CGTalk - Base Mesh : Planes of the Head (John Asaro) VS. Topology

gnarlycranium

11-02-2002, 03:24 AM

Ooh ooh, that is just spiffy! I've been roaming these forums and many websites for a long while now trying to figure out what the heck a good basic structure for a head is-- I've found that it's an absolute requirement for making a model that LOOKS halfway like ANYTHING. I wish I'd had somebody to show me a wireframe like this a year ago, it could've saved me months of Cranium-scratching. It looks like you've broken down the head into the basic edge patterns in a very coherent and well-handled fashion. That curl around the bony structures behind the ear is something very few people use, and it's nice to see. Everything's at an efficient minimum, which is great, showing where all the 5-edge poles are and how the different areas interact, it's lovely.

There's still a few bits I'm not so sure about. The hinge of the jaw, for instance. There's not much there to support detail in that area, for characters with defined jawlines. Also, the edges leading down from the corner of the eye next to the nose... it looks like they wouldn't blend into the slope of the nose and cheek very well, the outermost row is very narrow, and will probably give you too sharp of a crease where there shouldn't be one. Ooh, discussing this stuff will be fun!

phreaknasty

11-04-2002, 03:04 PM

bjotto - thanks for the comendations... and i hope you got a good nights rest

gnarly - glad to here it could have been helpful :). i certainly did my share of wireframe hunting as part of the effort to come up with this mesh and must thank many of cgtalk's own (Roy's, chamade, stuzzi, syomka, Stahlberg...) for unselfishly providing very useful wireframe examples in the past. hopefully my head can add a little to the effort.

a)jaw - i think i see your point. i'll run the mesh through some basic tests today i.e fat, thin, muscular... to see what i can get out of the current jaw configuration and improve it where i can.
b)eye-nose - in my smooth tests it worked pretty well, but i can see why it might be a problem, when i get to work i'll try spreading the lines apart especially as they approach the eye and see if that doesn't help.

thanks again for the help

gnarlycranium

11-06-2002, 02:58 AM

Heya, this page seems pretty darn astute and helpful for this stuff, you seen it?

http://coldfusion.art.msstate.edu/camenisch/thehumanhead/index.html

phreaknasty

11-06-2002, 02:33 PM

gnarly - yup, i ran across the human head site a year or two ago. the instructional parts are pretty helpful as a way to get you feet wet but i usually find it more useful to learn by example (i.e. the human head gallery and sites like this).

and an update per gnarly's suggestion....
the jawline, especially towards the ear, should now much more controllable and accurately defined.

more crits? anyone else? please? pretty please? don't make me rename this thread... "character: elven-hot-fantasy-chick anatomy"

My Fault

11-06-2002, 04:46 PM

Nice job, very well laid out head The one suggestion I would make is that it looks like the top of his head is cut off. General rule of thumb is that the eyes lie half way down the head, you've got yours set about 2/3rds from the jawline.

Still, really looks good so far. :beer:

phreaknasty

11-08-2002, 04:49 PM

my fault, now that you pointed it out, i can see that the eyes are obviously screwed. its amazing what you loose sight of when you're too deep into your work.

here is the correction they should now be very nearly halfway down. the image in the back is the original screwed one at the same scale.

thanks

My Fault

11-08-2002, 05:28 PM

Haha, that's what happens when you get in to the zone!

This just keeps looking better and better. Nice job!

seabede

11-09-2002, 05:17 PM

Hi,

I am impressed, very nice. Keep on working!

Regards Endre

Lunatique

12-02-2002, 01:40 AM

Excellent work. I'm really anal about topology too, and I've been doing the same thing you have--trying to come up with the perfect prototype topology for a head.

You might want to move a few lines from the jaw hinge/behind the ears towards the middle of the lower neck to accomodate the neck muscle.

bored alien

12-02-2002, 01:54 AM

I agree this looks really good in terms of topology, and it's about the level of detail that I like. Looks like the ear could use some more attention perhaps. Great work. :thumbsup:

kpalazov

12-08-2002, 12:28 PM

I very surprised you have not resieved much feedback?/!!

