Thread 1009306

A thread for work in progress of your projects, for recently finished projects or things you'd like critiqued.

Previous: >>1006887

Don't clown me too hard, first time trying to model a character (Laughing Octopus from MGS4)

>>1009306 (OP)
I'm planning on tweaking the anatomy a bit as well as the chin since it got a bit distorted when connecting to the neck. Mostly curious to see if others think the proportions look okay or if that could use more tweaking as well

>>1008959
Debuted!
https://www.twitch.tv/videos/2405743023?t=0h9m45s

>>1009306 (OP)
Use the fucking WIP image fucking newfag

Since I mentioned gigachad last time I decided to put him on the chopping block next. Still a "miss" on the likeness part, maybe even worse than normal because I didn't outright trace a camera-matched profile view this time (even though there is a pretty good photo for it) and eyeballed the entire thing. But I had fun sculpting his ridiculous body, so I decided to start retopologizing and rigging the guy. Only did the head so far as you can see in one of the pictures, but I'm suddenly not sure if I should proceed with the retopo or if I should try to match his face more correctly, at least by tracing a profile view as I normally do. And maybe fix any anatomy mistakes that you can see, of which there are probably a lot. What do you guys think?

>>1009313
Post the blank and then we'll use it.

>>1009351
Here

I fucking love MIGGERS :3
(I know I forgot the ventral fins)

>>1009356
it is circumcised

>>1009306 (OP)
Fixed the head and the chin. I think it looks less like an alien now

2 threads in a row without the proper OP image. /3/ really is fucking dead at this point. Jesus fuck. Tourists really have taken over.

>>1009390
like I said, most anon doesn't have the original: >>1009352

>>1009437
It's not hard to search the archives if people actually gave a shit.
https://warosu.org/3/?task=search2&search_subject=%2Fwip%2F
Search /wip/ in the subject field and bam, there it is in droves. Quicker than even posting a new thread.

>>1009446
>if people actually gave a shit.
you figured it out

>>1008912
Finally had some time to work on this more. I definitely need to clean up the edge highlights and fix seams, but bump mapping over the UV layout for bakes worked like a charm.

This course I'm going through does some really weird choices sometimes, like... Why there are houses built on mossy stones in the middle of the ocean, and more importantly how tf are trees growing from mossy rocks. Nevertheless, I've learnt so much useful shit already it was totally worth it. Gonna do some trees now, and the next lecture will be bushes I think, which is another weird thing to see growing on rocks.

Do you guys know how to change the color of shadow in PBR material? I mean I know there's ColorToRGB, but I only ever used it in NPR setting where I simply could output it to ColorRamp and configure the whole palette there. But here I need PBR material with all it's colors so it's kinda not gonna work, at least not that easy (I can't configure gazilion colors in ColorRamp). Sounds like I should try to decompose h/s/v components and mess with value, but that doesn't sound fun or easy either. Not sure how to recolor it even if I do that.

From the vag thread. Kitbash of different models, and retopologised parts. The hands seem way too long

The goal now would be to add a metahuman head so I can animate it easily and then add shape keys for customization.

I broke down and bought an "anime shader" for unity hoping it would help me on my path. I spent a long time making my own shader that could handle this 2d eye texture thing I want to do. So then I had to try and integrate that into this new shader i payed for.

>>1009444
.....update

>>1009496
Maybe just tonemap it? Why stop at shadows?

Vegetation is kinda meh from this distance, gotta find better textures I guess. Placing things by hand turned out even more grindy than I anticipated.

Another attempt. Those are kinda normie looking, but it works better imo.

>>1009525
I'm giving them a bit alpha now to soften the shadows, but full-image postprocessing is a valid option too I guess.

>>1009561
desaturated

Love looking at matcaps.

Had a mental breakdown and just now I've resumed work on this.

>>1009563
>>1009566
It's really damn boring to render the same view after every set of changes. As a motivation booster I should do renders from all kinds of angles instead of just one.

>>1009581
those are good for art inspiration

A bit different angle + some lights + fake evening.

This thing rendered for like 40+ minutes, for this shit quality :<

Now that teacher made me work on lights, I want to go somewhere in-between day and night cos I want day, but I don't want to waste lights. So it's kinda early morning I guess, or evening. Microtweaking stuff via randomizers. Randomizing hue/saturation of vegetation and tree leaves, randomizing emission of windows to make it seem like some rooms have lights on, some have lights off, and some pass lights from other rooms, lights in lamps are now point lights through translucent material instead of just emissive material. Give me some advice if you feel like it.

>>1009653
Another angle.

>>1009581
Looks nice, do you envision this character as MC or more of an enemy type?

>>1009645
I like it! 40 minutes is nothing, you don't even have anything silly like depth of field on.

>>1009669
Looks like a ninja turtle villain

>>1009581
looks like darksiders

Trying to learn actual texture painting

>>1009850
Sir this is /3/

>>1009849
you seem to be figuring it out

Sick as a dog and back to staring at this handsome fella for hours. Feels like the likeness is finally landing, but tomorrow me is probably going to disagree.

