3D-Blender Slowly learning blender [Feedback]

[rainbowpunfz]

I've been slowly learning myself how to do things in Blender, and have recently taken it upon myself to try and make a character from scratch. She doesn't have a name just yet but I am very close to finishing the entire modeling process, and plan to move on to rigging her soon. I've been staring at her for quite a while now and in the process might have gained some blind spots here and there, so any kind of feedback would be appreciated. If not, I'm sure you people of fine culture might appreciate the lewds anyway. I want to redo the hair, give it some more volume, and there is some more sculpting and coloring to do downstairs, but besides that most other things looks finished to me
View attachment bottom.webm
View attachment top.webm
View attachment full.webm

 

[snakeevabigboss]

How long did it take you to learn to use blender?
Also she would look better with her breasts being closer together in my opinion.

 

[rainbowpunfz]

How long did it take you to learn to use blender?

Learning to use it? doesn't take more than a couple weeks I guess. Learning the skillsets required to do anything with it, takes a fair bit longer. I've been playing around with it on and off for just shy of a year now. It really depends on what your goals are.

All major elements that have to be learned to some extend in order to animate a character would be: sculpting, topology, UV mapping, texture painting, (procedural?) shading, rigging, animating, and maybe some physics. Regardless of what software you want to use. That doesn't mean you need to master everything though, and there are certainly corners that can be cut. For example, the model I made is completely textureless, so I don't have to worry too much about also being a competent texture painter . But you could go as far as bridging DAZ stuff to Blender, as long as you know why and how the important stuff works.

There are quite a lot of things you need to familiarize yourself with and play around with. Things honestly looked quite daunting and scary when starting out. But there are plenty of very useful tutorials out there to help you along, and with each update to blender things seem to be getting more and more user friendly.

Also she would look better with her breasts being closer together in my opinion.

I tried going for something along the lines of Eva Elfie, and those magnificent bastards are quite a bit apart from one another. Maybe she needs a bit more volume or a bit less sag.

 

[Synx]

Its not bad at all, but take it from somebody that is actually trying to make a game in blender from scratch, creating your character from scratch is a complete waste of time. You are much better of picking up a base character (like from DAZ) from somewhere and morph that one your liking. The reason for this is that you will never get the same good topology from your own created character then from a base character. Besides if you are making multiple chars you save a lot of time as well.

I personaly used these:

You must be registered to see the links

for the body, and I used

You must be registered to see the links

for a male head. The first one has a ton of shapekeys to alter the body to your liking, and the second one comes with a shit ton of shape keys for expressions. As long as you don't alter the amount of faces used in the face and just morph the face around, the shape keys should still all work. I redo the skin and the hair but having a good base model with all kinds of shape keys is awesome.

This will really reduce the amount of time needed to create a character, especially with the shape keys (which is one of the parts i hate the most). I think you can even copy all the shape keys from the generic male to any DAZ head as i'm 99,9% sure Tyler generic male is made from a DAZ model.

 

[rainbowpunfz]

Its not bad at all, but take it from somebody that is actually trying to make a game in blender from scratch, creating your character from scratch is a complete waste of time. You are much better of picking up a base character (like from DAZ) from somewhere and morph that one your liking. The reason for this is that you will never get the same good topology from your own created character then from a base character. Besides if you are making multiple chars you save a lot of time as well.

Oh but where is the fun in that, I think I already said in here somewhere that simply porting from DAZ cuts tons of corners, (big asset library, standardized topology, etc.), but it doesn't force you to learn as much. If you just want to get a game out there, yeah makes sense, if not, this is a fruitful journey I think. I'm not saying this is the absolute way to go, but i wouldn't straight up call it a waste

As for topology, I can't say I'm displeased with the results. Base poly count of 20K, no poles in nasty places and the edge flow is generally going the right way. This might be easier to animate than the build a bod mesh, because of the reduced poly count. I know how and why things flow the way they do and where the detail is and isn't at, and that's super useful as well.

Spoiler: topology

You don't have permission to view the spoiler content. Log in or register now.

especially with the shape keys (which is one of the parts i hate the most).

Have you tried sculpting + face sets recently? it's absolutely amazing and makes making shape keys a lot easier. The changes made to multi-ress also help a ton.

