A thread for work in progress of your projects for (horrible) feedback by the community.
Previous Thread: >>1009306

needed to create a UI for the bones.

Made an RV with interiors 'n deployable awning

>>1014116

>>1014121
In the desert
You can't remember your name

>>1014116
load the rv in vr

do these hands look ok?

Working on 3D OC art, I am a beginner

>>1014140
They look fine to me.

Second model I've done, following guides obviously. Suggestions for it to look better or critiques?

>>1014295
1. perfectly sharp edges are basically impossible irl. there's always a little bit of bevel/rounding. the hardsurface parts of your watch definitely need it.
either bevel on the geometry itself, or look into bevel/round corner shaders in your renderer
2. look up imperfection maps and how to use them. you can use them to add a tonne of detail very cheaply.

>>1014140
please look at the side view and then look at your own hand. you have reference irl.

>>1014295
quality as fuck

>>1014295
>critique
I'd say the background could use some work
>but how
I'm not an art director, but the quality of the render (the watch) is so good that the background kinda brings the quality down
>inspiration
look for luxury tier watch brands advertisement and product placement, get inspired

>>1014092
bone UI made
eye textured
now to the main textures.

>>1014340

>>1014140
They are not quite right
>>1014295
Bevel the edges, and get a good studio hdri, lighting is not the best on your render.
>>1014340
Fur direction can help to sell it more as fur

>>1014343
sadly, I can't afford substance painter nor the addon.

>>1014277
Tried to fix it up a bit.

>>1014298
Alright, I beveled some of the case edges by 3 segments, which I'm guessing should be more noticeable on macro shots. I tried looking up beveling by shader render options but it wasn't working really, it just added some weird shadow. As for imperfection maps, I downloaded some but non were looking right, which is something related with uv maps if I had to guess. I'll keep looking for the right one, because most of the maps I found were for very rough metal surfaces, and not quite fitting for a polished gold surface.

>>1014337
I put the model on a very simple background cause I'm more interested in learning how to model watches than general scenes, but I did tried messing up with the background and add a few things as you suggested, a vintage watch advertisement had the watch sitting on a mirror with some blossoms and petals around it, so I tried to mimic that advertisement format.

>>1014343
Some edges beveled, not all of them because it just messes up with the symmetrical shape of the watch, but most of the visible ones at least. I'll look up different hdri to get better shots, but the main focus for me remains modeling the watch itself.

>>1014418
And here the bevels, in green what I could do with no big complication (I had to merge and re position some vertex but nothing really complicated), in red I tried but it would turn the shape into something really ugly if I tried to apply bevels. Maybe in the future when I have more experience I won't run into these problems, but for now it just turns it into a mess. Also thanks for the suggestions.

>>1014418
the bevel adds tremendous beleivability to it

>>1014342
https://sketchfab.com/3d-models/pla-buneary-retopo-and-retextured-06eeb2131f5f41de93e1e5db77fd9937

>>1014419
You can do them non destructive by using the Bevel modifier, but it looks much better now
>>1014344
You can project directional fur textures on blender mostly for the mouth part and then mix it with the original fur texture

>>1014418
>background critique
I see what the issue is now, in the other pic you included an actual photograph of the watch, in the to left. I confused that with your render, which is good, but not a photography of the actual object, sorry for the confusion pham

>>1014464
No its good, it adds to the presentation that advertisement suggestions of yours, I'll definitely be doing something similar when doing other watches. Putting simple stuff next to it that is, like petals, pencils, etc.

>>1014427
Top:Lips too flat
Bottom: Why not space it to avoid that bump?

>>1014357
The previous one looked cuter. Why are you making the head smaller?

>>1014116
Furry looking hot, RV looks really nice too
>>1014140
Finger tips too thin
Webbing seems to reach too low.
>>1014295
>>1014418
Somehow I thought the reference was the 3d model. From a layman’s POV, you should focus on the material

>>1014482
funny you mentioned it, but I'm trying to respect the doc as much as I can
as for the bump, I found a solution for the bottom lips.

>>1014489
pic related

>>1014489
>>1014490

I just think it would look better. Feel free to ignore me.

>>1014427
>no genitalia under the wool
I am disappointed.

>>1014501
I had to make it explicit If I were to do that (and possibly pull a Cumminham and charge you for it).