>>1009669
>>1009587
>>1009699
Thanks

>>1009674
Indeed it does look a bit like Shredder I guess

This is based on one of the Black Desert hashashin armor. I was asked to do the body the same and change the helmet. Wip.

>>1009468
Revisiting and cleaning up an older model... I think I posted it in a WIP thread before, but I don't remember.
I feel inclined to try to get this goose under 1K tris for consistency, but I don't want to lose any sense of him being a pudgy fatass - because that's funny.

Anybody remember the guy who was creating a human base mesh entirely in geometry nodes? What happened to that guy? How far has he gotten since his last post?

>>1010062
certainly he just went insane like the badguy from mystery men trying to invent his death ray.

In my opinion Blender is kinda nice at 2d vector art. That's even without grease pencil. I tried a bit doing vector in Inkscape and messing around with beziers took forever.
Just moving vertices around on a plane in Blender is 1000x times more productive, and you can always subdiv it later. And with grease pencil you can do 100x more.

>>1010088
oh for sure, this video is a very good show of it https://youtu.be/Xx2Pu09VnWA?si=k6IL-XoiEGLERZYK, nowadays all my UI work is pre-rendered from blender

This knight character of mine

thoughts ?

>>1010280
needs more detail

I think it's done. Just need to apply color

>>1010371
Looks great, looking forward to colors!

>>1010362
Very shiny. I think you need to turn the roughness up

>>1010372
Thanks

I think this is it.

Made a tiny bit of progress but I'm proud because I've been stumped on this for a long time. I figured out how to make a corner sample the index of the neighboring corner on the opposite side of an edge.
It's one of those things that's small but difficult to wrap your brain around. It's only a building block however. A corner stone required to make something more interesting.

>>1010380
>>1010381
Very nice, love the end result

>>1010380
Yeah boi we go golden! Awesome

>>1010392
>>1010440
Thanks

I've started on the female human version.

>>1010361
>>1010362
Good model, but I agree with the other anon that it's too shiny. Most MACs were given a very matte finish.

>>1010380
>>1010381
Good shit man, it's got somewhat of a middle eastern feel to it

nanachi i'm working on, I dont know if im going to be doing any more sculpting or if i should just move on to painting and posing. I was thinking the body is a little too small compared to the head but i'm not sure.

>>1010534
love it, I think the legs need a bit more volume

>>1010534
Impressive. Is this going to be 3d printed?

>>1010543
thanks! and you might be right, looking at some references the pants are almost like two spheres
>>1010571
possibly, but i only have an fdm printer and I've never printed a full character like this, not sure if it needs to be split into pieces for printing.

an elbow that collides with itself so it forms that curve in the skin folds, but i dont really know if this is the correct way to do it lol 1/2

>>1010600
basically, 1 bone sits on the Y-axis of the arm bone and has a sister bone on the forearm bones' Y-axis, they track to each other, and then have "limit distance" modifiers to each other with less than "1" influence and they will both give way to each other to create the effect of "collision", a 2nd bone copies its location, and a 3rd holds the deformation influence parented to the 2nd with a "stretch to" modifier, so the deformation is smooth. They need to sit perfectly equidistant and to get that location you can just make a circle with the center on the "elbow" joint, the edge along the circle is the exact point of the sister bone. Don't know if this is a correct and legitimate method to do this.

>>1010601
Does it get you a good working result? Then it's correct and legitimate.

>>1010602
im not sure how to get a more refined result without adding more bones and a higher density mesh, which makes me wonder what the industry standards for collisions like this are

>>1010608
They use volume conserving cloth simulations, but I also read somewhere they came up with a slick algorithm for elbow pit skin in tangled.

>>1010611
The tangled paper is there but it's not something you're gonna work into your scene without a team of engineers:
https://la.disneyresearch.com/publication/efficient-elasticity-for-character-skinning-with-contact-and-collisions/

>>1010611
>>1010632
ty

>>1010012
Decided that I'll make the armor and goggles separate models, both for convenience and as a cope for the polycount. Those will be next.
Feathers were done by stencil painting a bump map then baking lighting with it.
Beak is just a photo composite.
Feet look a little odd palatized, so they'll be tweaked a bit.
The pink on the wings is the color Quake uses for transparency.

>>1010601
That guy might be relevant to you, got a billion interesting deformation setups
https://x.com/sakuramochiJP

>>1010499
Done

Test

>>1010716
success

I haven't hit PEAK stupid yet, i'm open to suggestions for fixing this issue in this scene.

I really miss that procedural skin/flesh dude, despite him getting called "the raw chicken meat guy" i knew he was a true genius, i wish him success on whatever endeavour he's dealing with at this very moment.

>>1010765
I'm just dropping these two and fucking off for another month to keep the grind going.

I hate retopo.
How hasn't this been solved by AI yet?

>>1010088
That's not even a real opinion.
You just don't know how to use Inkscape.

>>1009997
>>1010371
>>1010380
Did you retexture a SC2 Protoss?
Looks cool

>>1010772
>How hasn't this been solved by AI yet?
Because it's been algorithmically solved for around 10 years...

>>1010772
Just stop being a sculptfag and gitgud at subd.

>>1010775
No, it hasn't.
I hope you're not referring to Quad Remesher.

