I hope this is the correct forum. I didn't want to post to my gallery because this is a work in progress.

I present my humanoid female version 0.9.0 for your criticism. It is the first complete mesh after three prototypes created over a six week period. It is heavily inspired by the DAZ unimesh and addresses what I feel are several shortcomings within my Renderman pipeline:

• it has a relatively low poly count.
• designed for rendertime subdivision smoothing
• rigging flexibility and ease (weight painting on smooth mesh or wrap deformer)
  - uses less system resources for SSS (using 3d "brick map" texture)

During its creation, I have found it easy to visualize other animals and I am looking forward to creating a dog. I also find myself studying (real) people more intensely and thinking of ways to capture their features onto the mesh. Next I will sculpt it into a human male. There are no plans to distribute it at this time.

I know the ear is crappy but I will be interested to hear your opinions, which I will use to refine it. Thanks in advance.

Hey replicand, this is looking great. Are you working from reference images? Where did you get them?

How do you plan to use the model? Even if it is just for practice, are you targeting it for a certain type of use, like games, film, stills etc.?

I know you use Zbrush, are you planning to finish this in zbrush? If you take it into zbrush, will you use the proxy or the smooth?
Will the triangles on the forehead and neck render okay after the smooth?

Thanks.

How do I plan to use? The answer is two tiered. Feature length animation, for sure. Rendering speed is key, so it drives the choice to use Renderman, which dictates the mesh's construction. Second, following the Unimesh philosophy (which IMHO is elegant), this is the base from which other characters will be created. Textures are no big deal per se since they will be procedural but the weight maps will transfer from character to character, which will save a bunch of time. I have already created a child character from this mesh. It took ten minutes.

Late in the modeling phase I noticed that moving vertices had little effect, so a sculpting program is the next step. I have ZBrush, but the Mac version of the Mudbox trial is now available and it rocks! I'm currently scraping pennies together to ditch Zbrush. Though I haven't decided yet, the "correct" way would be to use the low poly version in (sculptor program of choice), but I'll try both ways. I don't imagine it will make a difference.

Reference images are standard, unmorphed muscle mapped V3s, front, top and side views rendered in Poser. I've always felt V3's shape is weird (small head, big hips) but I wanted to concentrate on generating a clean mesh that I can easily modify. I'm a big fan of fashion magazine, so there's a plethora of faces to work from.

I'm still wondering about your reference images. I remember you mentioned that you had a new camera. Do you make your own reference images from people you know?

It seems like I read that the original mesh will sometimes need to be retopoligized based on changes made in zbrush. I have not use zbrush so I'm not sure about that, but if it is true then it might make more sense to use the proxy.

What is it that you do not like about zbrush? Is it the gui? Is the gui in mudbox very different?

replicand:  Looking good and very low poly suitable for game dev.
Ears are tricky and hard, I had a terrible time. There is a great tute for ears in ZB i will try to find it.

Careful! It's addictive. :biggrin:

Looks like a great start. Good idea to develop your own mesh - there are some really horrible tutorials out there - and to keep it low in polygons. You'll want to rewire often as you go, and that's so much easier that way. Get the shape right and then refine some portions only if you absolutely can't get the necessary detail otherwise.

Studying people is very helpful, too. Photo references keep you honest while modeling when you don't have a patient life model at hand, but nothing beats the understanding you get from observing actual people in motion.

Also, if you haven't done any life drawing before, I'd recommend you start now.

Noooooo! Not life drawing! I took a life drawing class a couple years ago and I hated it. I felt really uncomfortable for the model, if that's possible. I will agree that studying the real thing is valuable.

The ear work will be this weekend, probably a series of extrusions rather than two separate bridged models. Seems the more often I do things, the easier it becomes.

re: making my own images isn't that easier. I have very few friends, none of which would put up the humiliation of being an experiment. I even put an ad out on craigslist so that I could measure the people's proportions (I wanted to know where short people are short). In any case, I will photograph myself and reshape the mesh in my likeness so I can do all the things I wouldn't do in real life - like jumping out of an exploding helicopter.

Quote - Noooooo! Not life drawing! I took a life drawing class a couple years ago and I hated it. I felt really uncomfortable for the model, if that's possible. I will agree that studying the real thing is valuable.

Well, when I started with the life drawing, it was in a small group, and we took turns being the model. I guess that helps with the feeling uncomfortable. :lol:

If life drawing and taking measurements of real people is out of the question: I assume you know about 3d.sk and such?

I never volunteered for being the model.  I prefered to draw.  But I was always being watched behind my back by the teacher while drawing.  No one could see anyone else's drawings though.  I don't use drawings for reference.  I always have to have a photo reference in my hand while modeling.  I refuse to trace for some reason.

 Yeah, I'm better at drawing non-organic objects but the life drawing class showed me that I can draw organic objects; it just requires lots of patience. I have oodles of patience but I prefer the technical (programming) side of things, so I'm more likely to use a camera.

Speaking of which, working on the ear right now. I shot myself with my camera, and seeing my ear in "orthographic" projection reveals a lot more than trying to feel the contours. I will post the results soon.

And just to give perspective, there's a classical pianist in the next building. Sounds like he's playing Rachmaninov.