But anwyays, I have been reseraching just the very thing you have done . I think this is a great attempt at solving the problem and a very usefull one at that. I find that its good to mix and match different topologies to get just what your looking for, so this is very helpful to me.:)

I think its a great lesson to learn while starting modelling and I think what you have provided is a great resource.

I want ed ask though , if it alright, what was the process you underwent to achive this result----direction to changing topology direction ----conforming topology to gfavial features---stuff like that?:)

thanks once again defininety an outstanding piece of work

ps. can we see a front view:)thnks

take care

Kiril

phreaknasty

12-09-2002, 07:07 AM

hmm.. seems this thread lost a couple of posts in the mix up....

seebede - thanks

lunatique - thanks again, again. any chance you could repost the picture with your suggested changes. i got swamped at work and had to switch over to more pressing tasks for a bit and didn't get a chance to incorporate your suggestions. also you had asked a question in your last post about my modeling intentions which i hadn't fully read through yet, if you repost that also i'd gladly respond. sorry and thanks

kpalazov - thank you very much for your compliments. i do blush. :blush: i'd hoped that more people would respond for a good old shake-and-bake, but thought the responces have been somewhat scarece, those that have responeded have been very helpful. :thumbsup:

i'll gladly post a front view once i get to work, though i'd like to incorporate some of lunatique's suggestions that got deleted in the server mixup first.

i'm not sure if i'm qualified, and there really was no process, but... i constructed a poorly laid-out face using a very basic set of tools: split poly, append poly, smartCollapse, smartDelete, and smartExtrude. once it looked decent i ripped it apart. using the same tools i slowly built it back up trying to incorporate ideas from various sources i had collected including muscle and bone reference, a picture of John Araso's planes head, a mirror and a lot of other peoples wireframes. nothing special just time. it was at least partially as a learning tool so the process lacked refinement.

i doubt i answered your question to any level of satisfaction, so feel free to ask for clarification. have you constructed a face you’d care to share?

thanks again

Lunatique

12-09-2002, 08:37 AM

http://www.enchanted.prohosting.com/temp/cmw_topo01-fixed.jpg

Are you:

a)trying to make a head with the perfect/most efficient/accurate topology? If yes, then you are doing fine.

b)trying to make a base mesh with low level of detail so you can make other heads with it, and also use it as a wrap deformer for animation/making blendshapes. If yes, then your mesh is way too dense for that purpose.

phreaknasty

12-10-2002, 12:06 AM

lunatique - thanks for reposting, i sincerely appreciate the time you've taken to respond to my post. i'll try tweaking the head in the next day or two.

i'm actually doing a couple of things simultaneously (which might be part of the problem). one part is definitely an attempt to create a head with the most efficient, logical and anatomically cohesive topology i can muster. and thats what i'm striving for currently. the second stage that I hope will follow upon a successful creation of part "a" will be an attempt to aim for a head smack-dab in the middle of "a" and "b". unfortunately unlike Stahlberg the theoretical iterations built off of this head won't be created in a full fledged program. more likely then not the controls will be limited to self contained sets of predefined deformers. the ability to freshly generate ears and nostrils won't be available. conceivably there might not be any human input to guide the iterations at all. i may however get a chance to a couple of base heads (man/woman/child or maybe thin adult/fat adult/child) to ease things along.

cheers

phreaknasty

12-13-2002, 07:29 PM

kpalazov - here is the latest front view (with lunatiques chages) if you are still interested. i'll post the other view later today.

kpalazov

12-13-2002, 09:49 PM

hi phreaknasty,
I'm just comming from a bussy week , first time I'm reading your posts. THanks so much for the update and views, greatly appricated.

Firstly my work , the only work I have on the net at the moment is from long ago , but here it is just the same:):
http://www.darkside3d.co.uk/kircho/HeadMesh.jpg
this was WIP at the time I made the jpeg, sorry for the quality.

The mesh is more dense then yours but I found that , after animating the head the mesh conformed very well to the creases and stretches of the skin. Ofrourse I have not created a next and ears and they are some what of a challenge for me as they are very important, ear to look, neck for goood deformation.

Over all your mesh seems to make a compramise batween the desity of the mesh and the flow of the contours of the face which I thing is a much better way to head the problem.

