[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/3/ - 3DCG

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: 1724658894394525.jpg (1.17 MB, 1488x863)
1.17 MB
1.17 MB JPG
/wip/ - Works In Progress - "graphic design is my passion" edition

Post your work-in-progress projects, recently finished projects, or things you'd like critiqued here.

Previous thread: >>993026
>>
>>998180
you ruined it.
>>
>old thread at page 4 and not at image limit
>whole board only has 9 pages
yeah you're special
>>
File: shoulder mechanism.webm (1016 KB, 1280x720)
1016 KB
1016 KB WEBM
shoulder mechanism. flawed but it doesn't seem to break easily at least. also why can't IK just fucking work? why do I need to add so many extra bones and constraints just to make it so it doesn't shit the bed?
>>
>>998189
>anon doesn't know what a bump limit is
>anon is obviously a newfag because this has never been a problem before
>>
>>998195
You are so retarded, holy shit.
/3/ needs twenty more threads before any thread can fall off the catalog. It doesn't matter what page the old thread is on if you can still post in it.
Bumping has also been broken for eight years now. Sort by last reply and shoot yourself.
>>
>>998198
Again, it's never been a problem before, and it's not a problem on any other board. Creating a new thread after the old one has reached its bump limit is how it's done on literally every other board that has generals and no one shits their pants over it except you.
People are less likely to post in threads that have reached their bump limit, and I just like to see the board be active.
>>
>>998199
>Creating a new thread after the old one has reached its bump limit is how it's done on literally every other board that has generals
so why do you have to be such a little smug faggot with that op pic
>>
>>998200
I don’t recall feeling smug when I edited the OP pic, maybe just a little silly. I could have probably written “You don’t need to be someone special to do it.” Instead of “You don’t have to wait for someone special to do it.” because that would have gotten the point across better. It sounds stupid, but I have encountered people on this website before that think the jannies are the ones who make the general threads. I've seen the activity in the wip threads grind to a halt for weeks at a time because no one wants to make a new thread.
Maybe it’s my fault for wording things wrong, but your need to tell me I’m not special and that I’m being smug makes me think you’re projecting your own ego disorders onto me, because that's just not the way my brain operates. I’m just trying to have a little fun.
>>
>>998199
Here's some prior work and a reasoning for it: >>997228
>10 days ago
>it's on page 6 and still has space for images
Stop shitting up my thread watcher. You make the thread when you need it, and right now you don't.
>>
File: file.png (123 KB, 1376x989)
123 KB
123 KB PNG
Now modeling the gorget. I'll see if I can stamp some leaf-like pattern into it.
>>
>>998217
Can I cum on it?
>>
>>998220
No. Great shame. Ancestors cry.
>>
it is op's fault for making a cringe statement that he should have known would make retards have a temper tantrum

it is anon's fault for getting chipped by a throwaway cringe pic
>>
>>998217
Is it even worth the effort to use plasticity, bro? I have a real, paid, active license, but I also like modelling with polys because it just makes more sense to me, if you know what I mean.
>>
>>998180
I touch my pee pee at night
>>
File: file.png (134 KB, 879x802)
134 KB
134 KB PNG
>>998234
For surface modeling, absolutely. Polygonal modeling is quicker for most things, but I will say that I could spend an eternity trying to model that with polygons, and in the end I would have failed. I would never be able to get the perfect shading, the fillet washouts, the smooth transitions, etc. Maybe with subdiv, but even then... I wouldn't recommend it. It's not worth the effort if you're aiming the surface modeling look. The gorget, I could do with polygons, but not the rest of the helmet.

And I just think it's very fun. Time flies, which is why I only do this at weekends, or I wouldn't get any work done during the week.
>>
>>998240
are you using the blender bridge at all?
>>
>>998241
No, I'm using plasticity through vnc on a separate computer and I couldn't get the bridge to work over a different IP. I just export as OBJ when I'm finished.
>>
File: wow_rage.png (1.54 MB, 1416x1080)
1.54 MB
1.54 MB PNG
>>
File: file.png (341 KB, 1122x876)
341 KB
341 KB PNG
girls
>>
File: tbrender_Viewport.jpg (3.06 MB, 3000x3000)
3.06 MB
3.06 MB JPG
Still working on it, maybe another month of work left to finish this entirely and post it.
>>
>>998246
perfection
>>998264
What else is there to do? I assume the parts who aren't visible?
This looks pretty fucking good
>>
File: IMG_2690.png (1.02 MB, 794x594)
1.02 MB
1.02 MB PNG
>>
>>998258
super neat anon
>>
File: IMG_2693.jpg (317 KB, 1362x1584)
317 KB
317 KB JPG
>>998270
>>
>>998269
There’s still a lot to do with the wood. Some splintering here and there, some texture adjustment to add granularity. As for the metal part, it’s barely worked on, I’m still gonna work on it for at least three weeks before it’s close to what I want. I have multiple references that I’m following. But thanks for compliment my guy
>>
File: file.png (436 KB, 1919x920)
436 KB
436 KB PNG
>>998258
day 2
>>998271
thank you.
>>
>>998284
holy gemmy...
>>
>>998270
>>998284
gemstone
>>
File: skullb1.png (1.93 MB, 2140x1904)
1.93 MB
1.93 MB PNG
>>998190
more or less done. just need to mess with the rifle a little bit more I think, not very enthused by it.

what do you think, blue fresnel outline thing on the shader, or skip it?
>>
>>998338
how long have you been doing this?
>>
>>998199
>it's done on literally every other board that has generals
It's literally not. Generals stay until they reach image limit and then get refreshed. That's how it works on every board.
>and no one shits their pants over it
Everyone shits their pants over it when some retard bakes a thread early.
>>
>>998340
since january
>>
>>998388
doing good so far, but i think you only have the initial step. I think you should do rigging and animation on this character
>>
>>998284
high effort meme
>>
>>998391
You mean high effort art
>>
>>998389
ty. mostly rigged already, ik/fk limbs and parent switching for the gun. just need to add some constraints to the fingers and we're good to go. animation on the other hand... that shit is hard man. kind of have an idea of what I want do with him but it ain't gonna be anywhere near as good as it is in my head.
>>
>>998437
are you using blender for animation?
>>
>>998438
yeah, using blender for the whole thing.
>>
>>998180
i have not been to this board in over a decade and you're all still using this same image.
that makes me so happy.
>still haven't hit 1 million posts
glad to see nothing has changed.
>>
Cool thread idea
>>
>>998589
Bot, I hope whoever made you gets fucked in the ass with one of those baseball bats that have rusty nails embedded in em.
>>
>>998284
>>
File: lans.png (2.92 MB, 1920x1080)
2.92 MB
2.92 MB PNG
>>
How do I make a good low-poly anthro head?
>>
File: DesertWip9.jpg (1.08 MB, 1920x1080)
1.08 MB
1.08 MB JPG
>>996394
Been pluggin away at this in the background, nearly time to cut things and start rendering. I could probably work this to death, but I need to make sure I have the time to actually render the anim.

One question though. Anyone know of a way to make heat haze that's more or less "depth accurate"? Like yeah I can throw a refraction on a plane and call it a day (which is what I've done), but there's a bit more nuance to actual heat distortion than that. Not to mention since it's just a plane at a fixed distance from the camera, anything that passes that plane (for instance the train), will be distorted instantly like someone just turned it on, instead of a gradual thing as it gets further on. So a heat haze that was a bit more gradual over a distance would work, but I'm not really sure how to do that, and searching for it just gives the basic bitch "just throw refraction on a plane, bro".

Also, I'm pretty fed up with Blender at this point. It's a fucking dice roll now when I press render whether it will crash (about a 50% chance), get stuck on "initializing" and not render at all (30%), or actually fucking render (20%).
I've narrowed down WHAT is causing it to fuck up (the road and adaptive subdiv), but not how to fix it. All searches point to "disable GPU subdivision", which I have, but no dice. Switching to the 4.3 beta made it a bit more stable, but it's really that extra 10% to make 20% on the render dice roll.
Not to mention the random, "will this object render or not", or "will the displacement displace" roll when adaptive subdiv on.
The modifier is a fucking mess.

Thanks for reading my faggy blog post/rant.
>>
>>998947
get a more fucking powerful GPU then faggot
>>
>>998947
do it post-render and use a depth-pass to drive it
>>
>>998949
Next "more fucking powerful GPU" than I currently have is a 4090. Somehow I doubt that's going to make a difference. The GPU's strength isn't the problem. There's something fucky going on with Blender in conjunction with my project. Just weird shit all around, despite throwing much more at it in the past without issue.

>>998954
Yeah that actually sounds like a good idea.
>>
>>998958
have you tried a clean uninstall and reinstall drivers with DDU? Also, this is exactly why I dont buy the most expensive cards. Cards always break after a while. I've had so many cards malfunction that I just buy cheap ones like 60 series now. To be honest with you I think 60 series cards can do everything almost as well as a 90 series card and you can replace it easily
>>
File: 1000004467.png (707 KB, 1165x777)
707 KB
707 KB PNG
>>
File: lans2.png (2.75 MB, 1920x1080)
2.75 MB
2.75 MB PNG
>>998747
idk what to have him lean on but whatever it is im probably gonna use geonodes to scatter it everywhere
>>
File: 0001-0047.webm (207 KB, 720x1280)
207 KB
207 KB WEBM
>>998970
also:
>>
>>998960
I've updated drivers, but not reinstalled them. I'll give that a go.
Again though, I don't think it's the GPU. It works great. I've never had any problems with it, nor any of my previous ones (My 1080ti is STILL trucking in another computer as well). Like I said, the problem persists even with GPU subdivision turned off, so I don't think it's the GPU at all.
As for 60 series, I also use this for games too, and I run at 4k. 60s won't really do me much good.
I've got a 12600K (not the best CPU, but again, no issues), and 64gb of RAM. GPU, CPU and RAM all bought this year, all the parts have been working great.
For all intents and purposes, things should be well within my margins to render. I think it's just a freak thing. For some reason it's been crashing less, so who fucking knows now.
>>
File: lans3.png (3.03 MB, 1920x1080)
3.03 MB
3.03 MB PNG
>>998970
settled on pillars i think
>>
>>999012
I had meant to say cacti, but I like the chunky pillars better.
>>
>>999012
what mr bone contemplating about?
>>
>>998180
newfag here
why does this thread always have the same image or slight variations of it
the dedication is admirable
>>
>>998976
its good, but i want to see beyond a simple walk cycle into something "original"
>>
File: Wip Collage_Series 1_v6.jpg (1.55 MB, 1500x882)
1.55 MB
1.55 MB JPG
>>999063
>why does this thread always have the same image or slight variations of it
Tradition. One of the few things true /3/ fags can take pride in.
If I remember right, it was an oldfag's project that he made for WoW or something along those lines. It's been a good long while since I heard the reason, so that might be a bit off. Still, it's been the wip image even back in 2012 when I first started hanging around. /3/ was a wildly different place back then, that's for sure.
>>
File: cowboy.jpg (140 KB, 570x760)
140 KB
140 KB JPG
>>999071
ty, the plan™ needs a bunch of animations, just knocking out one I actually somewhat know how to do first

>>999060
idk. inspired by picrel. grew up in a very rural area and every other house had one of these out front.
>>
File: 19782738.5_image.png (198 KB, 783x841)
198 KB
198 KB PNG
>>998334
...
>>998940
always start by the eyes region and do the snout later
>>999073
i'm in that pic. nice of you to save that collage, good times.
>>
File: 0001-0200.webm (815 KB, 1280x720)
815 KB
815 KB WEBM
>>
>>999095
I'm going to be harsh with you but its because I see potential in you....

i think you need to challenge yourself more because right now what i see is very easy, like anyone could do it
>>
>>999099
by all means, harshness is why I post on 4chan instead if discord or reddit or something

>challenge yourself
so what do you recommend? that's quite broad but I'll give anything an honest try
>>
>>999100
you have to find something you're passionate about or else you won't do it. This can differ for everyone.