 

[Synx]

Well sure its not a complete waste of time especially if you just want to learn. I actually made my first char from scratch and while i'm very pleased with the result, I just came to the conclusion my character was qua mesh not better then a DAZ character. For animations in my very limited practice I didnt noticed any difference qua speed uptill like 400-500k polygons. My current model has 67k without eyes/hair.

I personnaly just thought putting my time in parts which i can actually make it look better then DAZ assets, like the skin texture, clothing, or hair using the hair particle systems, etc. felt more impacting then spending so much time on creating a base model that might at best have a slightly better topology then a DAZ model.

Have you tried sculpting + face sets recently? it's absolutely amazing and makes making shape keys a lot easier. The changes made to multi-ress also help a ton.

I havent actually will look into that. Last time i did shapekeys it was just annoying as the mirror shapekeys part didnt work very well. The head from my link above comes with 155 shapekeys though from full faces like surprising, to minor ones like mouth shape for the letter O. I dont really see a reason why i wouldnt use those as I highly doubt i can do a better job without spending weeks if not more on it.

 

[mrkhing]

Looks good, I've also been learning blender and trying my hand at creating 3d characters. While DAZ assets are really good for more realistic characters, modelling and sculpting really is the way to go for stylized characters like this one. Regarding feedback, I think you could add a little bit of bump or displacement to the skin. That's a problem I've also been running into where the skin looks a bit too "plastic". Some bumps and wrinkles as well as some tweaking in the subsurface scattering seem to improve it to some extent.

 

[rainbowpunfz]

While DAZ assets are really good for more realistic characters, modelling and sculpting really is the way to go for stylized characters like this one.

Especially on here, DAZ stuff is a dime a dozen. You don't have to be Xbox if you can be Nintendo. This way I get way more toys to play with, and aren't automatically pushed up against the uncanny valley so much..

Regarding feedback, I think you could add a little bit of bump or displacement to the skin. That's a problem I've also been running into where the skin looks a bit too "plastic". Some bumps and wrinkles as well as some tweaking in the subsurface scattering seem to improve it to some extent.

This is actually part of why I made a procedural freckle shader, breaks things up a bunch for the eyes. I did some playing around with a voronoi skin cell shader and some procedural blood vessel stuff, but i don't want to push the realism too hard because i'm afraid it'll end up looking weird.. maybe some simple skin tone variation is enough? (btw, if you know of some way to do procedural goosebumps and have it look decent, let met know xD) I might return to skin shading at some point, if done right it's generally low effort for decent mileage.

 

[mrkhing]

This is actually part of why I made a procedural freckle shader, breaks things up a bunch for the eyes. I did some playing around with a voronoi skin cell shader and some procedural blood vessel stuff, but i don't want to push the realism too hard because i'm afraid it'll end up looking weird.. maybe some simple skin tone variation is enough? (btw, if you know of some way to do procedural goosebumps and have it look decent, let met know xD) I might return to skin shading at some point, if done right it's generally low effort for decent mileage.

Some additional texture painting to add blush, dark beneath the eyes etc. would go a long way indeed. Yes, for stylized characters you do want to be subtle in some details otherwise it might look grotesque. Maybe you could try creating bumps procedurally using a Voronoi texture scaled to a high number connected to a color ramp and try to get some basic skin pores by tweaking some parameters. However, I didn't manage to get very satisfactory results in this department using procedural methods so far but I'm sure that there is some node wizardry out there that can make it happen. What I actually do is sculpting the details on the mesh itself through a multires modifier to have a low-poly version (for rigging) and a high-poly version (for fine-detailing).

 

[rainbowpunfz]

Maybe you could try creating bumps procedurally using a Voronoi texture scaled to a high number connected to a color ramp and try to get some basic skin pores by tweaking some parameters. However, I didn't manage to get very satisfactory results in this department using procedural methods so far but I'm sure that there is some node wizardry out there that can make it happen.

Voronoi is actually pretty simple to do and have look decent

Spoiler: node tree

You don't have permission to view the spoiler content. Log in or register now.

you'll have to adjust the strength of the effect though based on how rough the material is, (or primarily rely on clearcoat for simulating wetness)

Spoiler: results

You don't have permission to view the spoiler content. Log in or register now.

What I actually do is sculpting the details on the mesh itself through a multires modifier to have a low-poly version (for rigging) and a high-poly version (for fine-detailing).