In the next post I will try to give you a better crituque:) sorry for the rambling:)

thanks again for your time in posting the images.

take care

Kiril

Lunatique

12-13-2002, 11:08 PM

phreaknasty- I just noticed another problem with your topology.

Wrinkle your nose--notice how there are wrinkles that stretch from your inner eye corners to the middle of the bridge of your nose? You need to change some lines around there to conform to that.

Check out the topology on the head I have on my website in the 3D section. It' an old head that I never finished, but I did the topology in that area to accomodate the nose wrinkle.

phreaknasty

12-14-2002, 04:58 PM

kpalazov - the mesh is looking mighty clean. nice.

lunatique - funny you should mention that... i had a snarl built into the model long ago but somewhere along the way i must have tweaked it out since viscerally it seems so unnatural to have those lines coming to a peak. oops.

anyways here is the new nose.

Lunatique

12-15-2002, 02:17 AM

Oh yeah, much better.

psilipsycho

12-15-2002, 02:34 AM

Your model looks very nice. I would suggest pulling the forehead forward a bit if your going for the average human shape. The slope makes him look less than intelligent.

good job!

phreaknasty

12-15-2002, 06:15 PM

psilipsycho if, hypothetically, the model were roughly based on me, should i take that personally? :p ... i'll post a full view of my current iteration tomorrow ... i think the forehead should look a little better.

kpalazov - as i said before the mesh looks very clean. the only thing i noticed that might be a problem was the area between the nose and mouth. some of the lines closer to the nose have a tendency to swing up towards the eyes as the reach the "nasal furrow" one in particular meets the furrow at a 5-pole vertex. when you were animating this did you run into any problems such as pinching around that area?

cheers.

phreaknasty

12-17-2002, 02:44 PM

latest head...

anyone have any more suggestions?:surprised

levin

12-17-2002, 03:10 PM

wow. great clean mesh.

in your effort to achieve a base topology, haven't you pondered on a requirement for a different topology for different head types? i mean, can one base geometry be suitable to ALL kinds of heads?

andrew loomis' head types:

Lunatique

12-17-2002, 05:20 PM

I actually think it is possible to produce a base mesh that can cover all head types--taking for granted that it's not a MUTANT HEAD that we are talking about! :D

The point of a base mesh is actually not about the structure of the head, but more about the wrinkles and the muscles. So, even though head shapes can vary drastically, the underlining muscles and the surface wrinkles caused by them are essentially identical--just degress of severity. With that said, you can alter the base mesh's overall shape, but as long as you keep the main topology the same, you still have accurate deformations for animation.

phreaknasty- Here's another suggestion. If your mesh is supposed to be the embodiment of an "average" of some sort, I would suggest that you don't make it look like anyone(you mentioned that you used yourself as reference). Make it actually have generic proportions and features. Right now, your mesh is very distinct, and carries the idiosycrasies of your face.

edit: on second thought, it looks like you've changed some stuff in the latest version. I'd say if you changed the chin, you'd have a pretty good "average" head. What the hell DO you call the area between the tip of your chin and the top of your neck anyway? That's the area that needs a lift. Right now, it slopes down too much.

megaflaizer

12-17-2002, 07:24 PM

first, id liek to start by saying bravo - a very fine mesh

i have to agree witht he point above however, that different heads equire difff topologies regardless of similar muscular structure. this is because the fat of the face is structured diff and reacts diff based on how fat the person is.

might i recommend three base meshes, which could actually be derived from the same base mesh but with differences
- thin guy: more jawline detail, more lip and under the chin detail to meet with neck
- fat guy: less jawline detail, perhaps edge loops run down to neck diff as the neck is more joined to the head in fat people. some dont even have necks!!
- female: gotta make a separate female mesh, with diff edege loops esp down around the mouth and throat

phreaknasty

12-18-2002, 01:22 AM

yike... lots of post. sweet :D

levin - thanks for the praise. yup, i've certainly planed to, and considered creating various head shapes. i've been focusing in on an relativly average male for simplicity's sake and am just now begining to put it through some shape tests. you're quesion about the feesibility of one mesh being used to create ALL head shapes is very valid. as i mentioned in a post that seemingly got whacked in the server mix-up and as megaflaizer suggests below i may have to brake it up into 2 or 3 base meshes (woman/child/man or perhaps fat adult/thin adult/child (or perhaps female). i'll certainly incorporate the loomis picture you provided into my effort. nice image - thanks.