>>1010776
The 90's are long gone, Anon.

>>1010765
Welp actually i'm gonna drop another one and ask if i'm in the correct path, i know jack about character design so idk if this gives the same vibes as my main inspiration ;__;

Peace.

>>1010772
Done
Time to rig

>>1010653
The shadows on the feathers are too dark

>>1010713
sexo. good job anon

>>1010795
Very nice. I know it's not the same but the first thing I thought was King K. Rool

>>1010795
That looks awesome!

>>1010816
Thanks!
Yeah, I had a friend say the same, and I see why.
I have made myself a little project of making a set of characters that could work in a 3D platformer, so I guess it fits.

>>1010819
Thank you!

>>1009306 (OP)
made a super low poly C-RAM for the android jet fighter game, i'm developing with my brother

>>1010824
the C-RAM firing ingame. works fine an very low end devices (lenovo moto g8) btw

>>1010825
oops forgot to add picrel

>>1010826
>>1010824
Nice, reminds me of the latest anti drone platforms. Terrahawk paladin, flatbed trailer with a big gun and all the radars you'd ever need.
Does your engine support transparency? I'm looking at the barrel truss

Here's my current bullshit, drive train for a futuristic vehicle. I don't know what's the end goal yet, either something like a tank or Mako from Mass Effect. I try to keep it clean but worry not about the geometry too much, i'd like to render basic shaders and doodle over it in photoshop.

Anything I should be adding before getting to rigging and animating?
I've tried doing clothes creases and shadows to give it more detail, but it never really looks right no matter what I do.

>>1010961
thanks man. yes our engine supports transparency but I didn't wanted to use it on this model because you probably never come enough close to it to see it and we want to save every bit of performance.
Wow those wheels look really cool and they capture the "actual working sci fi" vibe pretty good. I think some rocket artillery would look really cool on it

RIP sticky

>decided to try making a Transformer (that ideally once its done will actually transform)
>realise I have no idea how to design a Transformer
I need more interesting stuff to put on the robot mode, maybe the problem is I chose such a simple vehicle basis.

cool sword I made for a commission

>>1011154
that is a cool sword

>>1011031
Fantastic looking model
Love how u have seperated the mesh for Easy rigging.
What does topology look like?

>>1011158
Thank you!

It's pretty low poly, roughly around what a PS2 or early PS3 in game character model would have I think, though I didn't aim for anything specific. Textures for the face and hair are 512x512 and the main body is 1k.

>>1011167
Here's a few shots of the finished model. Could probably do with some better posing and angles, but I'm very lazy

>>1011031
>>1011167
>>1011168
love her, so cute, for a game project?

Sup /3/, amateur here, I'm focused the shoulder/arm, this is my best/latest attempt at modeling/rigging the shoulder, the shoulder/upper arm bone damp tracts to the rest of the arm, and seems to deform much better than before. My issue now is that the excessive geometry collapse in the arm when both the elbows facing inwards and when the arms extends above a certain point.
I figure I could just split that bone into a shoulder bone and upper arm bone to reduce collapsing, but most examples I try to reference don't seem to do that/need that. Thoughts?

>>1011167
What software is this? Doesn't look like blender

Made some nice progress over the Hackation.
Beginning to grasp how to manipulate curves procedurally. Still not *fully* grasping it. But I can sort of lay curves along a surface, and make them spiral around a normal. And raise and lower the tips. So I shaped them into this short hair cut, with a bit of waviness to the strips.

After converting the curves into mesh strips, I then figured out how to apply a cylinder SDF to those strips, and set up the bump properly. Doing a bunch of math to make sure the cylinder remains proportional to the dimensions of each strip. Even when the strips are at random sizes.
Then the cylinder is repeated to fill out the strips, again, keeping perfect proportion. I'm very proud of this. It took a great deal of effort.

Then I figured out how to use the white noise node to create different values per strand inside of each strip. Using that to create variation in cylinder length, color, and bump intensity.
The length variation really helped to make the hair look less like uniform stripes, and more like actual hairs.
Then I use the white noise node trick again to mask larger chunks. Multiplying them with the individual strands, thus creating sort of "clumps" of hair. So it's not all randomly sized. It's a clumped random.

There's still more work to do. I want to figure out how to make the cylinders curl and wave, while maintaining proportion inside of the mesh strips. That's going to be the next big hurdle.
Also, right now, the hair consists of 96,000 triangles, because I've spread 2000 curves evenly across he skull. I will be able to reduce the number of curves by more strategically laying the hair to cover more surface area. I just haven't gotten around to doing that yet.

>>1011168
you got an art page or something? I'd like to see more progress or talk to you about this!

Remade my gosurori jackal girl a year later cuz she was the first 3D character I ever made. New model has half the polycount of the original

>>1011218
Unf.
How did you do the har?

>>1011221

Thanks.
Started with a rough blockout
then I would slice pieces off from the main chunk and dynamesh - remesh them slowly refining & cleaning up the shape - hair is a pain in the ass.

Another way I used to like is take a piece of mesh with desired profile + curve deformer - you can twist and flatten in in desired areas.

But this time I was too lazy to set that up.

