Modelling Heads


Yes, I do know that there is a good Dan Alban tutorial in his “Inside LW 7” book. I own it.

But I want to hear from you, the artist I look up to and how you model your heads, from PISONG’s caricatures, to Leigh’s Pirate.

I’m not trying to be a fan boy, I really want to know the in’s and outs of the different methods.

Thanks in Advance,


Wow, I will be watching this thread.

There are so many great artists who model heads with ease, but they all do it very differently.

I will email a few people on my side and get them to watch this thread. Tim Albee, Steve Warner, Todd Grimes and Leigh all should chime in on this one :slight_smile:

Thanks for the great thread idea Andrew!!


Box modeling is my primary technique:

I’ve tried just about every method (Box modeling, Spline Modeling and Stuart Atkin’s Detail-Out approach), and have found each to have its merits. But I just really enjoy starting with a box and going from there.




Point by point is my favoured technique :slight_smile:
It’s can take a little longer than box modelling, and some people find it harder to get their edge loops correct using point by point, but for me it is simply the most intuitive - it really suits my style of working.

When modelling, the only tools I really use are Extender and Bandsaw. That’s it :slight_smile: I personally hate using Smooth Shift and Bevel and all that kind of stuff. I prefer to keep it simple. A head takes me about 3 hours to make, after which I can spend any amount of time tweaking it or altering it to look different (for different characters).

If you’re really serious about head modelling, grab a copy of Jason Osipas recently published Stop Staring (published by Sybex) - it’s a book all about head modelling and facial animation. The author uses Maya but it’s very easy to follow in any program because every page shows clear wireframes and illustrations.


I’m no pro but i would use the point extender mode too. I find it much easier to plan things out.

Depending on what i’m drawing i’ll start out with the details, eye’s, nose, ears and mouth and then just sew it all together with poly’s

prolly need to refine it a bit but i like it :slight_smile:

I tried box modelling but i just couldn’t get used to it, everything looked very blobbly :stuck_out_tongue: but that’s just me being a bit crap



After reading these two post, and working pretty intensely on a model this week myself I have to say…

Do not try to emulate someone else’s style! Watch and learn from the masters but find out what method suits you and run with it (as long as your method produces good work).

This like no other talent is one that you have to kinda find your own way and that might be a combination of others styles.

Having said that you have to travel all the paths though to find one that works for you.

So look at all these methods, these folks have already done all of this and have years of experience on me at least. But I watch as many tutorials, books, videos and online reading as I can…then just model, model, model…

The best learning you can do is from trial and error, again IMHO.


My methods to modelling aren’t Ideal, and usually require some clean up at the end. But it’s efficient for me.

I use LW’s modeler like a sketch book. I’ll use a celebrity head as an example.

I start with a box, and begin to push and pull and bandsaw my way Untill I have acheived a very rough head, however even at this point it usually proportionally resembles the celebrity to an extent.

For caricatures I find it so much easier to box model. I can open up a ref. picture I’ve found online, or even go by memory, pushing and pulling to get the rough caricturistic shapes of the persons face. It’s fast, and I have a blast doing it.

I’ve found that box modeling will only carry me so far, so from here I literally select polygons and cut peices out, Usually the eye sockets, sometimes the mouth, ect.

I re-model these areas seperately using the point to point method then patch them back in.

It may sound a little messy, but it does feel alot like sketching. From the first “rough sketch” phases of box modeling, to the finer detail sketching of Point to Point, to the clean up at the end.

At this point I usually upload my caricature head for critiques. Once I’ve recieved a bunch of good advice I box model the rest untill I’m done.

I’m working on a writeup at the moment that illustrates this method further :beer:


yeah the way pi does it is close to how i do it
just using modeler as a kind of sketch book

but i don’t make a rough head first
i will just start with the left eye and work out from that
for me using point by point is a much faster and better way to model my heads
most take me about an hour or two to make

i never use any ref pics it’s just all as i go along
and sometimes it works out good… sometimes not :slight_smile:



Thanks guys, this is very inspirational material.

I feel fairly confident in my modeling skills, but the problem I have is head modeling.

I don’t know how to do the correct way to band saw to get the right amount of points, and so on and so forth.

I don’t know how many exact points to create for point to point modeling.

That’s the problems I ALWAYS encounter.

After this x-mas I plan on going all out humanoids.

As a gift from my parents I will be recieving the todd grimes DVD set and a Bill Fleming texturing book.

I just feel lost everytime I sit down to model a head.

I hope this thread can help not only help myself but others that visit this forum.

Thanks again, and I hope this thread stays a live for a while.


Originally posted by AndrewE
I don’t know how to do the correct way to band saw to get the right amount of points, and so on and so forth.

When I began modeling heads, I felt the same way. I kept looking for the right way to do things. But after developing my own techniques, I went back and looked at how others had achieved their results. I quickly realized that there is no “right” way to do things. Everyone works slightly differently. It’s the end result that matters.

Modeling, as a whole, is not about “how” but rather “when.” The question is not “how” do you use the bandsaw tool. I can teach you how in five minutes. The real question is “when?” And “when” is inherintly a question about problem-solving. In that regard, learning to model is the process of learning to problem solve. You have a specific set of tools. You’ve learned how to use them. Now you must learn when. Unfortunately, there is no simple way to do this. By and large, it is the product of experience. The more you try (and the more you fail) the more you will learn “when.”

