Modelling Heads

#362

Allenatl, yeah, agreed, they’re both great threads but aren’t that easy to find. Don’t know how I discovered Stahlberg’s (Maya) Body Topology thread as I barely have time to get round the Lightwave forums. The box modelling technique you posted looks really intuitive - they are invaluable pics, and much appreciated. Keep posting any progress.

Checked out your site - the female models are all really good and the spline modelling is especially well done.

It’s a struggle to make polygonal faces look lifelike and invest them with character. Here’s a model I made a while back and posted on the Newtek forum previously, though still have a lot of work to do on the surfacing…

#363

I’m not trying to start a fight here but you’re certainly generalizing a lot. Most point by point models you have seen? I would definately assert that anything modeled by the box method could certainly modeled point by point becase you are working at the point level. Working with as little as two points, not to mention building polygons manually, ther’es no shape or flow that could not be achieved. While I would agree that point by point modeling may not be the funnest method, I would say it is very dependable.
You keep talking about building for animation yet I don’t think your head or Chris’s for that matter fits that description. Looking at your head model, I don’t see how you could achieve realistic smile lines, SCM tendons, or even convincing eyebrow animation with only one loop aroiund the eye like that. Let’s see some endomorphs.
You also suggest learning about anatomy. I’m not sure what good that will do unless you’re willing to use it.

#364

That’s exactly the opposite. If you model something that correctly follows the facial muscles it will lead to a good animation result for sure. That’s the main reason topology is important for.

Also your example doesn’t look like it’s animation ready at all, the chin muscles aren’t animatable at all, the mouth and the eyes don’t even have a single animatable loop and the topology of the cheeks doesn’t help as well, not to mention the impossibility to animate any neck muscles, basically in your example there aren’t any. Once again topology and facial structure ARE something you have to worry about, they’re the most important thing to make an animation ready head.

Once again i disagree. One can’t start modeling a head if he doesn’t understand or know the rules he has to follow in order to do it, and topology is the first rule to make a realistic animatable head. Some characteristics and facial features are IMPOSSIBLE to create if the topology doesn’t help them: the smile lines, the subtle folds you got in your chin when you smile, the wrinkles near your eyes (crow feet), the big muscles on the neck, the massetere and so on.

#365

Thank you for all the words of advice - everyone:-) I am really looking for all the help I can get on this topic. I am getting my terminology mixed up though. I have studied some fces for a considerable amount of time and when I attempt to replicate in modeller, the mesh looks reasonable but not once it is shaded. This makes me conclude that I am reading the topology incorrectly or am trying to define the shapes too harshly?

Is anyone - seriously good at female high detail & animateable heads- interested in paid coaching with me? The only schools in QLD AUS teach Max game modelling and not much more and I can’t head off to full time uni unfortunately.

Does any one have other suggestions?

#366

Roux,

Have been working on a female model myself lately, and there are subtle elements, which are a real challenge to get right.

My girlfriend has magazines lying around, which are good source to study. In addition I’ve asked her, to let me do various sketches of her. None of them masterpieces, but beneficial all the same.

One thing I have found very useful on my hunts for source material etc. are websites relating on how to draw. There’s quite a few out there, covering all sorts of styles like anime, and more traditional, and covering topics like the head, hair, clothing etc.

What they tend to do in there tutorials, is breakdown the initial process into a simplified form, using simple primitives and shapes for instance, and then working up from that. Albeit this relates to sketching by hand, there’s often some really useful advice and tips to be gained from reading them, which can be applied into 3d.

Another apsect I find from studying illustrations, including anatomical ones, is that the ones I see tend to over exaggerate the muscles and details. This can be very helpful in conjunction with actual photo reference, as it’s not always clear to see from a photograph, what’s exactly going on.

These are just my experiences lately. I don’t know if that’s any help at all.

RPG:)

#367

this is my third head

#368

I guess good threads never die. Well since I’m no longer in the game I guess I’ll throw in some parting shots of my work, heads that is. Some are unfinished, some are high poly, some are lower; just a broad brushstroke of some of my heads.

#369

Jonathan, nice work… the goblin head is especially good. Are you really giving up on 3D?

fast4ry, that’s a finely crafted head. The perspective shot in particular is superb. Keep posting your progress.