>>1011218
Always appreciate seeing mob faces in 3D

>>1011167
>>1011031
That shit is beyond fire.

Fission mailed, we'll get em next time.
༼ つ ◕_◕ ༽つ

>>1011266
Thank you!
>>1011177
Cheers. It is for a game project, though it's basically a reference model for drawing rather than an in-game character model since the game is fully 2D. I want to have very detailed pixel art like in old PC-98 games, but keeping things consistent at a bunch of different angles is tricky, so I ended up modeling the character to pose for certain shots.
>>1011188
It's Wings 3D 2.1.5. Very old and barebones, but I find it's a great tool for doing low poly modeling fast.
>>1011203
Thanks! I have an old ArtStation account (@ wavelabs1) with a few pieces, but there isn't anything related to this project up yet. Soon, though...

Here's a few angles of the sidearm that's holstered.

>>1011218
Love this. Very nice model, anon.

>>1011405
Here i come Constantinople.

>>1010580
>not sure if it needs to be split into pieces for printing.
Yes, it does. You'll need to cut it up and make keys that fit together after the parts are printed. Theres a ton of videos about it on youtube

>>1009989
JD Vance?

>>1010267
a cute

>>1010534
posting an update on this, mostly done with the project now. first time i "finished" and posed a sculpt like this

>>1011563
Simple but effective. Cute.

>>1011579
thank you

I'm trying to replicate some sort of modern day version of an n64 game

>>1011725
Scary swolegator, nice.
On my end i ended up with something similar to a manned mars rover, crossed with a diesel loco.

some goblin I made for a game I'm working on. I'm quite new to 3d stuff, so decided to go with segmented limbs so I don't have to deal with deformations.

>>1011821
A decent goblin ,
well done.

>>1009306 (OP)
I have to say im kind of proud.

>>1011824
this one is good too.

>>1011823
thanks! I'm also working on this harpy. Thing is I'm not sure how to approach a low poly wing. I've basically divided it in 4 vertex groups each one with it's own bone. The edges you see selected in picrel are the limit between one vertex group and the next, with on loop cut on each side to facilitate deformations. Is this a decent approach? It looks kinda fine I guess but I wonder if it could be done way better somehow.

>>1011827
also here's the gal taking off

>>1011828
the legs are actually hands?

>>1011841
I based it (copied from) the Dragon's Dogma harpy design

>>1011824
I was expecting the letters to change as they spun, so that they spell out "NIGGER".

>>1011864
Nothing that advanced yet im working on transparent animations for 4chan in blander. Im exporting and importing models and animations from CS1.6 and i want to make few things but it will take some time.

>>1011885
*blender...
blander kek

>>1011845
based taste

>>1011792
Very cool design

>>1011889
lol, thanks DD bro. I really love that game, hope I get to make something half as cool as it some day.

>>1009306 (OP)
generic polymer pistol for the gooner scalie fps
>>1011436
nice.

been enjoying kanyes music, what does 4chan think of this model

Progress...

made this cute little cart in [spoiler]Trenchbroom[/spoiler]

>>1012045
>>1012046
>what does 4chan think of this model
Even putting the cringe of anime nazism aside, the design is an ugly, incoherent mess.
Then again, that's anime in a nutshell...
>>1012073
>>1012059
Are you doing full sub-D autism with supporting edge loops, or just bevels and marking edges?
Looks great either way.
>>1012073
Based and broompilled.
I don't see TB used very often on /3/, but there's a few Quake mappers in the FPS thread on /vr/ that post about it sometimes. Very underappreciated program.

>>1012089
i'm not a subD topology wizard so i just went with bevels. specifically, i'm using a bevel modifier and marking edge bevel weights so i can still preserve the low poly mesh, it's a nice and non-destructive way to define the subdivided high poly meshes without directly fiddling with the low poly one.

>>1012089
Thanks anon, went to /vr/ and asked the fine fellas there for help and already gave me some clues on how to improve.
>Very underappreciated program
Yeah I never hear about anyone on the gamedev threads talking about it but I think it's amazing, specially for making retro-looking games. Simple to use and fast to come up with results.

Working on a base for the female NPCs. Don't mind the head, I still haven't worked on that yet.

Working on multiple different things at once

>>1012436
This one is almost done but I wanna make it much more complex by continuing to add to it

>>1012435
Are you making a MtF nazi granny?

>>1012438
BBW*

>>1012439
So a MtF BBW granny? Sorry anon, to many fetish for me

>>1012045
>>1012046
Cris died but jesus gave us a new one, praise the lord

>>1011821
This is kinda good, keep going anon

>>1012442
Appreciate the compliment, thanks man.

Low poly model I just finished texture painting, now comes the tedium of rigging and weight painting.

>>1012434
I basically finished the base character sculpt, now I gotta figure out if I should do a clothing sculpt then retopo or retopo and then sew clothes or something...

>>1012482
Rigged and polished a bit. I think my topology could use some more care around the joints or maybe there's more to be done with weight painting, but it's satisfactory for now. Will try my hand at a little animation soon.

>>1012482
Good job you should pixelate it and turn it into a PS1 style model though :3

>>1012497
Looking nice. Ignore the moron suggesting to pixelize it. He's a hater trying to sabotage the best part of your work: the painterly textures.