>>1014418
1. do learn uv mapping it's fundamental
2. you don't actually need uv's here, you could also use something called triplanar projection. this is also very useful so look it up.


as for imperfection maps there are some here that will help
https://www.artstation.com/marketplace/p/3yAy/free-surface-imperfections-pack

these do look quite extreme but you can do a lot by just manipulating them
pic-related is an example of such a setup where:
1. value node sets your base roughness value
2. the float curve allows you to manipulate the values in the roughness
3. the roughness map's values are then added on top of the base roughness (whiter = rougher)

the clamp is there for housekeeping

adding isn't the only operation. subtract, multiply, overlay, screen are all just as useful.

adding imperfections to roughness is just the start
you can plug scratchy looking ones into bump to add little scratches
you can use them to mix colour corrected version of diffuse maps together (e.g. the same scratch map in bump can be used to blend in a darker version of the diffuse colour in the scratched area)

have a play, i think you'll enjoy it. look up some tutorials on youtube to see how people use them.

>>1014507
>manipulate the values in the roughness map*

Putting the finishing touches on this glock. Still have a couple more things to model like the takedown lever but here's some test renders.

>>1014534

>>1014090 (OP)
I have been trying to create a model from niche anime.
While collecting references, I'm realizing how much anime cheats with the size differences.
Character can literally appear twice as large in the same shots. And there is no explaining it with perspective distortion.

>>1014567
that's the world of 2D for you. Like all those 2 head tall characters with stumpy arms having no trouble reaching around things and grabbing them.

>>1014572
That's why I prefer 3D, character sizes stay consistent and on-model.

>>1014567
Woah... that dude's penis and butt is like totally out... in front of a girl too. Like imagine if he got a boner in front of a girl while naked and she saw it that would be so embarrassing

>>1014575
Fun thing is the his penis actually very inconstaint as well.
In one shot is it is literally the size of a woman's head, in another it is girth the same as a sword
I have never seen something so inconsistent.

>>1014573
not anymore. big production 3d is all about deforming meshes and forced perspective to mimic 2d animation. Side mouths and stretchy arms galore.

>>1014579
I hate it, why would they take one good aspect of 3D and ruin it?

>>1014580
you know why

>>1014580
>>1014581
the people want 2D, not 3D

I haven't gotten to sculpturing the ribcage, it's just a placeholder at this stage. Probably will start modeling his staff before returning to do more detailing

>>1014598
that looks amazing, anon

>>1014600
Thank you, anon

>>1014598
the horns are awful, and I say this because the rest of the character design kinda rocks, so take the cape, the skull, the neck and shoulder pieces, those are cool af, but the horns are ... idk pham I don't like them one bit, about the red spheres, maybe an eye of sauron treatment to them?

>>1014634
Not my design(It's from Overlord anime), so I can't really take out anything from it. But the orb inside the abdomen I intend to try and make it something like this(pic rel)

>>1014090 (OP)
scrangly fella

>>1014598
Mmm sexy Ainz
>>1014640
Looking good

working on potions. Might be me but I think this one looks kinda tasty.
>>1014640
based kobold

>>1014507
This is pretty cool, I think I finally got the hang of it and why it wasn't working back then, I had to adjust the uv of the faces by unwrapping them one by one so they look good and not stretched/blurry. This is great as it'll come just in handy for my future vintage watches models, thanks man.

>>1014640
His tail looks like a penis.
Tell me that's intentional

>>1014635
badass, nice nice

>>1014295 here, now I've done a third one. Not particularly the kind of watch I want to focus on, but fun regardless.

Had an incredible breakthrough tonight. I finally figured out a smart way to clump curves together. I was DYING trying to figure this out for WEEKS. All of my ideas involved rotations. First I would align the points, and then rotate them to fit the direction of the curve. I tried a dozen different variations, and they all failed due to one thing: gimbal lock. No matter how sophisticated the rotation method was, I just couldn't beat gimbal lock. Eventually, I had to face the harsh truth: That no matter what quaternion support blender nodes *claims* to have, the output will always end up in standard euler, not quaternion. Therefore, I will always encounter lock. I had to find another way.

So I finally put my mind to it, and realized there was a simple solution all along: Lerping. All I really had to do, was get the starting tangent, normal and cross product for each guide curve. Project the root points to each, dot product, and that gives me the relative distances of the root points to the guide curves. And then for each point of the curve, offset by those distances. And voila! Perfectly clumped curves. They don't gimbal lock, because they're not rotation based.

There is one problem however. I made it so that the normals of the clustered curves, rotate in such a way, that the mesh faces perpendicular to the guide curves. I did this because the mesh is flat, and so if they're all facing one way, then when you look side-on, the clump seems to disappear. So by making them perpendicular to the guide, it mitigates how much you can see though. That much works perfectly, the problem however, is if I try to mix between the regular normals and the perpendicular normals. It creates twists in a few of the strands. Not 100% sure why, but I think it's because parts of the strands are mixing "over", and other parts are mixing "under", so to speak. And the twist is point where it changes from over to under, or under to over.

Anyway, I'm happy.

i think it´s supposed to be a flowerpot dude?

>>1014598
Some updates on this fellow. Almost done with the zbrush part of detailing. Soon I'll have to take it out and retopo/bake it to be able to texture it properly. But maybe I'll do his mask first

The begining of this video features a gimp being deployed into the forrest and made me want to reacreate it. What a beautiful creature....
https://www.youtube.com/watch?v=EKtr96YdDEc

lowpoly bake done, will texture it prolly next month

anyone here working on a game or want to collab on a 3D game?

>>1015040
What kind of answer are you expecting when you're being so vague? Try again.

>>1015013
lmao

>>1015042
>What kind of answer are you expecting when you're being so vague?
I'm just asking a general question if anyone here is working towards a game or has an interest in collaborating on a game

Just reworked the eyes

>>1015046
Someone started a /3/ collab not too long ago. I forget who it was, or what thread they were in. So I can't really help you find them. But it's was maybe only a couple months ago? I think it was being organized by that dev who made the sexy pacman game???

>>1015040
Yes, lets do a MMORPG togheter! XD

>>1015040
>>1015046
>>1015053
Actually, I would be interested, if it were a serious offer, but given the circumstances,.it won't happen anyways ...

>>1015040
>>1015054
https://itch.io/jam/nsfw-minijam-19
This is starting just now, just to let you know

>>1015047
good ratios on the face.

>>1015040
Yes and yes.
I've tried to network with anons on /3/ a couple times and it fell apart each time, but I'm still open to the idea.

>>1015055
>theme is stealth
So like a game where you have to sneak around a city and find secluded places to beat off, and there's actual jerk off cops patrolling you have to avoid.

>>1015059
>>1015055
>>1015054
discords? I'm not an artist I'm a programmer

>>1015064
aight
https://discord.gg/gtkHDHJB
lets try again

Put one WIP up on characters on a game I'm making in the last thread. Why not post another? Not sure if I'll keep them as whole models or chop them up into segmented characters like ps1 and n64 games yet.

>>1015055
>>1015074
>>1015064
>>1015060
This is the jam's discord, find a team there if /3 disappoints
https://discord.com/invite/98m5PUgCU6

>>1014090 (OP)
Is this too retarded for topology ?

>>1015089
What are you trying to do?

>>1014483
how about now

>>1014995
Update with the masked version of Ainz

>>1014982
Small update to my hair thing. I switched from strips to tubes temporarily. Since I don't have to account for the direction of tubes, it avoids the twisting issue I was facing.

Using 17 guide curves, I generated this hair. The hair stems from points on a scalp, and clumps to the guide curves. Or another way to think of it: the hair extrudes from the guide curve.
Each strand's orientation is based on the randomness of the root points, thus they carry that randomness down the length of the strand. The randomness appears fairly natural. Though not entirely natural, because you can tell it's clumping in an obvious way.

I'm currently working on a method of interpolating between guide curves. It's proving difficult. But I believe that when I figure that out, then the hair will truly look natural.

>>1015130
Does it have a performance improvement compared to other hair workflows?

>>1015131
No. It's geometry node spaghetti. It's also barely developed at all. It's as performant as Blender typically is, when you generate tens of thousands of curves, and give them a 4 sided profile.

I believe my actual node groups to be fairly efficient. Taking only about 106ms total, in order to generate 17,298 splines in this manner. Though I am only generating the hair within the radius of the guide curves, I'm not generating a full head of hair. Which google tells me that 90,000 hairs is on the sparse side for a human head. I do plan on hitting that target. When the guides and interpolation are done, I'm going to create a "Fill" group, to fill in all the hair that the guides don't directly touch, and have them interpolate automatically to whatever they're nearest. So yeah, with a 90,000 spline target, it's probably going to become slow as fuck.