For me, I'm so into what I'm doing that I'm doing these subjects all the time, at home, at work on break, while I'm watching tv on my phone, when im driving in my car im listening to stuff thats related to what im doing. You have to find something big that you can do. You have to plan. You can create an animatic in resolve with images and videos
>>
>>999102
what a load of shit with zero feedback of any value
>>
>>999104
this
>>
>>999104
Re-read what i wrote. You are making extremely basic animations with keyframes. Me, I am planning like mad, making animatics, trying to stay at the front, coming up with new things that don't even come up with anything substantial on a google search. What this is, you have to find out because its personal for you. If you dont like what you are doing, you wont do it, period.
>>
>>999106
that was a different anon but I kinda agree with him. you come off as a bit pretentious with this, whether that is your intention or not.
>>
>>999107
I dont even know what pretense means.

You do you. I'll keep from replying to you from now on, but know this : what you posted is low effort. Most of what gets posted on this board is low effort.
>>
>>999108
post your work then, let's see the high effort masterpieces you have created
>>
File: 1709470141228874.webm (934 KB, 1440x1080)
934 KB
934 KB WEBM
>>999109
this was my old sim that i had to give up because the plugin company got sold and stopped supporting licenses
>>
File: aor02_01.jpg (300 KB, 800x600)
300 KB
300 KB JPG
An experiment with some kind of blue fog with my own raytracer.
>>
File: aor02_02.jpg (201 KB, 800x600)
201 KB
201 KB JPG
The white streaks are because of t-junctions in the model. I'll fix them at some point. The models are from Cube 2/Sauerbraten.
>>
>>999110
Kek, please tell me you're not the raw chicken schizo and that you're just using his work as bait.
>>
>>999114
It's they/them.
>>
>>999110
I'd love to see your animatic for this
>>
File: 1701904371014002.webm (1.21 MB, 1920x1080)
1.21 MB
1.21 MB WEBM
>>999114
show me some respect. I have years of simulation experience with fem solver

>>999116
my animatic is for my eyes only
>>
File: 3.jpg (210 KB, 1080x1350)
210 KB
210 KB JPG
reminded me of this, thanks for the laugh anon
>>
>>999118
lol, learning fem and then vellum to a high level is so pointless and silly am i rite guise? Time to get some kind of meme to post from mah folder?
>>
File: Ark Eye Test 02.webm (1.08 MB, 704x262)
1.08 MB
1.08 MB WEBM
I'm messing with parallax eyes again. It might not be obvious how they're different from the previous parallax eyes I made. But in the previous one, I used a very difference node set up. all the difference are under the hood. I pretty much made my own displacement node. Doing so gave me a better idea about how this stuff works. And I was able to get the parallax scrolling to be more accurate. Now, when you turn then camera, it perfectly offsets.
Though, there is still a problem with it I have to work out. The work isn't done. As the nodes grow, it gets harder to identify the problems and fix them. I don't know exactly what's causing the problems.

I didn't go crazy with the displacement in pic related. Just gave it some subtle depth.
The cool thing about this is: it doesn't require UVs. So when I finally get a handle on this, then it should be entirely independent of pesky UV unwrapping.
>>
File: file.jpg (382 KB, 1920x1920)
382 KB
382 KB JPG
>>998217
Still some stuff to do, like the interior cushions, buttons, micro fillets. Maybe a flashlight, I haven't decided yet. It's coming along nicely though.
>>
File: DogDraftsPrtSc.jpg (205 KB, 1973x1196)
205 KB
205 KB JPG
D O G D O G
>>
File: UpPrtSc.jpg (299 KB, 720x2458)
299 KB
299 KB JPG
RANDOM-3-D MONKEY
>>
File: blue Water acid.jpg (472 KB, 700x2980)
472 KB
472 KB JPG
>>
>>999130
dont you think there should be more "thickness" to it? Like, everything looks like it has only one side now
>>
>>999138
Yeah I'll make a cloth sim to add cushion on the inside. Hopefully it turns out nice.
>>
>>999143
why dont you just model it with thickness in plasticity, like you do in real cad when you have specs
>>
File: file.png (294 KB, 564x564)
294 KB
294 KB PNG
>>999144
Shelling is quite fickle when you have multiple joined surfaces. At least in plasticity, it tends to give undesirable results when you go too thick. It's already as thick as sheet metal, which is what I was going for, and even thicker in some places, like the interior of the helmet. Plus, I like the look of cloth padding. I'm thinking maybe a leather lining, but I haven't decided yet. Something like pic related.
>>
>>999147
Maybe velvet. I think that might fit better than leather.
>>
File: Pose.png (680 KB, 2560x1400)
680 KB
680 KB PNG
I had to follow a tutorial, but I got my first working rig. While the IKs work fine, I don't know why, after setting this up, I'm unable to manually rotate the shoulders. Has anyone got any good guides that could help?
>>
File: Untitled.png (2.8 MB, 1772x1761)
2.8 MB
2.8 MB PNG
Making a Scavengers Reign/Moebius inspired environment
Need to figure out how to modify the shader to get some of the shadows back into the scene
and work on the general landscape and foliage
>>
File: 1709655957970628.webm (250 KB, 920x1080)
250 KB
250 KB WEBM
keeping on with the arm since its really the basis of everything. Going to rebuild the arm again.
>>
>>999221
big flexa
https://www.youtube.com/watch?v=8g0v2SDe6MU
>>
File: wip5.png (3.76 MB, 1920x1080)
3.76 MB
3.76 MB PNG
>>999095
>>
File: Elf.png (2.06 MB, 1350x2160)
2.06 MB
2.06 MB PNG
https://files.catbox.moe/y4dbyq.png
>>
File: 1702853202555626.jpg (98 KB, 477x570)
98 KB
98 KB JPG
>>999368
>>
File: mantis3.png (292 KB, 1230x1294)
292 KB
292 KB PNG
mantis girl wip
>>
File: 0001-0250.webm (489 KB, 640x480)
489 KB
489 KB WEBM
Don't mind me I'm just testing parameters on my IK rig.
>>
File: Volume Color test 01.webm (307 KB, 844x592)
307 KB
307 KB WEBM
>>999575
I hate everything to do with UVs and texturing. So today, I thought I would take a swing at coloring the model by volume. I got the nodes set up to create proximity with points that carry color information. And I was able to balance the colors, so they're all normalized perfectly. A couple problems were immediately apparent.
1. Because it's all proximity based, the points of the mesh often find themselves in a midway point between two colors. Resulting in a lot of color overlap that makes it all look blurry.
2. Again, because it's proximity based, the mesh looks all spotty and radial.
So, the difficult task, was to somehow give certain colors priority, to reduce the blue. While simultaneously making neighboring color points blend connect on the mash, eliminating the spotty look.

I managed to reduce the spottiness somewhat. But it's still apparent. I haven't figured out how to eliminate it.
I figured that by increasing the amount of points to sample from, it will reduce the spottiness enough to trick the eye. So I spent a lot of time working out how to do that. Managed it, and the results came out a little better. But not good enough. Still king of spotty in places, and the colors are still mixing heavily. I tried my luck at fully interpolating the points, and that made it worse. And by "fully", I mean that my first interpolation triangulated the points first, and then only interpolated between points that could see each other. While the second interpolation didn't have the triangulation limitation on. So it interpolated all the points. Turns out my triangualtion node is good for something after all, seeing as how it produced better results.

But anyway, the way I'm doing this really is a fruitless endeavor. I can never get precise enough control of the color via geometry nodes. I did learn some few things though. Node comprehension raised.
>>
>>999586
Whatever you write is making less and less sense as time goes on.
>>
>>999592
not him, but i would be frustrated if I was UVing and texturing in blender, too. It stinks. I suggest maya and substance painter
>>
File: Head modelling 01.png (777 KB, 1481x552)
777 KB
777 KB PNG
>>999592
Yeah, sorry about that. I didn't check my post before submitting. Reading back, I see some spelling and grammatical errors.

The short of it is this:
I can make the mesh read color information from floating points. But it doesn't look good. I tried to make it look good, but haven't found a suitable solution. Now I'm sad.
The silver lining, is that I leveled up my node skills, and I think my face topology is a lot better. The edge flow around the eyes in particular makes me happy. I finally feel like I made solid eyes. I can move them around, and they remain smooth, and fit well around the eyeballs.

The actual face shape isn't quite where I want it to be yet. But it's getting there. I started the face two days ago, and it was atrocious. Yesterday, I did a lot of work to make it appear human. Just human was enough for that day. Today, I got it looking more aesthetic. But I think it needs more work. It's not an economical head. Standing at 11,940 tris.

>>999596
Fuck... do I have to? The very idea of textures actually offends me. Like, how are we still using rasterized graphics to create color information for 3D graphics? Makes no sense when you really think about it. You should be able just "fill" the area around the model with the color you want.
>>
>>999607
It's not because of spelling and grammatical errors. It's that you don't use words correctly.
With Ginecology Nodes alone and no math foundations you're not going to get anywhere.
>>
>>999628
I'm curious to know how I'm using words wrong.
True, my math sucks. But I'm steadily increasing my understanding. I tried to get help from /sci/. And they did help a little. But they aren't very verbose over there. I guess math freaks all have the non-verbal kind of autism. Here are my posts:
>>>/sci/16445040
>>>/sci/16445234
>>>/sci/16446497
Barycentric coordinates may be what I'm looking for. Except the more I read into it, the more it seems to only fit a 3 point system. While I need something that works for an infinite amount of points. It's too hard for me to grasp anyway.

Regardless, I don't think the math is really the problem here. If I'm being honest with myself, the task my be impossible with nodes. I'm afraid of the reality that true control over blender will come from learning python. Which I really don't want to do.

I'm looking at "generated" uv coordinates now. It may be what I was looking for. A UV system that simply puts a box around the model, and then projects the texture onto the model by the box coordinates. I'm struggling to make it work right now. Not sure why texturing painting in material mode shows no change to the texture. But in the panel displaying sold mode, it does show the change. Weird. Blender's interface still feels so jank.
>>
>>999630
Your lack of math foundations is the problem because it makes it impossible for you to understand the answers people give you.
>>
>>999645
Fair point. However, an explanation would still help regardless. They speak the language of math. They know the lingo, the symbols and the shorthands. I don't. But, I'm still capable of understanding concepts when they're explained in a plain manner. Which nobody seems wants to do. Or maybe they're just incapable.
>>
>>999649
I remember an exchange months ago where I told you something like: "you need to establish a metric" and you replied: "if by metric you mean..."
But the meaning of the word "metric" is not subjective. It's defined in math and it refers to a formula like "L = sqrt(a dot a)".
>Which nobody seems wants to do. Or maybe they're just incapable.
It's 100% a you problem. I've seen you ask variations of the same question for a year. I've seen several people giving you the correct answers several times and I've seen you ignoring the answers you were given.
>>
>>999654
That doesn't sound familiar. No doubt I've made an ass of myself in the past. But I don't recall a situation where a metric was mentioned to me. And then subsequently, I try to redefine what metric means. Because I haven't been doing anything where a metric is relevant.(I don't think) And I probably wouldn't try to redefine a metric. But rather beg for a simplified explanation of what a metric is. So you probably have me confused with another guy.

I have been steadily improving at math. Last year, I had a poor grasp on normals and dot products and that kind of thing. The year before that, I had no fucking clue what those were. But today, I can visualize most vector functions in my head. If you tell me to get a dot product, I won't be confused like I used to be. I can just do it. Add, subtract, multiple, divide, normalize. These vector math concepts are memorized now. Rotation is still a little iffy.