>>1012498
shut the fuck up retard.

>>1012497
looks sick, love the elongated face and style of this

>>1012513
You're mad because he's right

>>1012512
>look like an extremely aged out of date world of warcraft style model
>look like a modern, currently trendy but overall timeless ps1 style model
gee i wonder whats better

>>1012530
Both are timeless because both are not trying to be realistic.
Any styles other than photorealism are timeless.
Only in photorealism something can get objectively obsolete simply because there is newer better more photorealistic tech.

>>1012530
>Every game is painted textures is WoW.
This is your brain on AI. Shut up, Cris.

>>1012482
I like it, the texture work is really good, but the proportions are too extreme
>>1012455
I like the style anon, looks like kingsfield, where I can find more of your art?

>>1012482
>>1012497
Greatto jobbu anon-kun!

>>1012036
>>1012059
>>1012092
it's alive!!

>>1011215
Holy shit, that's a year of progress? Where did you learn to texture so well in such a short time? I'm impressed

>>1012484
Needs clothing including shoes, hair, and texture now. I might even do a few minor tweaks to the model, I was told it looks a bit bubbly.

>>1012579
also doing cloth sim for clothes

>>1012482
That's fucking sick, last time I did painted textures I did this lol

>>1012547
Great job anon
Did you make the hands and animations yourself? Or did they come from a pack off the Asset Store?

>>1012579
Have you seen a human in real life?>>1012581
Is that supposed to be Cris in a nazi outfit?

>>1012497
He has really tiny legs, looks so funny

>>1012602
He's just muscular and no idk who Cris is

poly modeling is fun

>>1012647
it is.

spitballing armor designs

>>1012536
>I like the style anon, looks like kingsfield, where I can find more of your art?
many thanks anon, really appreciate the kindness. No, there's nowhere to look for more of my stuff because I've barely done 3 models so far for the game I'm working on. Hopefully I have more stuff to share soon.

>>1012652
Much talk for someone who can't make a donut

>>1011436
I like this style. Both the model and the texture look amazing.

>>1012698
I can. I will not!

>>1012498
>>1012512
>>1012518
Thank you thank you, I appreciate the kind words :]

>>1012536
You really think the proportions are extreme? I made the legs very muscular but I don't think anything is particularly over-exaggerated.

>>1012603
They're not tiny they're just covered by his tunic you silly.

Spent most of my time grappling with learning the animation system and its interfaces. Does anyone know if there's a way to bind/bake manually keyframed shape keys to a single action so it only plays for that specifically in the NLA editor, or will I have to manually sync in the timeline. Does my question even make sense?

>>1012763
Post donut, or are you scared?

>>1012868
Then the arms are too long, he can almost scratch his knees without bending

>>1012591
i'm making the modelling,textuing and animating all by myself. it's a timesink but after you get used to the workflow it gets a lot faster

Any feedback would be appreciated, I really don't know how to improve my model right now, wasted like 4 hours trying to figure out better topology.

>>1012868
Proportions look fine to me. I don't know what the other anon is on about.

>>1012877
I do something like this. The drawing isn't 100% accurate. But a simplified concept. You might have to work out the kinks.
Also, I spread the legs apart for the resting pose. Which is unorthodox. But I find it easier see the topology that way. And with the way the edges flow, spreading the legs is more relaxed. Plus, I think I get better deformation in the butt and hip area with a resting spread.

>>1012906
> Also, I spread the legs apart for the resting pose. Which is unorthodox.
Chris?

>>1012907
lol, no. Does Cris do that? Well I guess a broken clock is right twice a day. It just made sense to me after a bunch of testing.

>>1012906
Topology retard here, but aren't you supposed to avoid having a lot of loops meet at one point like those 5-way intersections? Or does that only really matter on hard surfaces?

>>1012911
5s and even 6s are necessary to make just about anything. Because they're how you "turn" loops. I need the loops to turn, so I create a 5 that allows turning. Simple necessity dictates its existence.
5 way poles are not going to cause harm to shading. The disadvantage with 5s, is that 1 vert dictates the volume of that area. So a 5 area might appear flatter in locations that required fullness.The trick is to place them into relatively flat areas, where the edges will be mostly spread out evenly. Or inside of creases where the edges will get tucked.
So in the image, the 5s are placed along the crease lines. Where I expect the groin area to crease in the front, and where the butt is expected to crease in the back. Perhaps I could have made the back crease lower. But, eh.

From my testing, 3s are the ones that fuck up shading. You don't want 3s on convex surfaces at all. That's mainly what my topology avoids. A lot of tutorials place 3s on the butt area, because they want that fullness when the leg bends. But if you set up the butt just so, then you can get the fullness without 3s fucking up the shading. Here, the 3s are on the "inside", the concave part of the ass, along the inner crack. They will be hidden away there.

You can't get around using 3s or 5s. It's actually impossible. You have to turn loops to make smooth topology. When you turn a loop, you create a 5. And when you create a 5, it has to be resolved by a 3 somewhere. Don't ask me how it works, it's just the laws of math and topology. Best not to worry about them too much. Except try to keep 3s away from convex surfaces if you can.(Unless you just need a corner.)

