muscle definition seems none-existent, and the anatomy and planes of the face are off. I don't know what reference you're using, but either it sucks, or you're not following it very well. Here's a thread that'll help you get the basic shapes of the head down:
It's a simple question of art practice. If you can't sculpt a good head, then you don't have enough knowledge of it and/or practice. The only way to fix it is to sculpt a hundred heads using reference.
It's a simple question of art practice. If you can't sculpt a good head, then you don't have enough knowledge of it and/or practice. The only way to fix it is to sculpt a hundred heads using reference.
Legion, I think you're going to be in this weird rut if you keep modeling the exact way you do in Blender.
What I mean by that:
As a 3D Artist making a head, you're trying to combine two big ideas: Topology and Form.
Topology is simple enough, there's examples in the Polycount wiki. You would just need to see the general methods and copy from an existing example.
What you need to look at more closely, and put into practice by modeling even SIMPLER forms and drawing it out on paper, is form.
The previous guys and gals have posted this. I would like to add additional resources like the Asaro head.
You need to notice and study these planes of the face, along with proportional relationships of elements on the face.
If you're only going to go about modeling this in lieu of drawing (which is how I originally started), I recommend literally just recreating the Asaro head in blender. No need to turbosmooth it, just get it modeled exactly with all its planes. It sould be easier to push verts that way. The assumption is that you'll learn about the major planes of the face, and you'll know what to do even though in the end, those places are hidden under smooth, rounded skin.
Others already gave you a good critique. I highly suggest drawing, like 2D and from real life. Read a book about face, observe the face and study it the anatomy.You can probably find Andrew Loomis head book pdf somewhere. Look it up. I highly suggest it.
It's been posted on some other threads recently but I think it will be extremely helpful to you to start thinking about the face in terms of its planes. The asaro head is a great reference for that. Think of it as a bunch of hard planes, at different angles, not just one giant soft bundle of meaningless features. Once you get these planes in, it'll really start to look much better.
It looks like you're making front only and side just for shape and wish that it will become a head. You really need to think of it as of 3D object and not just 2 planes.
When it comes to pure modeling this tutorial is my favorite approach -> [ame]https://www.youtube.com/watch?v=xls25e08sSg[/ame] 10min tutorial and you are on a very good path. By working with as little geometry as posible and automatically creating good edge loops it makes it easy for beginners.
Good shape comes with practice, but using 2 views from good allined pictures is the way to go for a solid start.
On a related subject to OP, if you have done asaro head what should you do next for practice?
I've made the model in maya and I feel like I didn't learn that much from it. The turnarounds I see for 3D models tend to not be orthographic or if they are their are subtle differences.
Edit: Character modelling seems really hard to judge compared to hard surface or environmental modelling. I'm still a beginner through.
Asaro head teaches you what you want to nail in the beginning before you move on to smoothing things out, no matter the proportions of the head.
You're not learning a specific thing, you're learning something you'll be eventually subconciously doing over and over again for human heads.
It also teaches you that good visual communication doesn't necessarily mean smooth forms all the time. You can communicate forms quickly by using planes. Imagine someone saying "I don't know WHAT kind of dog I want sculpted, but can you make 20 different ones for me to decide which one I want?" Easier to just planar sculpt 20 out instead of meticulously making each one with the right curvatures.
I akin planar modeling to 3 value paintings. You get communicate the big ideas quickly without committing a lot of time into any one thing, which is good for iteration. And as game developers, iteration is something we need to become close family with.
@Legio_Studios: Nothing changes with the wireframe. You're topology is on the right track with the modifications, but what's more important is that you model a proportionally sound head.
Just work with a far simpler mesh until you get the proportions right. This applies to modeling, sculpting, hard surface, drawing,... actually everything in life. Understand the basics before going into details.
