or just break it when required ?

Until now i have had no problems with my models, even when animating triangles cause no trouble, not even with UV maps

I think it matters, but not as much as a lot of people make it out to.
Yes, it makes it nicer to unwrap objects with less distortion, and thinking about how you are laying out the polygons will leave you with a more efficient mesh, but some people take it to extremes.

In your case, since you aren’t running into issues, It sounds like there isn’t a problem.

and for mechanical modeling, forget about it. This is a part from a project I worked on a while ago:


I certainly could remesh that to an all-quad, manifold mesh, but it isn’t worth my time, and it certainly isn’t worth my client’s money (just try explaining why you logged an extra 6 hours on that project to fix the topology because it wasn’t ‘right’. you’ll get a lot of funny looks). If there was some crazy distortion effect, with ripples rolling across the face of that component, then it might be worth considering, but it’s a flat plane, it never changes shape.

Long story short, proper topology is good, only as long as it is helping you. if you are bending over backwards to try to do something with quads that you can accomplish just as effectively with a couple tris (or even an n-gon!), you might be wasting your time.

1 Like

Yes, when animating and smoothing using subdivision, topology does matter. That isn’t to say the entire model needs to be quads laid out perfectly along the form of the figure; only that you should try to hide areas where you need to break those rules somewhere less visible.

I think topology always matters, but what constitutes good or bad topology is dependent on what you’re doing. Good topology for a subsurf model and good topology for a low-poly model often looks very different. A triangle is bad in former, and often great in the latter.

Animation you want to avoid bad deformations. Subsurf you want to avoid pinching. If you’re baking normal maps you need to setup your topology to avoid shading errors. If you’re doing a dyntopo sculpt just for practice and only plan to render it – probably not much need to get nitpicky over the topology. It just depends on what you’re doing.

They go into it in fair detail here, Yes everything at render time is converted to tris. But those tris are still following the flow of the edgeloops that made them in the first place

Topology is what separates bad modellers from good modellers… or so I thought…

It is one of the many things that separates them. Speed, artistic talent, ability to match & modify concept as required, etc also play a large part. :wink:

All depend what you are after, render with smooth shading, render with flat shading, 3d print , animation etc. post #2 front panel is good flat example what you can do with only 1 curve object and forget topology.
Many times smooth shading put you a situation that topology matters (even if you dont need smooth shading there but you just set it on).

For some reason I feel like this could be the script for the intro of the next need for speed game.

I don’t know why

Topology still matters to a decent extent, but with new animation and deformation technologies like Delta Mush, it’s not as critical as it once was.

That’s not to say you should discard the idea of a quad-based mesh as many tools in both the modeling and texturing department depend on their presence.

Uh, what? Delta Mush won’t make bad topology animate any better. If an edge is missing or misaligned where a bend is required - it’s still missing/misaligned whether the deformation comes from Delta Mush, linear blend weights, or a cage deformer.

Everything else being equal perhaps. I took time to understand topology, because I’m a nerd. I’m still a poor modeller though!

Good topology is desirable. It makes unwrapping easier, it prevents shading errors, it makes for smoother deformations.

But, if you have enough experience to predict that a particular topology deficiency won’t cause problems down the road, and it makes your work easier or faster, by all means, go for bad topology.

For Efficiency I would say YES. There are people who know very well topology and do very good models, even using Tris but knowing where to put them to make a good flow. Even for non organic objetcs it does matter for a more soft looking or good shading.

In lowpoly games of today its even more important because the flow of whatever thing (muscles,etc) is defined by the topology even if its not very dense.

Some people are just blessed with a head for topology. I always admire modelers like Dan Platt if you go to his website the guys has some complicated shit modeled and a lot of it is 100% quad. Personally organics I can almost always get 100% quad but hard surface modeling is b*tch.

If I tried to model something like what Sterling posted with 100% quads it would drive me up the wall so I would end up using a lot of n-gons and triangles.

I did a bit of face rigging last year and my 2 cents is if you are going to be doing a lot shape key work spend time on nailing that topology. But even the pros can get it wrong just look at Victor from Cosmos Laundromat there is a bit of weird creasing every time he pulls an expression involving his eyebrows particularly brow squeezes. Shoulders for me would also be another place I would insist on having solid topology.

I missed that. Can you perhaps post a screen shot with some grease pencil strokes to illustrate this?

There are ways and tips to make good topology without being a genius at it. For example, if you look at the human body as an organic group of forms you will see that there are planes that are “together”, like deltoid, and pectoral muscle if you mark an arrow that goes above that plane and crosses above it you see there is no interruption, so good topology is about planes that are “non interrupting” whatever is the form of the thing you are doing and everything that goes outside this same flow is also known as edge loops and they meet at the point of disrruption, called e-poles.

An easy way to see it is just separating every muscle and organ, even the nose and the mouth completely, then try to join everything without breaking the outside boundary that makes the silhouette of every muscle/organ, instant epoles in proper places and instant flow/topology.

Well, sorry because I explain very badly in english.

[http://www.hippydrome.com/

http://www.hippydrome.com/Modeling.html](http://www.hippydrome.com/Modeling.html)

Isn’t this simply a case of ‘your mileage may vary’.
If your workflow and/or artistic style requires a particular style of topology then yes, topology matters. Otherwise no.

Some modelling tutorials recommend minimising/hiding tris and ngons but I’ve never read one said you should exclusively model with quads.

Ya’ gotta do what works for you.

Clear topology is still very important to have especially for deformable characters and assets. I would be cautious about getting into any habit’s of not worrying about topology.
Quads basically, but it’s normally ok to get away with a few tri’s here and there. I agree some people get a bit too obsessive and caught up on the tri thing.
Subdivision routines like to have clear quad based topology. If it’s non game pre rendered work all meshes ideally need to be optimized like this in some way while working in the viewport on most CG scenes .

One thing not mentioned yet is that good clear topology is just another essential part of having a clean pipeline. Keeping things clear to all, through all the area’s of production. It’s part of the mantra of keeping a good shop in order.
Normally keeping it all clean and tidy is one of the hardest things to control in a CG or animation pipeline when a large to moderate sized team are working on and sharing the same files and assets.
You have to always think of the next person along the chain. Working with messy meshes and other assets is horrible when you are dealing with complex projects and schedules. Also having a sightly casual attitude to any element of a production can cause big mishaps further down the line. The problems will usually just begin to mount. Messiness piles on top of more messiness and it becomes very hard to fix or change things later on.

A good topology also acts as an invaluable guide to how the model should deform. If the modeler or designer passes it the rigger then the rigger knows where best to position the joints and deformers. This even would apply if they were also using a skeleton mesh guide inside the skinned mesh.
It makes rigging and skinning much clearer to all and helps a lot in the modeling of shape keys or blend shapes, especially with faces.
In almost all modern studios the topology becomes an essential part of a finished characters design.
Apart from all of the above. A clear topology is just much nicer and better on the eye while working which I don’t think can be overstated either.