more later tonight....

phreaknasty

12-18-2002, 02:39 PM

...
lunatique - as i think we discussed previously i agree with you about one mesh potentially being able to encompass all of the necessary structures to generate almost any shape. if need-be one could imbed latent lines that would be used in some instances and "smoothed" into the surface in others. however it may come down to an issue of
cleanliness vs usability. mhhh... also i see your point about the jaw/neck region. i started with my head simply so that i had some easy to come-by reference material. then as part of my effort to run it through some shape tests i hoped to make a more generic face. i'm relatively generic looking :p (for a 20 something white guy) so i hope it won't be to much work.

as to the name of that region. you got me there. perhaps hyoid
region... now you've got me intrigued... mhhh...

megaflaizer - thanks. i certainly agree that i might have to create a couple of base meshes as i put forth when lunatique first broached this subject (both his initial question and my response seems to have been kill in the server mix-up). my initial thought was woman/child/man or perhaps fat adult/thin adult/child but yours seems equally valid.

however i'm not sure if its impossible to create a base mesh that can take almost all of the variables into account. looking at loomis's head shapes my mesh should be able to handle almost everything there to a decent degree - especially in the second smooth iterations. i've already tried incorporating many of the necessary lines. for instance the "fat" face in the loomis image above (second row, last column) has a really fat neck which creates a fat fold where a thin persons jayline would be. i intentionally created an edge loop that swings from the eye, down through the check area and hooks under the chin for just such a need. conversely though i'm sure i've left something out and i'm also open to the possibility of needing a couple of archetypes. thanks for your suggested breakdown.

phreaknasty

12-19-2002, 10:37 PM

lunatique - after much scratching of my head and internal debates as to why i couldn't let it go... i was able to determine that the proper term is SUBMANDIBULAR as far as i can tell. and the guy below takes good advantage of it. :p

here is a fat version with no changes to "layout/topology"... only deformers and some hand-tweaking of verts:

http://www.geocities.com/~wet-nap/forums/cmw_fat.jpg

and the image i loosely based it on:

Lunatique

12-20-2002, 03:16 AM

Hahahaha. Guess you just found the proper name for a double chin? :D

megaflaizer

12-20-2002, 08:37 AM

wow, thats "phat" hehe.

i like the way the chin area and underneath adapted so well to becoming folds of fat instead of the trunk that accompanies a normal person.

so, you gonna pass this base mesh around once its done man??:cool:

after careful review of your posts, i thikn your right. youre wel on your way to creating a base john/jane doe to create char models from.

ThirdEye

01-03-2003, 07:33 PM

This is the coolest base to make every kind of head i've ever seen :surprised :thumbsup:

ThirdEye

01-04-2003, 11:43 AM

@phreaknasty: why don't you make a (video) tutorial on that? It would be great! :cool:

The Magic Pen

01-05-2003, 02:20 AM

Awesome work phreak :thumbsup:

phreaknasty

01-06-2003, 04:51 PM

megaflaizer- i hope you're right that i'm right :) . Honestly I had thought of giving some version of the head away but due to unforseen issues i'm not sure if i can, at least not anytime soon.
ThirdEye- i've considered it, but my process isn't realy that interesting nor perhaps my skills worthy of the effort. the useful part seems to be the end result, but as stated above i'm not sure if that will work out. the other problem is that the head is already done so i'd have to reverse engineer the process. besides that being really boring and time consuming, i'm not sure what value that would have, being such a contrived effort at that point. thanks for the compliment
Magic Pen- thanks

bjotto

01-06-2003, 05:11 PM

Very nice, just one thing:
I would change the flow around the nose to something more like this:

Joel Hooks

01-13-2003, 05:57 AM

I really like the results of your study. My favorite is the spiral around the ear - something i will try to incorperate into my own works. I've seen a couple people useing spirals as of late, and they seem to be a good addition in some cases.

The pudgy variant is great too, I'd like to see you push it into a few more of the shapes on that chart that was posted!:thumbsup: