Modelling Heads

#321

Kind of like the “Bachorette” method. Coolness… In the past, I have done Point to point modeling, and the results were breathtaking (?). Box modeling was nice for me, but I find that using the splines to layout polygons is sweet as ever…since I most use splines on objects in the background that do not require massive amounts of polygons.

My focus with spline modeling is for video games. And that drawing I sent you was just a rough sketch…the detailed sketch of the same person is for texture maps…

Of course, I will try that method as well…

#322

Theres lots of ways to model but when it comes to learning my experience is that using a more traditional approach that starts with the general layout and then works into the details is easier to follow and understand.
Knowing the subject is a very important key and if you dont know the subject you intend to model intimately then having reference is important.
I think splines or variations on the box method are great ways to rough out a general form and get the basic proportions right before detailing. I think point by point is better used for detailing rather then general way to model.

Lazynok,
Ive been known to work out arrangments with students.
Email me. :slight_smile:

#323

I have to say after reading the post some parts are interesting, but I must say that I disagree with Mr.Shultz. Those who are using the point by point or spline modeling system normally produces results that are either not useable for production or lack substance. With a background in Industrial Design, I can say that Spline modeling is for accuracy and tolerence of geometery so it can be put through Rapid Prototyping, since none of the models I see here are going to be used for that reason its not necessary. For use in visual fX its nearly useless, when does it come in handy when you are resurfacing a data cloud from a scanned model. Yet to be revealed is the true benifit to modeling from a box, turning edges, draging points with symmetry etc.

I initially started spline modeling when I worked in the game industry and let me tell you it made me dislike modeling to the point I wanted to stop learning 3D all together. It wasn’t until I was revealed a system of modeling for the purposes of animation by modeling from a box and turning edges did I find my self creating my designs only possible on paper and in my head.

Truely understanding your designs and the objects which you approach to model is really the first step and for that its not even necessary to use a computer.Grab some some clay and try sculpting, grab a pencil and start sketching. That’s my two cents on this matter.

I guess the truth is I don’t really understand why someone would start to create something in 3D space from floating points rather then create a chunk of geometery and keep adding that digital clay. To even the novice I would say that this is a clearer more defined way of starting to understand 3D forms.

Chris Reid (www.monkeyshitfight.com) and Timur Baysal (www.taron.de) both of these guys have sites which you can truely benifit to seeing how this intial method of observation can produce very intended/designed results.

#324

I guess we will just have to agree to disagree on this one. Ive seen people model in all kinds of ways and turn out LOTS of models that are/were production ready. The end result depends on whether or not youre modeling for yourself or modeling for somone else. In the end the goal is to produce a model that is correct in terms of form and flow.
The examples youre showing here dont look any better or worse then anything else Ive seen regardless of what method was used to produce it. I think your generalizations are off. :rolleyes:

#325

IMHO I believe simply that if you can produce a good, accurate mesh, who gices a s**t what method was used to model it?!
Everything is down to personal preference. It’s the same as an opinion. No one can change someones opinion.
Just like artists who like to work in pen or pencil, colour or b/w, etc!

#326

Mr. Shultz,

I guess we do disagree. But modeling for form and flow isn’t necessary going always give you the desired results for animation. Case in point a model built for animation may not have the accurate desired flow for muscles in the face or the rest of the body. Models for animation are subdivided specicallyed and edged turned to allow for deformation to occur properly.

Most splinemodels I have seen don’t necessarily do this, and if we really are talking spline models then your talking beziers and nurbs to which case are really bad for choice for animation (which proved a point to my old T.D). Spline modeling in Lightwave can be extremely time consuming and much more of a mechanical process then designed/artistic process.

I’ve never seen a spline model that convinces me that underneath the surface there are muscle tissues and bone indicated. Not to attack your personal work, but its a good example of what I am talking about.

My point here is that your methods are your preference but there not going to get anyone a job.You even make mention that your method comes from drawing books. What drawing books are you refering use a system that reflects spline modeling. Even when one approaches things with Rapid Visualization, your drawing cross-sections but never are you using those as the basis for your form.

Also I got this tip from Meni. Model with lighting in mind. The wire frame shaded mode can be decieving as it doesn’t indicate really what your form is doing(you just see the facets). Modelling from Smooth Shaded mode and using the selection tools to select deisred polygons so you can see your wires. Yet still with smooth shaded look. Your only going to be able to take advantage of this when you are modeling/sculpting your geometery, you won’t get to this point until way late in the modeling when you spline and patch stuff.

Doing some chrome studies actually proved to me that accuracy of a surface is all light driven and just looking at a spline cage won’t tell you that.

