It's almost like he was posing for a rotoscope.

Click to view attachment

you know, I had wondered a while ago if mug shots would be a good source for rotoscopes!

Interesting.

Now you've got me thinking about the process and trying to distill the lesson's essence. The question that forms in my mind is "Where will the modeler get lost in the interpretation and how can they learn about optimal spline placement?" The answer is not an easy one and so I'm adding a second boundary (the opposite extreme to the photo reference) in the form of the attached silhouette (a quick outline of the photos via MS Word).

Somewhere between the fully abstracted silhouette and the realistic photo reference (too little information versus too much information) is where the true work for the student will be.

100% right on this. Face modeling is such an advanced topic it almost doesn't belong in beginner tuts. But it's demanded, i think.

The front and side will be inadequate for all but the most able 3D modelers who can look at that and visualize the 3D form they represent.

Early on in the curriculum I have them lathing a few real objects that they actually have, like a can or a dinner plate. I want to get them thinking about real-world contours. But it's a big jump from that to a face.

I don't have a sure answer for this yet.

When to do a head is always going to be a tricky decision.
As you say in many ways it is beyond the first time modeller but people will still want to do it. So I would say, if you are going to do it, put it towards the end of whatever modeling projects you have lined up.
And whilst we all know you can make a lifelike head from reference photos that doesn't mean we have do one strait away.
The potential for frustration and disappointment with the end result is quite high. No matter how good the teaching. Such photo realistic heads come only from a lot of practice.
I would suggest starting with something more caricature/cartoon based. With bolder more simplified facial forms. Then, as its not meant to be a 3D representation of a real person any lack of skill on the part of the modeller will be less critical.

I always have trouble with the depth. Working points, describing the process is one thing, but actually doing it is quite another. I think for beginners, the subject needs to be addressed because that will be the expectation, to be able to create human, or human like characters. Perhaps a less realistic head model would be better for the beginners, one with less details, but enough to get them used to the process.

I liked Al Capone because he's kind of roundish and smooth doesn't have many wrinkles. But maybe that's a drawback. Maybe someone with very angular features would be easier.



The face on the Candle is a beginning face that I think novices could do but it's still a big leap from that to a human head.

Caricature seems to be a very appropriate approach. The A:M modeler will be doing that regardless of whether they begin with a caricatured face. Starting from a more caricatured abstraction would simply accomplish some of that work for them. In the case of Al Capone he could be simplified or the photos saturated a little more... just enough to demonstrate optimal spline layout. For instance, the Disney designers were masters at abstracting the Disney style from real world references. (Ref: Andreas Deja's blog). While that example of Capt Hook is a bit more caricatured than what I'd expect to see here it provides a nice example of the art of caricaturing a reference.

What has me especially intrigued is how Al Capone will fit into the over-all scheme of TaoA:M. I assume he's not just attractive because of his profile view but will be the main villain of the piece? From spaceships to lasers to smoky club rooms and fully loaded tommyguns... if you thought you already knew all of the plot twists of TaoA:M... this changes everything.

Real references... real history... a nice way to sneak those into TaoA:M.
In the future I hear they clone and bring back a whole bunch of Abraham Lincolns!

I made this model to use as few points as possible, but she's touchy when you adjust the face. You don't have so many points to deal with, so it's easier there, but you might struggle with some wrinkles and whatever you call it when it just looks unsmooth.

Thanks RU2D4, I'll look at that.

Whoa! Rodney is from Japan! I thought he was a white guy from Vancouver, B.C., Canada. Now I am truly fooled. No monoboom 3D world for you! Ha! Unless you want me to move to Japan. No Boomtube for you! No monoboom bub! All copyrighted here in US. You in big trouble!

Regarding when to model a face, I think it's sort of a backwards question in a way. Of course I honestly have no real basis of professionalism or education to back up this claim, only my own trials and errors. See I started modeling faces fairly early on, probably a couple months into having Animation: Master. Now it sucked and I don't think I completed it, but I never came on here and asked for guidance (which would have sped up the process considerably).

Now I sorta did this off and on and never ever actually completed a face really. It wasn't until about a year ago that I completed one, got critique and modeled another full face.

So my suggestion is make it a topic that's addressed early in their modeling career with AM yet heavily stress the bringing it here for critisism and push the fact that they need to complete it no matter what and then model another one. I personally found that once I modeled a face completely, I saw where I made mistakes and with guidance was able to smooth those out. Of course I still don't model nostrils yet and I'm obviously no where near as good as some like Eric here, but yeah, my two cents on the matter

Rodney is IN Japan, an MP in the US Air Force. I've met him, he actually looks like his avatar.

Well, regardless, this is the profile of the head I drew that became Gina.