I also don't recall receiving an answer for my current problem. I've asked a similar question before. Where I wanted to know how to tell if a point was inside of a triangle. That involved doing some area math with triangles, that I wasn't accustomed to doing. But I did eventually figure that one out. You're probably confusing that one with this one, because both illustrations show a point moving within a triangle.
>>
>>999656
You're ignoring me.
>>
get a room, autists
>>
>>999658
autism website
>>
File: 0001-0250.webm (530 KB, 640x480)
530 KB
530 KB WEBM
I was watching BCON and they said that "Animation 2025" is not going to happen.
>>
File: 0001-0250.webm (640 KB, 640x480)
640 KB
640 KB WEBM
I've fixed like 100 bugs in Blender's IK system but there are 1000 more and I don't trust it, still I'm amazed I was able to make it work at all.
>>
>>999663
Nice, what sort of things did you find?
>>
>>999664
There's a lot of sloppiness which dates back 15 years. If you're able to read C code:
https://sourceforge.net/projects/snes9l/files/blender-2.79b-debian12.diff/download
It includes a fix for a self-shadowing bug that's in the internal renderer and, oddly enough, in Cycles too. I was planning to write an article at some point but I doubt I'll ever find the time.
>>
File: 1699815917812110.gif (145 KB, 300x375)
145 KB
145 KB GIF
>>999660
who could have predicted this!
>>
File: mh_itasc.jpg (284 KB, 1440x797)
284 KB
284 KB JPG
For >>999660 and >>999663 I've left whatever comes out by default from MakeHuman untouched including the default skeleton. All I've done is I've attached some IK endpoints to the hands, feet and head. I didn't touch the mesh, I've done no weight painting, no shape keys, no drivers, no Geopolitical Nodes. Just bug fixes >>999666 to the IK system.
If you're a Blender user you'll likely never going to experience anything like this because looking at BCON, they're going in a completely different direction.
>>
>>999607
We use textures because we have hardware that can draw billions of triangles a second with fixed functions for mapping textures to the perspective of the triangle.
3D graphics for surfaces is just approximating the contribution of light on that surface (or subsurface with materials that let some light scatter through them).
11,940 tris isn't bad for a face if it is the highest level of detail and you're not going to be doing any extreme close up shots of the eyes or mouth.
Your topology looks really good.

You should watch women's makeup tutorials for a headstart to making petty women.
>>
>>999595
Substance painter yes, Maya no, Maya it's a waste of space
>>
>>999698
The thing that about 3D, that I have trouble accepting, is moving on to the next step of the process. I keep seeing ways to improve topology. But editing the mesh sees to break everything.
When you work with blender's particle hair system for example. If you make a change to the mesh while the hair is still connected. Then it causes your hair to scatter all crazily. The simple solution is to first disconnect the hair, and then work on the mesh, then re-connect the hair. Except, what if the hair is already hidden. Out of sight, out of mind, right? You go to change the mesh, without remembering to disconnect the hair. You make your changes, and save your work. You just saved the fucked up glitched hair, losing hours of work.

Similarly with textures, if you make changes to the surface of an object, the UV map gets all messed up, and you have to meticulously push it back into place. God forbid you unwrap the UV again, it re-proportions everything to account for the new geometry. Making everything misaligned.

These issues would be less of a problem if I could just stop fiddling with the mesh. But I can't help it. The more I look at it, the more I see things that can be fixed or improved upon. The idea that I just have to leave well enough alone, and move on to the next part of the process, really bothers me.
>>
>>999706
Baking and projecting exist, this is not the 2000, you can break a model and reproject textures in a few minutes
>>
>>999709
How does that work? If I delete a face, then what does the texture bake/project to?
>>
>>999706
you should go through the standard process once to learn the hard way but there are way around most things now.
wrap3d will let you wrap bad topo to good basemeshes
will even transfer the texture over to match the new uv's
for character grooming, you never want to do the groom on the actual character mesh. separate the scalp mesh and groom on that. then you project and attach the scalp back onto your character.

that said, you should avoid constantly changing your topology because it's a PITA.
>>
>>999706
I would look into blenders hair geonodes. The old system is ass and not worth anyone's time.

>>999709
This would be good advice but Blender's baking is dogshit for retaining details in the original model. You will lose a lot of information doing this.
>>
>>999711
>separate the scalp mesh
just to be explicit, i mean duplicate and separate - don't just leave your character mesh scalpless
>>
>>999711
I know of the scalp trick. Why does it need a scalp anyway? The strands of hair all have their own point in space. They should just work relatively off that
Even Geometry node hair requires that you have a clean UV unwrap, in order to make the hair work properly.(Overlapping UVs cause confusion and slow down for geo hair)

I'm probably going to do geonode hair. But honestly, I get lost in the sauce. Nodes are convoluted. Even trying to use their prebuilt assets, it still requires you build the nodes in the correct order, and then you have a hundred variables to tweak. I'm just whining now. I just hate how fragile everything feels in 3D. It's like setting up a house of cards. The taller the stack, the more prone everything is to falling apart. I never feel like "ok, that's locked in." It's always "I hope this doesn't break." And then it breaks.
>>
>>999718
Maths is hard anon. Data structures for flexible creativity are going to leave you aimless.

If you want to just make a female goblina then use DAZ or just knuckle down and spend time trying shit and reading documentation/source code.
>>
>>999718
the roots for hair curves will store their position on the mesh either using:
1. uv space
2. primitive number + intrinsic primitive uv's from the rest pose

this allows them to look up attributes from the mesh or from uv space, including but not limited to, colour and animated position information. these position lookups are fast and precise.

i haven't done a groom in a bit, but in houdini it is possible for guides/groom to exist separate from the mesh/scalp (usually something's gone wrong if you're here). houdini will actually allow you to just do a new look up from a different mesh. in this case the 'fragility' you're feeling is probably because:
1. you don't know what you're doing
2. most dcc's are designed as programs for artists and will try and abstract away a lot of stuff and ask users to work in certain ways so they don't have to deal with technical problems
for 2. there are usually tools buried somewhere for technical artists to come in and save the day.
>>
File: Shape key eyes.webm (528 KB, 892x658)
528 KB
528 KB WEBM
Made some changes to nose and mouth. Now there's a lot more geometry to shape the nostrils and nose tip, and bridge gets 6 sides instead of 4, for more roundness. Gave lips more geometry for shape. And I found the magic loop that made the inner eye corners look more natural. They eyelids feel even more solid before. So I celebrated by making a shape key. Look at that topology. Clean.
Also played with hair nodes. But it looks really shoddy right now.