This is already happening, she's actually 20k polys + a multires. I might add some more levels to manually sculpt some more creases (lips fingers etc.) and bake those down to the current mesh ( I really don't need a couple millions of polys for rendering), but I'm afraid that'll end up being a lot of bucks for not that much bang and overcomplicate the design.

 

[rainbowpunfz]

redid her hair, improved the shader for freckles, and did an endless amount of tweaking of her face. I must say, it is hela hard to translate what's in your head into the right lines and proportions on the screen. I'm not entirely satisfied just yet with the way things turned out, so i might tweak it some more. Or maybe im just looking for excuses to not start rigging her face, because i for the life of me can not get her to smile without it looking... wrong.. Upper lip goes up, cheek muscles tense, lower eyelids go up, yada yada yada, and before you know it i've made her looks like she belongs in an asylum ¯\_(ツ)_/¯...

Spoiler: 360

You don't have permission to view the spoiler content. Log in or register now.

also, dope hair physics:

Spoiler

You don't have permission to view the spoiler content. Log in or register now.

 

[rainbowpunfz]

So, i was going to do facial rigging, but I got sidetracked, and ended up rigging some naughty bits instead.. Couple of issues here and there, but I'm generally quite happy with my first attempt. No shapekeys completely bone based

Knowing me i'll probably do the breasts physics before facial animation as well..

Spoiler: rig in action

You don't have permission to view the spoiler content. Log in or register now.

 

[rainbowpunfz]

So lets talk breast physics for a minute, (yes i was supposed to do facial stuff, which is also slowing progressing, but boobies). I could have gone the easy round and slapped some wiggle bones or deformation cage onto it and called it a day, but where is the fun in that? Also, i'm going to dub this post breast physics for autists

There are various systems within blender to simulate something, for breasts your options are going to be either soft-body, cloth (also known as the better soft body sim), or rigid body. I ended up using the rigid body system as that provides the most knobs and leavers.

View attachment boobtest4.webm

Boobs are essentially blobs of fat on the front end of a torso held together mostly by elasticity (duhh!), using the rigid body system we need something that reacts to forces (gravity, motion and G forces) in somewhat angular fashion.
As forces are applied to the breasts the skin pulls them back to the center. As breasts get bigger they become more 'limb' like and also slightly rotate. This is surprisingly difficult to do in Blender and as of yet I still don't have a satisfying way of doing this. The main problem is that everything is in x,y,z. Blender does allow you to set up a system akin to a

You must be registered to see the links

but the rigid body simulation is unstable.. so you have to convert stuff to polar and back to get anything 'faithful' to the math. while with smaller breasts most stuff is purely dominated by elasticity, bigger breasts usually have some more slack allowing for some free falling action. There is no clean way of specifying this in blender but we can use drivers to lower our spring values near the center and increase them near the edges. Last but not least, breasts also squish and squash, something that in theory a simple piston inside the entire system can measure (though i'm not entirely satisfied just yet with the results, the position of this piston might be crucial)

Okay, slap some bones to it and we're done? I wish..

I said bigger breasts act more like limbs, but that doesn't make them limbs. Skin still very much follows them around as they move.

View attachment boobtest2.webm

Another lesson that i learned is that multiple bones to deform the breast can lead to better results, and it is better to not have them be connected. This allows for stretch and squash in specific directions.

View attachment boobtest3.webm

putting all the lesons i learned together and the final result looks like this:

View attachment boobtest.webm

It isn't perfect but keep in mind that everything is done with just bone deformations, and ran purely on the physics, no manual corrections or adjustments anywhere.

 

[rainbowpunfz]

I got around to doing facial stuff finally lol. Turns out it was quite a challenge.. Some things look good static, or from one angle, but completely wrong when animated. On top of that, i'm not that big of a fan of the default facial rig that comes with rigify. So.. I learned how to write rigify templates, and made my own template (like any sane person would do of course, hahaha.. help me..) There are still some i's and t's to dot and cross, but right now it is at least functional. Once things are polished enough i'll probably go back to the rig i did for the vagina and do the same there. If there is interest i might dump the templates online somewhere for others to play around with.

View attachment 0001-0250.webm

The main difference between my template and that of rigify is in the mouth area, there is lot more attention given in how the mouth and the surrounding cheek areas deform. As a result it's doable to deform and look reasonable without the use of shape keys.

here's a quick 'n dirty render in Cycles:

View attachment 0001-0480.webm


I'm really fond of the eyebrows when they are just static, though they currently don't deform all that nicely, so'll have to figure out what to do with those.. Getting close to having created my own completely original model, once she is done and animated i should probably think about giving the poor lass some clothes lol.

edit: Fixed the brow deformation issues, cleaned up some sloppy keyframing

 

[Deeze]

You seem to be progressing well!

 

[rainbowpunfz]

Most of the rigging is now complete. I had to redo the breast physics, as rigid bodies don't link nicely between files, so it is now based on cloth simulations with forcefields, again driving bones which in turn drive the deformation. At some point i might play around with simulating the skin elasticity as the breasts deform (think stretching and squashing that happens to bags with fluid in them) since i have a relatively straightforward idea of how to do that. bone deformations alone are completely static, so adding such a simulation on top of it might help make things look a bit softer and more like bags of fun lol. But i don't yet know if it'd play nice with clothes so'll probably hold back on it for now.

The only thing missing is some jiggle physics for her butt, but that shouldn't be too hard to do. Shading also lacks some polish here and there, but this can be added without messing with anything else later down the line. Which means that pretty soon i can start working on building a pose and action library and seriously play around with animating her (and also give her some clothes at some point lol)

here's a rather lazy animation with breast and hair physics:

View attachment 0001-0157.webm

 

[ddakkal]

Hi, nice learning path you got here, you managed to do some amazing stuff!
I recently got into character modeling in Blender and would like to ask how did you model the genitalia?
Like, can you point to some resources/courses/articles or just describe the process of how did you learn to do that?

Thank you for sharing you learning experience!

 

[rainbowpunfz]

I recently got into character modeling in Blender and would like to ask how did you model the genitalia?
Like, can you point to some resources/courses/articles or just describe the process of how did you learn to do that?

Look at a lot of reference material (hehehe..) and focus especially on what kind of proportions and shapes there are and how they connect together. For instance, note that the place where the clitoral hood and labia minora meet, varies between girls, for some the hood connects near the top of the labia, and for others it connects lower and off to the side. Once you have a decent understanding you can sort of chose where to stylize and/or simplify the design. for proportions it might also be helpful to simply look at a variety of

You must be registered to see the links

(jeej for science). Also take note how and where the genitals are situated, otherwise they might look fine on their own but not in combination with the body as a whole.

Genitals are somewhat complex in shape, so I prefer to just sculpt with dyntopo and worry about topology afterwards. When it comes to topology, chances are that you need a lot more of it than you do in the main body, for woman a good place to create all of that additional topo is in between the labia minora and majora, and if you need a sink you can always hide that inside the vagina itself.

I still have to figure out how to do guys, but the main complexities there are more in the animating department rather than the modeling.

 

[rainbowpunfz]

Still alive, just haven't been working all that much on this project. Felt like i should maybe throw another post up or something.

Since last I posted a lot of my effort has gone into polishing minor things here and there, and deciding on how i want to tackle the subject of clothes. I have a proof of concept in the works that should be done soonish. The goal here is to create clothes that behave believable while still maintaining reasonable artistic control without needing to spend exorbitant amounts of time making/animating them. The way to do so, i belief, is a structured workflow and clever marriage between physics and animation systems such that one can enjoy the benefits of both. Ultimately it should be easy to animate a scene in which a bra is removed by unclipping it, sliding the straps from the shoulders, and let it fall to the ground. Incidentally, that is also the animation i'm aiming for with the proof of concept.

I'm also getting a bit more practice with the animation side of things, and it feels like I'm starting to get the hang of it. Here's a walk cycle she'll be doing while taking of her bra:

View attachment 0001-0250.webm

Expect a properly rendered video with that cycle and a bra coming off soonish

 

[rainbowpunfz]

View attachment bra_test.webm

There is still some minor clipping happening here and there as a result of the mesh being squeezed on both sides by collision boxes. I've some conceptual ideas on how to solve that problem, but i need some more time to think/tinker with things before decided if an how. The bra itself is quite a simple model, but can be made more complex with the appropriate shaders/textures/surface deform. All in all I'm decently happy with the result (don't pay too much attention to the face, I was too lazy to add facial animations) In principle the techniques used here can be extended for most other pieces of clothing, the only exception being things like hoodies with zippers (which i intend to tackle next). If anyone is interested, just let me know and i'll cobble something up to show what's happening behind the scenes.