I've never actually done a head from a photo. I ought to give it a run through ...

Click to view attachment

Just Decal it for starters. You do a front image and decal an image onto it (in modeling mode if I recall), it will be lousy for starters, but you'll find what you're after eventually. You can decal from any angle.

If it wasn't Al Capone, I might try to do it myself... tomorrow... I'm tired. I'm thinking current figures possibly, but then Capone for practice... sleep now. I've done the talking face before... with audio... Max Headroom style. I had fun with it and would do it again. It doesn't have to be that good... just consistent, and I like the idea! I'll stay on this post.

It just occurred to me, if you're decaling from any angle, using the (H) hide option is very helpful. Helps with ears and stuff that you don't want to spend time with modeling... but if you have the comedy... you could really Monty Python style it up. Hair is really difficult... Lip Syncing is hell... do not let those concepts get in your way. It' s way too easy to find reasons why you cannot! I've been there>

I wish I had the head of Al Capone in a jar to study, that would help. Fleshing out the face some more...

Click to view attachment

Darn, you're good!

He's looking good Robert!

My laptop is in for repair and you've got me itching to spline something. I could no longer handle the thing powering off every few minutes if the power cord wasn't just so-so. The good news... without access to a reliable computer I have been seriously ramping up my sketching/doodling.

(technically, I am not in Japan at this exact moment but I should be again fairly soon)

What if you use a rotoscope of an existing model in shaded wireframe? Maybe use your Al Capone model as the roto, once you're finished. That way, they could use the spline layout as a "connect the dots" and it would show spline continuity.

Here's an example. Maybe a simpler model, as yours is.

Some other photos of him make me want to rethink the splining a bit.

He's looking wide here...




The splines I have now wouldn't accomodate a jowly smile pose like he does here...

Modeling a head is not my fav but I've had reasonable success using Anzovin's "Model a Face with Bill Young". It walks you through almost CP by CP. Those mug shots would make excellent rotoscopes for this procedure. Both photos were made at basically the same range with the same lens and not too much telephoto. A big advantage. Since they have abandoned AM, is it possible those videos can be posted as a tutorial?

I think example meshes are very valuable, although my goal is to get them to start identifying the important contours before they spline. If i do too much for them they are still helpless.

This ain't going to be easy.

that's a pretty successful Obama effort, Bruce. i bet you could have hyped the ears more for good effect.

I haven't seen that one, I don't know if it's been released or not.



Just as an aside, a likeness is a tough thing to do in any medium. At the art museum here they have a bust copied from an actual life cast of George Washington and it only looks approximately like the paintings we are familiar with.

Here's an experiment with re-routing some splines to more closely mimic the sagging folds on his cheek and chin. He appears well-fed.

Click to view attachment

That looks amazing! Very realistic. Fortunately he is usually seen wearing a hat... that may make it easier than doing hair.

Thanks, his ears stick out more in the modeling window, the chor camera adds some distortion. I also thinks his ears have spread over the last three years. The office ages them all.


That CD was released 2004 maybe (v9.5) It has half a dozen .mov video files that walk you through the process. I'm not sure who has a relationship with the Anzovin folks and can ask but if they were willing to donate TSM2, they might be receptive to releasing this one.

Capone looks great BTW!

I think Al Capone needs to roll in a nice ride like this.

This has all got me sidetracked and I'm now modeling Captain Kirk even though I have more priority things to do

A pink Deusenburg! Perfect for Valentines day events.

This is all great exploration but I think if the question is how to teach head and face modeling for the beginner, it needs to be rethought a bit.

It would make more sense if it were worked out in advance just what the successive lessons should be. Then the question is, are we teaching photo-real modeling (Boring, in my opinion, though I like Robert's Capone) or best practices for modeling and rigging a good, animatable head? And do we have any sense of what most beginners want most to learn?

the approach should emphasize simplicity, which means a simpler, more cartoony head, and a simple rigging process that would emphasize simple eye and mouth mechanics.

Decaling for realism is a good goal, but I think modeling and rigging should take precedence.

I'm doing this head to think through the process (I've never done a real head before) and understand what the major problems are. I do that because I regard the current head tut as overly complex and inadequate and I dont' see people getting good results with it. It can be taught better but i need to figure out how.



That's what this whole sub forum is for, to figure out how to structure it all. But I'm not necessarily exploring it all in order. I'm testing out things as i think of them and there are obvious gaps that will need to be filled in.



My goal is not to teach them how to make "Al Capone's Head", my goal is to use it as an example case of how to analyze the problem and teach the skills to solve it. There are some elements that are common to all faces.



Martin says the of the two questions he is always asked when he's selling A:M, one is "Can I make my face with this?"