Stumbled into a node called "point density". It sounds like a node that's made to perform the color volume stuff I was doing. Except, I can't get it to work that way. And all of the tutorials are like 9 years old. using the old versions of Blender. I found 1 tutorial that shows it working how I want, but following his steps doesn't yield any results. It literally just creates blank black material. And the comments are off on the video, so I can't even see what others have to say about it. https://www.youtube.com/watch?v=KogFovV-QoA

>>999722
>If you want to just make a female goblina then use DAZ
Nah. I want to make my own thing.

>>999723
What's DCC?
>>
>>999825
And of course editing topology breaks shape keys. I edit the the cheeks, then deleted half the mesh and symmetrize. Somehow, the right eyelid now malfunctions when the shape key is turned on. How the actual fuck does a symmetrized mesh, not have symmetrical shape keys?
>>
>>999831
editing topo breaks shape keys but it doesnt break what is actually being used in the industry for years (see avatar way of water) which doesnt use shape keys but simulation driven by mocap.

tldr. Dont use outdated shape key workflow
>>
>>999834
The way of water uses machine learning for driving facial deformation because they had the time and money to crunch all that facial data.
>>
>>999836
the way of water uses mocap and tissue simulation

https://www.youtube.com/watch?v=IPQ5vTqqdgE
>>
Welp, blender started slowing down for no reason. Went down every object and checked the modifier stacks and materials. Can't find the culprit.
I guess that's enough Blender for one day.
>>
>>999713
>This would be good advice but Blender's baking is dogshit
You can use the dozens of free software to bake, I never said anything about blender
>>
>>999710
To the original texture anon, you want to modify the mesh and keep the texture.
You project the old mesh on the new one with the new UV and bake the textures from the old mesh.
>>
File: 1587346362803.gif (682 KB, 198x247)
682 KB
682 KB GIF
should i buy one 180-day premium or two? will there be another premium sale in the next 180 days?
>>
>>999869
FUCK wrong board ENTIRELY
>>
>>999837
>avatar 2009
Don't be wrong on the internet.
https://www.fxguide.com/fxfeatured/exclusive-joe-letteri-discusses-weta-fxs-new-facial-pipeline-on-avatar-2/
>>
>>999834
>>999836
>>999837
>>1000054
Doesn't matter, because I'm not going to use AI or Motion capture for faces. I'm not shooting for full realism. Face capture was a mistake. Everything looks uncanny now, because everything uses face capture.

Also, I figured out what is causing blender to lag. For some reason, my mere 12k face mesh, magically obtained over 2 million verts. I tried all kinds of things to make it go away. But I just couldn't find the excess verts. Eventually, I used select by trait: loose geometry. It selected a single loose vert. This vert, I place deliberately in order to always have a centered vert to help with symmetry. Since I can always rely on it being dead center to the X axis, even if everything else changes.

Well... somehow all 2M excess verts were concentrated in that one position. So they all appeared like a single vert. Amazing. I'm not sure what I did to get 2M verts concentrated there. I have no clue when it happened.
>>
>>1000054
You didnt watch the video. The video is clearly Avatar 2. They go over all the new tech. READ the description of the video, it goes all over Avatar TWO
>>
Invalid surface UVs on all the curves. Why?
Why does god sent Blender to me to make me suffer?

I don't know what causes this, why it matters, or how to fix it. I'm entirely lost. First of all, what even is the surface deformer even doing? Secondly, why is it a unique node that can't be created by the user? So if I delete it, then what? It's just fucking gone? How do I get it back? How do I reset it?

Hang on, let me consult the manual.(lol)
Wait shit, the manual actually helped for once. It was in a roundabout way. But it helped. I have to begrudgingly admit. If you run into this situation, here's the problem:
You changed the topology and/or UV of the mesh. The curves don't know where their roots should be on the UVs anymore. You have to tell them to re-detect the surface. Go to the hair's sculpt mode, and then Curves > Snap to nearest surface. That should do it.
>>
File: Colosseum second levelS.jpg (487 KB, 1920x1440)
487 KB
487 KB JPG
How do I make the UV mapping for each instanced of these marble columns (array+curve modifier) to be different from the last, so that they don't all look the same
>>
File: 001.jpg (208 KB, 1765x969)
208 KB
208 KB JPG
>>1000077
There's a UV offset option in the array modifier that you can use to offset each instance's UVs. 0.5 or 1.0 will probably give shit results, but anything that aren't those values should work. You can offset the U and V separately, so I'd do that as well.
All of this is assuming your columns aren't using fancy pants unwraps and are more or less laid out pretty flat on the UV space.

If you're using something like object coordinates in your shader, you can use something like this to offset them as well. You still need UV coordinates for it to work, but you don't have to do any work unwrapping your objects, just go into edit mode and U>Unwrap and take it as is (they can even be scaled to 0 as well). Then just tweak the UV offset in the modifier to get it to shit out a random value per object.
Then just use that random value to drive XYZ location of a texture coordinates node (optionally using a map range in between to make it go more than 0-1).
That's all slightly more advanced though, so just using the modifier's offset would be easiest.
>>
>>1000095
Thank you fren, I totally missed the UV property
>>
>>1000101
No problemo. A few modifiers have that as well. Mirror is one I know for sure, not as useful with that one, but there's certain cases where it's perfect.
>>
>>1000059
You're an idiot and I have work to do.
>>
>>1000069
make your own thread if you want to write blogs about every retarded mistake you make, bro
>>
File: GET THAT MOTHAFUCKA.png (916 KB, 1083x822)
916 KB
916 KB PNG
Finished modeling and texturing a piece of clothing. I only need to cleanup the weights and it should be done
>>
>>1000134
what? The video came out when avatar 2 came out. It details what avatar 2 improved upon avatar 1 and MCU stuff. It shows this in video and writes about this in the video description. What is wrong with you?
>>
File: cassida2.jpg (442 KB, 1920x1920)
442 KB
442 KB JPG
>>999130
Almost finished. I was going to put some buttons on it, but I'd hate to ruin it with too much stuff. Maybe I'll put some stitching decals on the inner fabric, and of course the mandatory, half torn off pokemon stickers.

I'll model the female next.
>>
>>1000140
It's coming along nicely but what is that? Some kind of covid mask for LGBT folks?
>>
>>1000144
It's a helmet for ladies (female (not ma'ams)).
>>
File: 0001-0250.webm (655 KB, 640x480)
655 KB
655 KB WEBM
More than the usual bounciness, I'm trying to capture the effects of gravity on certain body parts.
>>
File: 0001-0250.webm (797 KB, 640x480)
797 KB
797 KB WEBM
I've never done hair simulations before but I've given it a try.
>>
File: msquestion.png (518 KB, 2560x1440)
518 KB
518 KB PNG
Not sure if this is progress post worthy, but it's the first not a tutorial project I've tried. The blade tip needs adjusting and the guard needs much more work, but it looks like I'm on track to finishing something for once.
>>
>>1000191
Nice, looks good anon, keep with the good work
>>
>>1000144
Looks alien I like the design, not sure about the colour
>>
>>999095
T-bot? That you?
>>
File: Head modeling 02.png (812 KB, 1115x573)
812 KB
812 KB PNG
>>1000135
I guess I was oversharing a little bit. I'll try to contain my complaining.

>>999825
Made some changes to the topology yet again. More looks going down the length of the nose. Felt it was necessary to sculpt the planes of the nose. Now the bridge is distinct. The nostrils are better defined. Still not sure of the topology is optimal, so it might change again.

I finally buckled down and learned hair. I think I get it now. Like really get it. It finally clicked, and I'm comprehending. But even so, it's not easy. It's going to take some time to refine. It's pretty messy as it is now.
Glad the model finally has eyebrows too.

Created a procedural skin texture. And then painted in the pink around her lips and eyes, and some blush here and there. The UV map is atrocious. >>999857 I need to learn this now. Is there any more insight you can give me? I'm not really understanding what you mean. Note: I never baked a texture before. And what does "projection" mean in this context exactly?
>>
>>1000191
Looks very nice and clean, anon. Good work. Will you keep it low poly, or will you add a subdiv?
>>
File: totk master sword.png (674 KB, 966x445)
674 KB
674 KB PNG
>>1000191
You didn't take my advice about connecting the grooves. Check out the official model, they use triangles to connect the grooves, like I suspected. Though, admittedly, I was wrong about two of the spots I circled in my previous image. Still, the other 3 spots are simple triangle connections.
>>
>>1000146
>>1000153
I've asked several times and it's been explained to me several times but I still don't know what the fuck is happening here. What is the point?
>>
>>1000138
awesome!
>>
>>1000284
I make those short clips when I'm working on some aspect of my own "animation system", which is really a number of patches on top of Blender 2.76. For testing purposes.
Since the Blender developers seem to be busy discussing HRT brands or whatever they do at the compound and I happen to be a developer as well... But don't worry because nothing of what you see is ever going to become mainstream, just enjoy the clips if you like them.
>>
File: msquestion.jpg (396 KB, 2560x1440)
396 KB
396 KB JPG
>>1000274
The guard has taken me twice as long as every other aspect so far, so I'm tempted but gotta do it right.
>>1000277
I put some but not enough, didn't have time to do the grooves before work. I didn't have a reference of the top of the wings, didn't know they were diamonds. Making the golden crest on the hilt was easier but makes the inset for the pommel harder.
>>
File: file.png (134 KB, 391x544)
134 KB
134 KB PNG
>>1000207
The color isn't set on stone. I myself am not really satisfied with it. It's too aggressive. I want something softer. I'm hoping for something that will make people think, "oh, that's a flower!" Maybe a soft pink with the inner petals in a lighter tone.
>>
>>1000316
it looks a lot better with opaque materials than with the transparent glass material. I could see it looking good if it stays opaque and was textured nicely, but as of now with the glass it looks a lot worse than it does in the grey opaque viewport
>>
File: file.png (111 KB, 470x532)
111 KB
111 KB PNG
>>1000320
Thanks for the feedback. Do you think it's because the glass just looks bad, or is it the interior that is not interesting enough?
I could make it into an opaque visor, but that's a completely different aesthetics than what I'm going for... It's supposed to be cute and delicate, not menacing.
>>
>>1000287
Thank you anon-kun, I'm glad you like it
>>
File: cone head flower edit.jpg (700 KB, 2834x1920)
700 KB
700 KB JPG
>>1000316
If you're going for a flower bud, then I think the design is bad. Because all of the detail that marks it as a flower, are very high and on top of the head, where it's barely visible at all. Looking from the front view image, it just looks like a huge cone head with a notch in it. I thought you were going for cone headed aliens.

What I would do, is begin the bevels of the petals much lower on the helmet. the outer most pedals would look better lower. Like one of the outer petals should bevel around the ear piece. The visor itself is part of one big petal, and so it terminates lower too. And then fill in more petals from there. You don't have to follow the lines I drew exactly. I might have drawn too many petals. Dunno. But it should get the idea across.