#370

I still model here and there. Not nearly as much as I used to. I’m not selling my software, that’s for sure. Then I’m not acquiring any either. I’m still using LW7.5 and my long term goals are now more into writing, and doing things that I’ve been putting off for years with the fleeting dream of becoming a permanantly self-sustained artist. Things change and responsibilities continue to mount. Perhaps I’ll finish some unfinished stuff in my spare time, however it’s like I say, the passion for doing it has greatly diminished, although the community still cracks me up, it also kindles much of my disappointment as well.

I’m off work until monday however so who knows, I may fire up Lightwave once or twice and add and tweak a few models.

#371

Nice models, fast4ry and Jonathan. Nice signature,too, Jonathan.

Here’s a couple more from me. The first is a spline-modeled head with a little different topology, based on one of the heads posted previously here. The second is a small update on an older model (remodeled the lips).

#372

“That’s exactly the opposite. If you model something that correctly follows the facial muscles it will lead to a good animation result for sure. That’s the main reason topology is important for.”

take a look here at this example below. Its modeled not necessarily with something that correctly follows t he muscles of the face, but rather modeled as a function design for animation. It doesn’t follow any sort of designated muscle system, but indicates that in its overall form for the sake of function.

Look at the flow it doesn’t follow any medical text book I’ve seen. Its built entirely for its purpose.

“Also your example doesn’t look like it’s animation ready at all, the chin muscles aren’t animatable at all, the mouth and the eyes don’t even have a single animatable loop and the topology of the cheeks doesn’t help as well, not to mention the impossibility to animate any neck muscles, basically in your example there aren’t any. Once again topology and facial structure ARE something you have to worry about, they’re the most important thing to make an animation ready head.”

go to www.monkeyshitfight.com use the password pinK
you can see that model I mentioned earlier. The chin is perfectly animatable. What you might read on the web or in a book about an animatable concentric loop isn’t necessaril the best result for animation. This all depends on what your animating.

“Once again i disagree. One can’t start modeling a head if he doesn’t understand or know the rules he has to follow in order to do it, and topology is the first rule to make a realistic animatable head. Some characteristics and facial features are IMPOSSIBLE to create if the topology doesn’t help them: the smile lines, the subtle folds you got in your chin when you smile, the wrinkles near your eyes (crow feet), the big muscles on the neck, the massetere and so on.”

Not to bash you guy, but again. As the countless examples shown here by others just learning they don’t require the knowledge or the study to learn how to model. I encourage them to jump in and leave the technical stuff for a later point in time.

#373

After reading the feature with SplineGod (Larry Shultz) in the September issue of 3D World I got inspired and did this model in a day:

436 quad polygons for the whole head, trying to focus on flow rather than technical aspects. Created poly by poly, and then later on Bandsawing, Smooth Shifting and Spin Quads for detailing.

Thank you Larry :thumbsup:

#374

Hey Great job! :slight_smile:
I havent received my issue yet. Im glad the article was worthwhile for you. I got to meet the editor for 3DW magazine. Great guy and Im very impressed with the magazine and direction. :slight_smile:

#375

Which 3D world issue…is it this month’s?..I need to buy mine at Barnes and Noble

#376

Hey Jess… :slight_smile:
I think its the Sept issue. :slight_smile:

#377

I have posted a micro-tutorial-description of this subject here:
http://www.cgtalk.com/showthread.php?p=1517445

#378

Hi all.

Man, I have not yet read the whole thread, only the first four pages, and I allready think this is the mother of all threads.

Thanks for all this valuable information!:applause:

Greets…

#379

Hey, Petterms,

Fantastic economy… just shows what you can do by keeping things simple. :thumbsup: Are you going to animate him?

#380

No direct plans at the moment. It was only ment as an experiment to see how low poly i could go, to create the basic poly flow of a head. I was treating the MetaNurbed polygons as splines (inside my own head) :slight_smile: instead of thinking of them as the “final” geometry.

I also used it to study where the inevitable 5 polygons shares 1 vertex (those star-shaped places) of other models to see where they normally occur, and also try to avoid them as much as possible:

After the polyflow is ok, freeze the model with 1 subdivision and start adding detail (not started it yet, too many other things to do at the moment :slight_smile: )

#381

Thats pretty much how I think of the SubD cage too. Thats more or less the advantage of using SubDs instead of manipulating the polys directly is that you use less data (less time) to control more data.

What I do with those star shaped intersections is use spin quads to push them out of the object to an edge or to an area where they can be hidden or they just dont cause problems (areas that wont be deformed).