Quote - I never volunteered for being the model.  I prefered to draw.  But I was always being watched behind my back by the teacher while drawing.  No one could see anyone else's drawings though.  I don't use drawings for reference.  I always have to have a photo reference in my hand while modeling.  I refuse to trace for some reason.

I agree that drawings do not make good references. The reason I recommended life drawing is that it forces one to look very closely and interpret what one sees, which leads to a better understanding of the human form. Sculpting in clay might be even better, and because it's an inherently three-dimensional process, one can do it from photo references (just find a set with the same model in the same pose from various angels).

And yes, tracing is evil. Never, ever trace. There is no orthographic camera in the real world, so your photo references will always be distorted.

PS: No one could see anyone else's drawings? Man, that's serious business. All the life drawing I ever did was in rather relaxed, non-academic settings, so there was no question that we would look at each other's drawings. Usually the models would look at the drawings, too. Personally I  think one should not be allowed to draw or photograph naked people unless one is willing to take on the role of the model at least occasionally. But that's just me.

Yes, I'm aware there's no such thing as a real world orthographic camera, though you could simulate it very well using a telephoto lens with narrow field of vision from far distance.

Anyway, here's the new ear 50% complete. I deleted the "ear core" and all surrounding edgeloops and built it as a series of extrusions. 

Yes, the ear looks much cleaner now.

 Thanks. Still working on it. Will post improvements late tomorrow.

The good news: that's the same figure, although of course with lots of sculpting and rewiring in between.
The bad news: it took me three years to get from left to right.

Just so you know what you're signing up for. :biggrin:

(Of course conventional wisdom says to scrap your early attempts and start from scratch several times before you actually try and go the distance. It makes sense, because that's how one gains practice and experience. So don't feel tempted to imitate my craziness.)

I still haven't done ears yet on mine.

 Yes there's a marked improvement between your two meshes. The crazy thing is that the detail level / poly count appear the same, its just the placement that's changed. 

Three years ha, I laugh at you! I'm an obsessive compulsive vertex slinger with no money, no family and no social life. All I have is time. Oh no, that made me sound really pathetic.

I have no problems starting over again but I'm interested to see what obstacles and shortcomings I find during production. Then I can target the problem areas on the next mesh.

You should start your ears. Be like Jack Sparrow and run into the mouth of the Kraken.

Hey, I remember that model.. nice improvements!  .. After I scrapped mine, I didn't go back to it. 
Nice going Replicand... but gotta be honest, the shape of the head is bugging me.  It's like the sides of the head are trying to create a reversed V shape,... like a roof on a house.
Don't know if you'd find these helpful.  I'm not really fond of the shape of Kryoboy's ear, but I do like the edge flow. 
http://forums.cgsociety.org/showthread.php?f=7&t=108412&page=149&pp=15  - Kryoboy's
http://forums.cgsociety.org/showthread.php?f=7&t=108412&page=135&pp=15 - Stahlberg's

 Divine Rain, I'm not looking for praise I'm looking for honesty so I appreciate the fact that the head shape bugs you. I'm too close to see it objectively. 

Your two links are a goldmine. I never go there. Maybe I should start.

I've just realized that I'm not even close. I've just seen a mesh created by Lee Wolland which has much more relevant and well placed detail that my mesh doesn't even hint at. Rather than start from scratch, the body will receive heavy modification.

Pic is the latest ear refinement, which I will put on the backburner for now.

Hey DivineRAIN! Fancy meeting you here. :biggrin:

That CGTalk thread is definitely a must-read. I also recommend subdivisionmodeling.com, which has some great tutorials and other interesting stuff. The site is more specialized and less glamorous, which means that the relevant bits of information are easier to find, and people tend to be a bit less pretentious.

 ODF - I'm not worthy.

subdivisionmodeling .com is awesome!!! On the surface (no pun intended) it seems strange that there's no gallery. Otherwise, I really wish I would've started this (whole undertaking) sooner.

k, glad ya liked the links and you're ok with critiques. 
There's another good (and old)  thread on cgtalk called topology research. 
I searched the thread hoping this image was there.  What I like about this topology is the blue part because I tend to think using a topology to define the jawline makes sense.

forums.cgsociety.org/showthread.php   the first page, just 1 wire shot is left... a lot of the older posts have missing images.  But the image that's left is pretty good.
forums.cgsociety.org/showthread.php  nice edge flow in post #411
forums.cgsociety.org/showthread.php  Gollum wires..

 Hey, I really thank you guys for all your assistance. I think I'm going to freeze development at this point because (a) this is my first humanoid, (b) I want to see what obstacles I will encounter in a production environment and (c) ODF, I value your opinion but Divine Rain - you have pointed me towards a lot of material which will take time to absorb and integrate.

I guarantee that I will share my improvements in a future thread entitled "What I've learned since my first humanoid"

I'm looking forward to that future thread. Absorbing and integrating are good things.

By the way, if you'd like to have a play with an actual mesh and see if and how they implemented all that theoretical knowledge, I recommend Victoria 4 from DAZ. I think it's still free, and for once they did a relatively decent job with the topology. As male figures go, there's also Apollo Maximus. There may be other free models with good topology, but I'm mostly a Poser person, so that's all I know of.

Finally, if you enjoy watching the unwashed masses struggle: phantom3D and I both have WIP stuff in our file lockers here that you can download, and there are number of threads where we discuss them, although usually more focused on rigging and Poser-specific things and a little less on modeling.