Now, with regards to the human form, I should probably note that a head is different than most other man-made objects because what you see is the end result of what you do not see - in this case, bone and muscle. When I’m building part of a car, I’m only worried about the final shape of the metal. But the shape of a human face is determined by the shape of the skull and the flow of the muscles over that skull. The trick, then, is to model your polys in such a way that they mimic the flow of the muscles. You don’t have to do this, but if you ever plan on animating your character, it will really help.

One of the best way to learn about flow is to look at the wireframes of other artists. Everyone accomplishes this flow in a slightly different fashion. But they generally follow a similar pattern. Study these wireframes. They will be your guide to understanding the structural layout you need to achieve.

Good luck!



I’m pretty new myself and this thread will help me a lot. I always see incredible heads in here and i always wonder if i can do that. I guess after all your advice its just a matter trying it out and not getting frustrated.

Leigh: yeah i flipped through thta book at a store. I wasn’t sure if i should get it, but if you recommend it, imma go pick me up a copy :slight_smile:

PiSong: Dang your models are pretty tight. I always know who it is even if i dont read the topic. Yeah do a write up. I want to see how you get it done.


This post is great! Thanks to all of you.
If we can’t borrow other people’s technique, as Wordwarepub wisely said, we can at least realize what kind of techniques and variations of techniques are used. That’s the coolest thing about this thread.


Theres a difference between modeling techniques that work if you are an experienced artist VS. you are trying to learn the software AND trying to learn the art of head modeling at the same time.

Ive taught people to model characters for years. In doing so Ive come to notice a few things;
The point by point/extender method works better for people who already have a good art background. They tend to know where things are on a human face before they start modeling. Proportions and flow are concepts they understand already.
I h ave found, in my experience that the vast majority of people who DONT have that art background fail miserably when following this method. Why? Simply because this method breaks many of the rules that most good traditional art books, courses or instructors follow, namely that you start wtih the biggest, most general details first and then work your way into the details.

Starting with the details when you dont know exactly where youre going first (experienced artist) usually ends creating more problems then it solves. If your basic model, the foundation of the rest of your model, has problems, those problems only multiply with every additional vertex you add. Ive never seen any books on drawing, sculpting, painting etc. that were designed for a beginner EVER start with details like an eye and then build out from there. Its always start with a primitive shape and work into the details. Thats the best way to plan where you intend to go.

If you noticed on the little sections of my tutorial I have up, I start with a simple polygon and at ever step make sure its as close to where I intend to end up later. I layout the proportions in exactly the same way any good drawing book does.

Once you have the basic tools down and get a better handle on the artistic aspect you can skip, like many experienced artists do, the basic steps and jump right in. :slight_smile:

Again, this is the difference between an experienced artist using Lightwave and someone new to both the art and technical aspect.

For myself I GENERALLY reserve point by point for detailing in small areas rather then as a general purpose modeling method.

I use splines typically as a way to layout form and flow very quickly. The great thing about splines is that they use very little data therefore I can layout things far quicker then point by point.
Its like sketching in 3D and the splines not only show you a “3D” sketch but they also define the flow. Theyre also easy to change quickly. Patching them is a no brainer. Splines are an ideal way to trace reference loaded in the background if you need the model to match the reference exactly. You dont get the clutter that you do when throwing a bunch of polys over the reference.

I will sometimes start off with a primitive shape such as a single poly, box, tube etc and then sculpt and refine as I go. This is more of a free form approach.

When Im handed a model sheet and its a dead on front and side view that usually tells me to make the model exactly like that. Usually Ill go with splines. If the model sheet shows a 3/4 view of a character or a couple of poses that generally means to make something kind of like that. I will usually resort to starting with a primitive shape and sculpting or a combination of splines and primitives.


That nails it perfectly!


I would love to try and do it the spline way, but I can’t find anywhere on the net to do so.

I know larry, your the god of splines, would you be able to show a quick step by step guide to modeling with splines?

When I try to box model a head I get so close but then I get stuck on how they add more mesh. I find that there’s never enough information on how to add more mesh. And if anyone can ever teach me or show me how, I would be greatfully theirs!

That’s my two cents on that.


Hey Andrew,
Before jumping into a full blown head modeling thing using splines I would get the hang of how they work…just simple stuff.
I have a very simple quicktime up on my site that shows how to model a reptile head using splines.

If you need to add more vertices/edges to an object you should look at bandsaw and the knife tool. There are other free tools that allow you to add cuts along polys as well such as quickcut and jigsaw. Smooth shift and beveling are also great ways to create detail for things like eyesockets and noses. If you saw those quicktimes I have up showing the intro parts of modeling a head youll see knife and bandsaw used as well as smooth shift.


I know this is going to sound stupid, but which website?

I have problems navigating through it…


The part of my signature that says free tutorials or
at :slight_smile:


Before I start modelling I study the shape I’ll be making before making a decision on what modelling approach is the most appropriate. For heads I usually use the extend technique. I don’t feel too comfortable making a head from a box.


Originally posted by Leigh
If you’re really serious about head modelling, grab a copy of Jason Osipas recently published Stop Staring (published by Sybex) - it’s a book all about head modelling and facial animation. The author uses Maya but it’s very easy to follow in any program because every page shows clear wireframes and illustrations.

I recently purchased this book and think it’s just brilliant. It teaches some great theory. Other books I;ve read teach you “how” to do things but this book teaches you both “how” and “WHY” to do things certain ways.

PiSong: Love your work. I’m really looking forward to that write-up.


Cool, thanks guys for the post.