Hard surfaces can handle 5s too. But I'm less experienced with hard surface modelling, so I don't feel like I can speak with authority there. Only that if you look around, you will see 5s in the work of professionals. Because it's unavoidable.

>>1012906
>>1012912
Bless you anon, thank you this is a big help. I was wracking my brain trying to figure out how to achieve my desired flow , I want in the end to be able to select the major body-parts/areas of my model as though its an action figure if that makes sense? I'd like to use this base for animations in the near future and from my understanding having easily defined loops for areas that will contort or deform like the hips,elbows, knees etc. is the best approach for animating

Getting off my ass to work on this. Got a lot of things right that I like, but a lot of things wrong. Like the transition from her top to her skirt. Texture on her hands are sloppy, and other details. Other than that, its coming along.

>>1012999
GODDAMN

>>1012999
Good shit. Accidentally hit /3/ when I meant to reply to another thread, clearly God intended me to. Keep it up.

I think Im beginning to understand...

>>1012999
Checked. Nice camel toe. Arms too short

>>1013100
Looking promising, you walk a noble path

>>1013100
yes the grid pattern is the best topo for big round ass cheeks.

>>1013123
Are you making fun of me or are you deadass

>>1013147
no really it is and ive fought with anons on /3/ over it. the horizobtal edge loops can open like an accordion to keep the shape during deformation.

could any anon help me with the body im working on? If you could point out how to improve the proportions or what kind of features to add to it, i would be so happy <3

im planning on adding more defined collarbones

>>1013184
looks great.
except the widest part of the leg is 2 low.
and shoulders and neck is pushed 2 far back compared to the breast

>>1013184
damn that looks solid. the image says lowpoly, did you sculpt and retopo or poly model this?

>>1013184
clavicle, pubus.
but your damn near ready for the next steps.

>>1013184
Show me a 45 degree?

>>1013100
>>1013163
I'm not a fan of that grid shape. The way it transitions from the lower back to the ass crack is really bothering me. And I doubt it scales well for larger asses. Probably not a big deal for small mounds.

Pic related is my actual ass topology. The one I sketch earlier. Not sure if it's the best. But I think it's pretty good.

>>1013214
I asked in the thread the other day about good topo and honestly yours is great anon its a good balance i think this'd deform well when moving.

>>1013215
Thanks. I was hesitant to post it today, because I like the way the other anon's ass uses edge sharping for the crease between cheek and thigh. It's so clean.
Though I'm not sure how good that edge sharp will look come time to move the thigh. Since you would want it to smooth out as the thigh and glute stretch out. But maybe there is a way to control the crease through nodes. Perhaps capture the normals along the sharpened edge on the face corner domain. Since the face corners will be on opposite sides of the edge. It will be as if you captured split normals. And then capture an interpolated version of the normals. And then plug the split normals and interpolated normals into a mix vector node. The mix vector node is a lerping function. You can lerp between the two vectors by plugging percentage values into the factor. So you would want to get the difference between the resting area of the faces, and the posed area of the faces. Scale that difference, and plug it into the factor of the mix value node.
If everything works how I'm imagining it, then the normal will transition from creased to smooth as the mesh stretches out.

>>1013214
all that matters is how the surface looks when it deforms. you wont see wire frames in the final render, and even if you could youd be better off using fake wires that look good and real wires that deform good.

>>1013217
All true. And still, I feel like my my topology will get better deformations than the straight grid. It also has better overall coverage for the shape of the surface. Not sparse in the cheek zone.

>>1013218
watchyu mean? >>1013214 is a grid
the interface between the thigh and the posterial cleft are not what im talking about.

>>1013219
It's not quite a grid. Pic related is the simplified concept.

Never make me look up posterial cleft again. I'm not sure if you're using the right terms. Since google doesn't know what you're referring to. It pulled up a bunch of cleft palate pictures. Which is just disturbing to look at. Even after putting your terms into quotes, and going around the "did you mean?" option, google still can't identify what those terms mean. So uhh, I'm not entirely sure what you mean either.

>>1013220
this is the part that matters. this is the part that's a grid. It;'s the meat of the ass cheek that takes the brunt of the deformation from the femur. That's where you risk losing volume. The ass crack skin is attached to bone. It doesnt, or normally doesn't, move around. Your transition from the cheek to the crack is fine the way you made it. That's just not my intended focus.

>>1013222
That grid *section* is a large chunk of the ass. But it is just a section. If you only make the grid without the other sections, then you find yourself in a situation where you lose vertex density. You can see it here >>1013100 Where the butt bulges, the faces are quite larger than the rest. Density becomes low. By forming the crack section in my topology, it's actually adding to the vertical edge density. Without creating loops that flow all the way up the back. They just terminate inside of the ass crack. The crack section also doubles as a method of creating depth for the cheeks. Since it's sort of turning "in" so to speak. Or "out" depending on how to see it. It's all tied together. Each section is needed The grid portion you highlighted is just the "top" of the cheek. The crack is one side of the cheek. And the thigh crease zone is the underside of the cheek.

>>1013214
this is almost the exact topology i ended up with after years of noodling around. poles in those same spots. must be the logical conclusion of subdivided ass.
still deforms pretty poorly for anything more than basic standing/sitting/walking

>>1013226
Really? So you're saying that more extreme poses alike splits, and fully extended running and all that won't look good? And you believe there's topology out there that can do those poses better? I would like to know what you know.

>>1013227
the crotch is just an exceedingly difficult area to skin. can't really do more topology-wise, i don't think, because unlike the shoulder the legs are already in an "extreme" pose when at rest. try to add more edges for the inguinal crease and you have to deal with more poles.
any better examples crotch deformation you come across are getting the rest of the way there by using shape keys or delta mush or helper bones or something.

>>1013228
In order to make the deformation supremely good, I have no choice but to learn how to create implicit skinning in geometry nodes. :^)

>>1013227
I said it was fine. I have done nothing but say so

>>1013214
I consider myself kind of an expert on topology of that region, in my days I model a lot of females on Korean mmo, and that topology is good, but to doesn't have the amount of density needed to make round cheeks. It's good for that kind of ass but the density of the ass compared to the back is not good.

>>1013289
Would you go for more of a circular arrangement then or just add more loops to the region?
>CAPTCHA : 4KT0Y

>>1013184
Upper torso looks a bit too far back, and legs may be a bit too long even when accounting for the missing head, but shit son other than that it's looking pretty damn good, best of luck seeing the model through

Oh one more thing I'm noticing, there's a chance the arms may be attached slightly too low, you may want to push them a little further up which should result in the shoulders also getting a bit shorter on the Z axis, maybe yes maybe no, I'm not sure about this one, I suppose if you tested it with some armature deform in that area it's going to clear this aspect up

Rabbit retopology

working on topology for a basemesh, the pose is done with metarig and automatic weights.

>>1013358
where hole

>>1013364
Cool, now disable/delete the metarig, and make your own. For practice. You want to know how it all work, without the fancy schmancy automated stuff. Start by adding a single bone, and then building a basic skeleton, and then parent the skeleton with empty vertex groups. And manually weight it.
Trust me, you want to do that now in order to encounter all the trials that arise from rigging. That way you know what the metarig is actually helping you with.

>>1013364
Have you ever considered doing something else? You kinda suck at 3d, also we already have a Cris, we don't need another

>>1013291
Not necessarily, we used a similar loop arrangement to yours, but less dense to begin with it (it was more than a decade ago), then we added a few loops for density in the ass. I use the back dimples to hide the poles

A collaborative effort between friends worldwide.

>>1013385
Whats a Cris?

>>1013400
The guy you just replied to.

>>1013401
I wish I was Cris

>>1013400
It sure is summer in here.
>>1013406
>>1013401
>>1013400
Chris just stop fucking around.

>>1013423
Maybe you should change your name to "The Cris Nerd", since you know so much about him.

>>1013423
Cris died like half a year ago. And you still can't make a donut

>>1013385
It doesn't look bad, you sound bitter

>>1013385
>>1013443
I also don't think it looks bad for a lowpoly model. The raised hand deformed weirdly after the shoulder though.

>>1013443
>>1013444
You have a really low standard on what it looks decent. If he wants to get better at 3d we shouldn't lie to him like a kid

>>1013444
>>1013443
Cheers! :^)

>>1013385
>>1013471
>Have you ever considered doing something else?
>If he wants to get better at 3d
You obviously don't want him to get better. Is the irony of what you're saying to him lost on you, or do you just have the memory of a goldfish and can't remember what you typed a day ago?

>>1013474
>That's so good Timmy I'm proud of you I'm going to hang your masterpiece in the fridge.
Learn to read, I. told the truth, I think what he made is shit, but if you want him to get good don't lie to him like a child, I never said that he could get better.

>>1013480
No, you essentially told him to quit. Quitting would make it impossible to get better; that's obvious if you're not retarded. I already know how to read, but you need to learn how to write.
>I never said that he could get better.
I'll assume you meant to say "couldn't", otherwise you'd be contradicting yourself. Typos happen, but that "n't" makes a huge semantic difference, unlike misspelling a word. My point is that I suspect you don't read what you type before you post, and that's why your posts are so idiotic.

>>1013501
Do you have some kind of retardation?
I said >I never said that he could get better
That means I don't think he can make it, but if you think he can and I mean (you) not (me) lying to him like a little kid wouldn't be a good idea.

It's not that difficult.

>>1013503
You know what? You’re right, I’m a fool. I guess I misinterpreted what you said because I assumed that there was something wrong with you, but I guess you are actually just a cunt and you meant what you said. His model may not look great, but everyone starts out like that. You would know that if you actually started instead of spending your days being a permabeg nomodeler that likes to drag people down.

>>1013506
>ur a meanie!!!
Grow up or fuck off to your discord hugbox maybe

>>1013472
based guy working on his art while others waste time fighting

>>1013513
He might be based, but he sucks at 3d

Just did this one tonight for funsies.

God I'm terrible at the game too

>>1013521
You wouldn't suck either if you obsessed over 3d as much as you obsess over that guy.