So far, I find it more intuitive to use. Which is why I'm endeavoring this at all. The guide curves define the shape of the hair, and they don't move when you tweak the mesh. Because it's not UV dependent.
It's probably not going to be a good method for creating real performance hair. Since it's reliant on the randomness of point distribution to establish where the roots of the hair are. And that *does* change when you tweak the mesh. Which isn't an issue when you have tens of thousands of strands, it doesn't matter if the stands move 0.001cm. But if you were to try and work this at a larger scale for say hair cards or something, then it would be really annoying to have the placement of your cards change, every time you tweaked the mesh.

>>1015092
np already solved shade smooth was fucked up but now it works as it should.

Defias cutie for a WoW themed project

>>1015132
Still haven't figured out how to interpolate between curves. But now that I know how to do this basic "extrusion". I figured I'd give a shot at creating braid. I don't know how to do that 3 way overlapping braid. So I settled on a simple twist.

This is fun. This is the kind of stuff that amazed me when I first started learning. Now I'm starting to actually understand it. Not merely copying tutorials.

>>1015099
First one was cuter, I prefer the first version, you can take specular roughness so the render looks better

>>1015152
Fixed a minor issue that prevented the twisting curves from rendering.(The radius of the curves was nearly 0)

>>1015199
Go to bed Cinder. You're drunk

>>1014116
yiff in hell furfag

>>1014090 (OP)
Is this ok for first subd modeling ?

Legless alien dude and his handicap hoverchair.
Alien is sculpted and has vertex colors. Chair is just metaballs at the moment.

>>1015220
Looks pretty good. Might want to tighten some of those edges with extra loops depending on the look you're going for.

>>1015074
invite expired, brochacho

>>1015248
Sure ok.
I wanna make a game. I'm open to participating in someone else's idea. I have my own ideas. Theyre quite degenerate. If that sounds rad then great. If not I can compromise.
https://discord.gg/Gf5kcgAN

>>1015250
invalid invite

>>1015250
>>1015253
Both links work you POS

>>1015254
???
why so angry

THEN WHO AM I TALKING TO?

>>1015257
You seem really dedicated. Clicking 2 links and posting a screenshot is 1,000x the effort most people are willing to make. Post a link and I'll click it to come to you.

>>1015259
i think im already talking to you bro

>>1015167
Made some headway on the curve interpolator. In the video, the hair looks extra clumped at the start. That's normally what it looks like without adjustments to the clump dimensions. Just straight cylindrical clumps.
Then I enable the interpolator, and it *mostly* disguises the clumping. It's still a work in progress. But I got the general sketch of the thing down. Working out the issues from here

I broke out my ever handy voronoi triangulation node. I made that thing like 2-3 years ago, and still find it's one of my most useful nodes. It's gotten even better, since I was able to replace a slower component with a simple math node. That optimization made it able to handle a lot more points a lot more quickly. It's still not a lighting fast node though, because it has to do the dirty job of creating hundreds of thousand of new points. Potentially millions. And when you do that, Blender is going to slow down regardless. Still, I'm super happy with the speed increase.

The voronoi node makes it so that points can copy data from other points within its "field of vision" so to speak. Like a convex proximity checker. It looks around for neighboring points, and creates copies with shared indices. I call it a "cloud"

I take the cloud of data, and capture a bunch of attributes, then collapse it back down by deleting the copies. The whole process takes some getting used to. It's fairly convoluted. But with this method, I can mix a bunch of attributes together, and normalize, and have them come out the other end a perfect blend. In theory. I still have to do the math in order to make it perfect. But functionally, it's doable.

Also, I figured out how to prevent curves from interpolating across the parting of hair. A simple trick of getting the dot of the curves normals and the copy curve normals. And any copy with a dot less than 0 gets deleted.

Anyway, I think I'm going to take a break from this for a few days. It's taxing.

>>1015244
did you paint that manually or baked AO ?

>>1015316
Manual then applied 'dirty vertex colors', which I guess is basically AO. Still need to do another pass on it.

Will make a video out of this, but i'm still deciding on the lighting

>>1015244
Found an addon called 'Blob Fusion' that is way better than meta balls. Remade everything on the chair except the main body with it. Probably going to end up replacing that too.

>>1015102
An update on the real-time version of this fellow. Still have to retopo and texture the mask and the staff but it's getting there

>>1015455
Okay, I think I'm done with this until I can muster up the autism to retopologize.

>>1015245
Roger that.
Will play with it more.

>>1015514
good shit pham

>>1015538
thanks anon, it was a fun mini project

>>1015539
some details are sublime like the gold embroidery on the cape

>>1014090 (OP)
I had this brilliant idea of trying to recreate a manga panel by posing.
Fuck posing is really hard...

>>1015569
Now try making the models from scratch, drawing the hierarchy, binding the mesh, rigging it, and then posing.

>>1015569
Ignore the other guy, do a pose a day and you'll get better

>>1014428
link kinda related, but found a good tutorial on youtube about blender and fur painting: https://sketchfab.com/3d-models/pla-lopunny-retopo-and-retextured-8d8ccf24276b449c80ccfc75de78480b

Working on geo nodes eye.
I now comprehend nodes on a foundational level. I'm no longer held back by ignorance of the system's workings. However, it still takes some problem solving to do what I want. It's still not easy. But it's understood now.

>>1015577
whats the benefit of doing this with geo nodes instead of something like bones/shape keys?

9000 hours

I don't know what to do with the shape of the torso of hips, aiming for mechanically plausible industrial machinery.

Where do you guys find inspiration/reference from?

>>1015638
I used to think that mechs would be easy to make because of the relatively simple shapes until I realized that you actually have to handmodel plausible joints and such instead of just having them hidden under the skin of an organic creature

>>1015638
this looks good, very efficient vertex use

>>1015638
In real life industrial machinery the thing gets thicker the lower you go. Especially in an "unflattering" way. Like this mech is human shaped. If it were a human male, broad shoulders and narrow hips would be ideal but a machine built like that would break. Every component would need to be built to lift every component on top of it.
Pic related is a real industrial robot. It tapers. It doesn't look like a sexy human arm. It looks dumpy because it get wider as you get near the base.

>>1015644
>until I realized that you actually have to handmodel plausible joints
aint that the truth, I'm trying to get away with using the default human skeleton too.
It's for VR so I have to account for all the ways a human can pivot and roll. basically have to make everything ball joints.

>>1015646
thanks anon

>>1015647
very good point. I'm willing to suspend enough belief to keep the humanoid proportions.

>>1015649
>basically have to make everything ball joints.
Welcome to the universal joint

quack

>>1015577
Remade the eye entirely with a different approach. It's more flexible now. The iris size can be adjusted. The depth of the pupil can be adjusted. The mesh is air tight and UV mapped. Still a work in progress.

>>1015614
Precision.
I'm not certain it will be useful for animation. I haven't delved deep into animation yet. So I don't know where that rabbit hole leads. However, if you just want to create an eyeball. This node group makes the initial creation process easy. And because they're generated onto empties, you can pose them with ease. Anything an empty can do, the eyeball can do.

>>1015704
>Anything an empty can do, the eyeball can do
doesnt this go for any object? i guess this is useful if youre making 100s of eyeballs but how often are eyeballs THAT unique between characters?

>>1015708
>>Anything an empty can do, the eyeball can do
>doesnt this go for any object?
Yeah, sure. I could make a new object, like a random cube or something, and then replace the cube mesh with the eyeball. Manipulating an object works just as well as manipulating an empty.
I guess I just like how empties look. Especially the sphere ones with their axis enabled. They make it very clear, the scale and rotation of the object. And they don't create any extra geometry that gets in the way when you go into edit mode. I guess it comes down to preference.

Creating this eyeball is the smallest of my ambitions. I have some much more complex stuff I want to use nodes for. But I've been ignorant for so long, I haven't been able to accomplish any of the higher goals. I'm happy that the fundamentals of nodes has finally clicked for me. It feels like a fog has cleared from my mind.

>>1015704
How can you add geometry nodes modifiers to empties? For me it doesn't work and empties cant have any modifiers stacked on it. :0

>>1015738
I don't. I add the modifier to a mesh object, and then use the Collection Info node to read a collection of empties.
In this case, the eyeball modifier is on the body. And it's reading a collection with a couple of empties in it. I'm not sure if this is the best or most efficient way to do it. It's just the way that seemed best to me.