There is a beginner expectation that they will do something like a real human head.

In the Maya beginner manual, the closest they come to an organic face they rig is a one-eyed blob. We can do way better than that.



I agree. Remember my candle face? That project would certainly come before a real human head. That's the simplest, reasonably functional face I've thought of so far. I'm thinking an added intermediate case, like my "Punch" puppet head would be an appropriate project before the real human head.



Modeling will have to come first. I agree that rigging is essential (I think that makes a total of two of us )

I want to cover decaling somehow because that's an essential part of the "finished look" .

Well, at least three, Robert...and I know of at least three or four others that would also agree.

Yay... but they're kinda quiet.

I would put an extra mouth spline in there, between the existing mouth and nose/chin area to define the lips more. Other than that, this is beautifully done. I also probably add more splines than I need too, haha, so take what I say with a grain of salt, .

Yeah, if you look up at the previous version I had another spline on both the upper and lower lip and I think i should put them back.

Of course I agree. If one intends to animate the model (face), then it is equally important to model with a spline topology that can be rigged (animated) to mimic the muscles/facial movement in addition to resembling the actual static posed character geometry of the rotoscope. The face rig hierarchy, cp weighting are almost inseparable with spline layout.



I know I sound like a broken record, but Malo's modeling method does just that. He presents a very general scheme and actually uses a complex character to demonstrate the general concepts. He presents how to think about spline layout (for any body part, including face) so that one does not have to rethink the spline topology each time. His method is well suited for animation, and well suited for A:M. He presents excellent step by step images, however the text needs some work (awkward language). If the text were clarified, I feel no other tutorial would come close. This is probably why it hasn't been embraced the way it should be.

Maybe that's all we need then. Who can clarify it?

Hey nancy, what and where are Malo's modeling tutes? I'm not familiar with them and it sounds like I should be!

There is a pdf in this post. Download it. It's a gem. I've printed mine out (in color!).

The name of Malo's thread in the Tutorial Forum (modeling section) is "Modeling a character by extrusion". Perhaps the first example of "awkward language" perhaps?

I have many times tried to decipher some of the more awkward text in the pdf document, and usually end up just looking at the images. When I have tried to come up with alternative text, I usually end up thinking "I'm not sure what Malo meant to say, but I sure understand the pictures". But perhaps if we tried to come up with more english friendly text, and perhaps augment with explanatory tidbits (perhaps reasons, perhaps alternatives?) here and there, I suspect it would be more immediately helpful to those who might be struggling with modeling for animation.

To summarize (and sorta using Malo's language), Malo presents in this order, for modeling a complete character:

1) a general method for "extruding" any lump (my word) that needs to arise from a set of patches (eg arm, leg, eyelids, ears, fingers, etc)

2) using 3 splines for things that need to be articulated (elbows, knees, shoulders, etc)

3) How to create a 2D rotoscope to be used for final modeling, by first doing a 3D primitive proxy

4) Then finally: Modeling the character using the rotoscope, lathing a shape, and then refining the lathed shape using his "extrusion method" for
A) Body (with head shape)
B) Members (Arms, Legs, Hands, Fingers, Feet, toes)
C) Face (eye holes, brow ridge, mouth, nose, eyelids, ears) -
D) A generalized FACE spline topology (flattened) - illustrating result of step C
E) missing from pdf tutorial: tongue, teeth, interior of mouth - probably not needed if you got this far.

Nice, thanks Nancy!

What's the other question?

although i have some experience from a few years back I would consider my skills very beginner now. As far as what beginners would be interested -I think for sure most would want to get a good tutorial on doing a face - the goal for most people I think would be able to ultimately create and make our own character and to rig them and then animate them in some sort of short. At least those are my goals. Understandably a lot of learning and practice is required but a goal is where you are looking to get to . So yes including doing a face would be great but I also think perhaps a less realistic one for starters. If ones splinemanship is not up to speed and folks get frustrated because of that --it kind of defeats the purpose. That said if you have a very complete set of tutorials that takes one along a nice path having some challenges in the end part would be good to do and I have to say ---this kind resource you are working on is going to be very helpful in getting more folks interested in AM and sell more subscriptions as well as well as make more happy and creative AM's- So thanks for all your efforts.

This would certainly not be the first tut. A beginner face tut might make something like this

Robcat--that would make a great introductory face learning project ....its more simple but its also funny to look at and would be fun to be able to learn to make.
Rich

I don't recall... it may have been "Can I get a job with this?"

Rob
I have always hat rouble with cheeks, mouths and eyes.
He has beady eyes, fat cheeks and little mouth so I would like to see how you handle these areas.
Even Dimples.