The fun part is when you model in this way you get the bonus of no geometery amplication issues when you paint in Zbrush. The model comes in with the correct subdivisions in place.

#327

Don’t mean to break into this speculative argument (ok, I really do mean to), but I wouldn’t necessarily say that spline modeling has to take any more time then box modeling or point by point especially when you consider that Larry is only talking about using splines to lay out a basic frame work upon which he uses other methods to add detail and finish the model. I happen to agree with him on this point.

I don't use splines to model characters but I could see how one might find it a more precise method, perhaps those who don't want to spend a lot of time tweaking and pulling points which happens to be my favorite part of the modeling process.

What I do disagree with Larry about is his blanket statement about how certain methods aren't the best ones for beginners with or without an art background.  Whether that is his experience or not, it's still a blanket statement and people should learn whatever method feels comefortable to them and not learn another way becasue someone who teaches tells them to.
#328

Side Note: Does anyone know CGtalk works on a internet based palm Pilot…amazing…

Hr: Try spline modeling the head you see from monkeyshitfight.com, it ain’t going to happen, not even if one could layout all those areas could one model them with splines, one would spend so much time with it. Think about this, in a production environement are you really going to sit down and realstically do a model point by point?? Or patch and spline?

When I met larry at the unveiling of [8], I heard him say that blanketed statement, i just thought he was kidding.

#329

For me, splines are perfect as a layout tool. I can create a blueprint of the entire model with splines. I spend more time creating the splines but far less time reworking polygons later. And once I have this blueprint in 3d space, then I can do the model any number of ways -patch, point to point, etc. To some this may seem like creating the model twice but I can’t emphasize enough how much time is saved because of less tweaking later.

That said, I’m still learning all of this, and I fully intend to increase my skills in both spline modeling and box modeling. But, crossbone, it seems to me that getting good models for animation depends on the skill and experience of the modeler rather than the technique he uses.

#330

Crossbones,

I might have missed something in your posts but what exactly is the benefit of modeling without a centerline?

#331

Originally posted by crossbones
[B]
Hr: Try spline modeling the head you see from monkeyshitfight.com, it ain’t going to happen, not even if one could layout all those areas could one model them with splines, one would spend so much time with it. Think about this, in a production environement are you really going to sit down and realstically do a model point by point?? Or patch and spline?

[/B]

Well, first of all, as I already said, I don't model characters with splines.  It doesn't mean it can't or shouldn't be done.
Secondly, I wouldn't try to model the head from monkeyshitfight.com because first of all, it's not the greatest head I've ever seen and I hardly think that the polys have that great of a flow anyway.  Where are the neck muscles (scm) anyway?  I assume you're talking about the head you posted above with the watermark on it.  I tried to go to the site but it required a password (hence-not going to bother with it).
 Am I really going to sit down in a production environment and model something point by point?  Let me think....you bet your ass I would.  Don't make assumptions about how long it takes someone to do any one thing.
#332

Damn it, I am treading water when it comes to the nose…I cannot seem to get the nostrils correct…can anyone here offer help? Thanks

#333

If it will help Halo, give me your email address and I’ll send you an example nose to look at…

#334

Well, okay… email me at RedStorm2000_@hotmail.com. I also built my head from a spline cage so I will show you what I am having problems with…

#335

Actually, scratch that last message. You can download the file here http://giger3d.com/Nose.lwo but yes, it might help if you post what you have so far.
I wouldn’t use splines to model a head but it doesn’t mean you shouldn’t. But there could be easier ways.

#336

Do you want to see where I am stuck? I will send you and email with the image in it…

#337

Yeah, send it on over…

swhite8018@wideopenwest.com

#338

Alright Halo, I looked at the model and I made some adjustments but I can tell you the big problem with the nose is that you don’t have enough geometry coming down the front of the nose to define that shape accurately. It seems you have too much geometry in some areas (you really don’t need that many rings around the mouth or eyesfor instance, you just need enough to define the shape you’re after) but then not enough geometry in others.
If you notice, I also reconstructed the area between the eyes because you had several polys all meeting at one point which will cause you creases later. Another potential problem area is right near where the jaw meets the neck. You have 6 polys sharing one point and considering this is the origin of movement for the jawbone and speech, if you plan on animating this character, you may want to rethink this polyflow.
Perhaps it has to do with laying out your model with the splines. I imagine it’s probably not the easiest to see how your polys are going to flow together when you’re piecing together many patches of polygons. You might be better served with a method that gives you more immediate feedback on how your polys are going to work together…

#339

Okay…I can see the forehead. Could you diagram the jaw area? I could simply fix that with Spin Quads…

#340

Man,
I’ve followed this thought …

Helll, wherever…

Find a master. That’s what the Budhist boook tooold me…