Rule of thumb: the mesh should be as simple as possible and as just as complex as necessary. I guess you could get a stylized recognizeable likeness of that picture with 1/4 of the edges you are using right now. They just confuse you and add little to nothing to your actual goal at the moment.
My recommandation: Start anew (because it is easier and faster). Use one of the approches we mentioned here. Add vertecies/edges only when you really need them to add form, but force yourself to keep the mesh as little subdivided as possible. Once your proportions work out, you can add additional edgeloops to increase the details.
But again - the basics are the key to all of this. Get the anatomy right and everything falls into place later on. Screw that one up and no matter how many triangles you use it won't look good.
You are modeling that way because you are not looking at the top view. Everybody does this when they start out, they look at the front and side but not the top. If you look at this head it will be square from the top or bottom. You are also not really looking at how a head is shaped otherwise you would see the errors on the 45 degree sides of your model. You are not modeling what you see you are modeling what you know and that is normal in the beginning and it gives results like you show.
Replies
Once you do that you should have a better idea of where you went wrong. Source: loomis
http://www.polycount.com/forum/showthread.php?t=76733
The polycount wiki also has a section on anatomy:
http://wiki.polycount.com/wiki/Anatomy_Reference
It's just anatomy problems here. You need to pay attention and make what you see.
A "head" is vague. Is it a man; woman, child, fat guy, cartoon character etc? All of those require specific approaches.
Exactly this. Practice practice practice
What I mean by that:
As a 3D Artist making a head, you're trying to combine two big ideas: Topology and Form.
Topology is simple enough, there's examples in the Polycount wiki. You would just need to see the general methods and copy from an existing example.
What you need to look at more closely, and put into practice by modeling even SIMPLER forms and drawing it out on paper, is form.
The previous guys and gals have posted this. I would like to add additional resources like the Asaro head.
You need to notice and study these planes of the face, along with proportional relationships of elements on the face.
If you're only going to go about modeling this in lieu of drawing (which is how I originally started), I recommend literally just recreating the Asaro head in blender. No need to turbosmooth it, just get it modeled exactly with all its planes. It sould be easier to push verts that way. The assumption is that you'll learn about the major planes of the face, and you'll know what to do even though in the end, those places are hidden under smooth, rounded skin.
Good shape comes with practice, but using 2 views from good allined pictures is the way to go for a solid start.
front
side
and my wire frame
I've made the model in maya and I feel like I didn't learn that much from it. The turnarounds I see for 3D models tend to not be orthographic or if they are their are subtle differences.
Edit: Character modelling seems really hard to judge compared to hard surface or environmental modelling. I'm still a beginner through.
You're not learning a specific thing, you're learning something you'll be eventually subconciously doing over and over again for human heads.
It also teaches you that good visual communication doesn't necessarily mean smooth forms all the time. You can communicate forms quickly by using planes. Imagine someone saying "I don't know WHAT kind of dog I want sculpted, but can you make 20 different ones for me to decide which one I want?" Easier to just planar sculpt 20 out instead of meticulously making each one with the right curvatures.
I akin planar modeling to 3 value paintings. You get communicate the big ideas quickly without committing a lot of time into any one thing, which is good for iteration. And as game developers, iteration is something we need to become close family with.
@Legio_Studios: Nothing changes with the wireframe. You're topology is on the right track with the modifications, but what's more important is that you model a proportionally sound head.
Rule of thumb: the mesh should be as simple as possible and as just as complex as necessary. I guess you could get a stylized recognizeable likeness of that picture with 1/4 of the edges you are using right now. They just confuse you and add little to nothing to your actual goal at the moment.
My recommandation: Start anew (because it is easier and faster). Use one of the approches we mentioned here. Add vertecies/edges only when you really need them to add form, but force yourself to keep the mesh as little subdivided as possible. Once your proportions work out, you can add additional edgeloops to increase the details.
But again - the basics are the key to all of this. Get the anatomy right and everything falls into place later on. Screw that one up and no matter how many triangles you use it won't look good.