>>1013385
You are a faggot, no two ways about it

>>1013546
Don't get mad at me for telling you the truth, even the anon defending you said that the model doesn't look great.
Ok sorry, your model is fantastic, I'm proud of your masterpiece.

>>1013549
I don't mind reading your opinion, btw you were responding to someone else while assuming it was me.

yeah im sorry guy but that model is absolute garbage ! why is he squatting ? thats gay as hell !! learn to model wat tf is this shock content !!

>>1013561
Why are you french?

>>1013560
I will play your Itch io meme game

should be finishing these today

pic related is my first attempt to model a person fully from scratch, would love some feedback on my topology, I'm aware my proportions may not be the greatest.

>>1013607
>blender
quiet impressive 2bh

Improving my procedural eyebrows. I've made something similar to this before. But it was less precise. This version more accurately keeps the roots of the hairs flush to the surface. And keeps the hair strips oriented to the face the root is connected to. So that when I rotate the individual strips, they remain shapely and pointing outward for maximum visibility.

I was struggling for a couple days, because I didn't realize that when I moved the curves, their normals weren't updating to suit the new orientation of the curves. So I had to figure out how to make the normals of the curve move along with the curve in a consisted way.

The node group isn't perfect yet, but it's getting there. I still have to work a couple more things out, in order to get that perfect eyebrow swoosh shape.

>>1013613
it looks very nice for a first attempt. We do know that the hands feet and head are the hard part right?

>>1009997
Good stuff anon. My next project i was thinking of some Dark Souls type Knight and this has inspired me.
>>1010713.
Nice.
Once i gave finished summoning my inner basedboy i will be trying some kind of human armoured warrior thing.

>>1010795
Nice texture on that

>>1013646
we do, I've done some hands prior for a more toon model and i've sculpted a mannequin style head before. pic related, i believe my first attempt i complicated my topology too much?

did this stupid animation while bored of coding. Should work on modelling some 3D items like potions and such soon, since I plan to start working on menus and inventory later this week.
>>1013607
damn that effect looks super cool. How do you do that?
>>1013613
they look amazing for a first attempt imo. The purple lady tits are defying gravity quite a bit tho. I'd lower them a bit and maybe make the nipple point a little bit upwards.

>>1013624
ok, I figured out how to bend the curves in a precise way. Now the brows have that little swoop to them.
There are controls to tilt the hairs left and right, and also curve left and right.
The controls are for the inner and outer ends of the eyebrows, with everything in between interpolating.

The node isn't fully featured. But for what it can do, it does it precisely and cleanly. So I'm happy with it.

>>1013685
https://youtu.be/kB1KA9AxF9Q

>>1013682
Yup. Too complicated.

>>1013700
many thanks for this

>>1013703
keep it all quads or is there a specific kind of pattern i should be aiming for in the joints of the fingers for instance?

>>1013685
its done trough camera projection
>>1013682
not an expert but this should more or less explain the "rules" when it comes to geometry transition
files.catbox.moe/pfr110.fbx

walk animation looking stiff as fuck and I can't figure out why

>>1013783
I'll take a look, thanks anon hope it doesnt explode my computer!

>>1013784
you should try looking into layered animation, it is slower than pose to pose when it comes to getting the first draft out but it gets rid of a lot of issues early by animating the "big" bones first then moving to smaller details. The biggest problem in the animation you posted so far is the lack of enough vertical movement in the torso and the fact that its just too low to the ground.
https://youtube.com/shorts/YIXdlLwVh_4?si=9PE8cR8NOyx37kvq
Also just add more bones in general, a lot of more subtle issues will just be unfixable otherwise, even old ass games knew to not cheap out and at least split the torso in 2 and not have the head fused with the neck

model for a boxing game i'm making.

>>1013786
thanks for the feedback friend, I'll try adding more movement to the body (including a slight rotation on the Z axis) and look into layered animation as well.

>>1013805
looks good, do you have more?

>>1013784
You need to raise the heel of the foot and turn the hips to stop the knee popping, not just bring the hips down constantly, like you have.
Counter-rotate the shoulders from the hips too.
Raise the body up while it's over the foot touching the ground, then down again at the widest part of the step.

Where did I fuck up

>>1013786
>>1013816
many thanks anons. I put some more work into it and made those changes:
>shoulder movement, the body rotates so the arm that's pointing forward has it's shoulder taken forward as well
>more arm movement, including lower arm which was static
>brought the body up, and adjusted it's position in several frames so the knees don't "pop" so much
>made sure that the body moves up and down when it should
>heels move
I think it looks way better than before! many thanks again, you anons are very helpful here

>>1013819
Probably scaled the teeth in object mode, and then forgot to apply scaling.

trying out squiggly outlines. I wonder how ill be able to implement this into a game engine

reacting to deformation

Learning how to sculpt heads. I suck at this

>>1013835
Previous attempt

>>1013823
It's a bit off. Looks like it's walking on snow. Some strikes are too long, some too short.
Do the walk in place and then move it. A single loop and repeat it.

>>1013685
Stiff but charming

>>1013653
Thanks

Female werebeast version

New Thread:
>>1014090
>>1014090
>>1014090

>>1010216
best thing in the thread