As for color, perhaps make the outermost petals a different color from the rest. As flower buds tend to show, the outer/lower petals seem to take on the color of the stalk. Which is typically green. But you can make it not-green. Since purple appears to be her color scheme, then make the outer petals purple, so they fuse with the rest of her suit. And then the rest of the petals are that gray color metallic color you're using for trim. Or perhaps, a brighter whiter tone. But then it transitions from white to a third color. Like a pink. Again, you don't have to follow my color scheme exactly. But I'd advise not sticking with a single monotone purple for the entire head. It makes everything look too uniform. Doesn't help convey a flower.
>>
>>1000378
Thanks. That makes sense. I think I got the method down now that I can do this in a few hours instead of the days it took me the first time around.
>>
Does anyone have a vid or can tell me how to do animations with multiple camera angles? As in switching between cameras/removing them after the angle is done with? Not sure if I've explained that very well.
Is it just a case of having cameras in the scene and keyframing them as you need them?
>>
File: hat_problem.png (811 KB, 990x623)
811 KB
811 KB PNG
I have a question about a problem: I imported my fucking model to mixamo and downloaded the animation with mesh so i can have the rig ingrained in the mesh since i dont have a rig, but the fucking model came back with wrong materials like the hat which is transparent and i don't see any fucking way of fixing it.
>>
>>1000316
Why not a gradient? Like those lotus flowers
>>
>>
>>1000529
The turnaround is shit, but the model is good. I think the texturing is a little to clean, but that's really just my personal opinion. Not everything needs to look weathered.
Nice job, anon. Proud of u.
>>
File: file.png (2.73 MB, 1920x1080)
2.73 MB
2.73 MB PNG
I rendered some gems. I wish I got more colorization on more sides of the gems, but I think this as is look really nice.
>>
>>1000531
thanks man. i think it looks a bit too clean as well, i was kinda cautious with the wear and tear because i didn't want to make it seem like i just lazily threw some edge wear and dust filters, while they look cool they also make too much of a presence, so i just made it them very subtle.
>>
>>1000561
>i didn't want to make it seem like i just lazily threw some edge wear and dust filters
Literally me when I make a weapon
>>
File: 1729694679503595.webm (799 KB, 1280x720)
799 KB
799 KB WEBM
new toon shader I started working on. Still got a ways to go but this improves on stock
>>
File: Cat Head.png (960 KB, 1119x667)
960 KB
960 KB PNG
Took a break from human head to work on cat head.
It's an old model, but revitalized with all new cleaner topology.
I'm trying to get over my fear of texturing, so I painted it a little.
Added fur. It cost a lot of tris, but not too much. Nearly doubles the tri count. Which, is about the cost of an inverted hull outline. So it's not a big problem. I just wish I had a more efficient method.
I wanted to do something very different for fur. Inspired by this video: https://www.youtube.com/watch?v=-wfgxtbXPPk
Tried to recreate it in geometry nodes. But it seems some coding is required to truly make it work.
Also encountered what I think is a bug with sampling points in geometry nodes.

I also revamped my eyes system to require zero geometry nodes. And a lot fewer shader nodes. Basically, I figured out that using texture coordinates set to object, and making the object an empty, get all that data I was transferring from geometry nodes automatically.
Discovered a pretty bad flaw in my eyes system though. Eyes can look wonky from the side. But luckily, it should be fixable by just having good eye topology.
>>
File: baiku.jpg (49 KB, 975x465)
49 KB
49 KB JPG
Makin a hover bike
Not sure if I really like it, but I'm probably just gonna finish it since I want to make a good number of these.
>>
File: 64356345362.png (1.8 MB, 1586x928)
1.8 MB
1.8 MB PNG
im so fucking slow holy shit.
>>
>>1000584
Are you using blender? If so, are you using the layer weight node?
>>
>>1000591
Don't give up alchemy lab kun, I believe in you, 's looking good so far
>>
File: 1706788157263125.webm (874 KB, 1280x720)
874 KB
874 KB WEBM
>>1000597
using maya. I tried blender, but couldn't get the results that I wanted. Here is a version 2.
>>
File: 1728256882372157.webm (2.62 MB, 1280x720)
2.62 MB
2.62 MB WEBM
>>1000643
test with imported suzanne. Getting closer (have constant line widths of a few set widths that respect dot product) but still am missing a lot from the shading technique I use when I execute a specific rendering style in ink in traditional media with fineliners
>>
>>1000652
Not bad, but I think the problem with toon shaders is the ambiguous shading of large areas that don't really mesh well with the line. Like it doesn't come across as actual shading like you'd see in an illustration, but moreso just a 3d artifact. This isn't just a problem with yours though, it's an issue with like 90% of toon shaders.
Which is why I think the best toon shaders don't even bother shading. It's mostly flat colors and lines.
>>
File: Cat Head 2.webm (3.88 MB, 1202x570)
3.88 MB
3.88 MB WEBM
>>1000586
Check it out. I figured out how to make fur with textures.
It took me HOURS to figure out how to give these faces proper UV coordinates. They're all individual squares, so it should be easy. All I had to do was plot the 4 corners of each square to the coordinates (0,0,0) (0,1,0) (1,0,0) and (1,1,0). That's literally it. It took me FOREVER to figure it out. And the solution ended up being relatively easy. My solution avoids the UV unwrapping node too. Which is great, because that node is slow.

After that, I stumbled into another problem. The textures were not in the proper order. As in, ones that should be in front, were in back. It was some kind of depth issue. At first I thought it was because all the squares shared UV coordinates. So I fixed that by making them sit next to each other. Because the texture repeats, they still showed the texture normally, save for the depth problem that persisted. Then I changed the blending method from alpha blend to alpha hashed. That fixed the problem. Don't know why, but alpha blend seems to have depth issues. Which is a shame, because it blends better than alpha hashed, which is all grainy. Alpha clipped also works, but I don't prefer the sharp edges it creates.

After that, I came up with a way for the fur to lay down. I might not be able to sculpt it. But at least, when it's laid down at an angle, it covers more of the underlying mesh, and it appears to give the fur a pattern. Rather than just sticking straight out.

From here, I can improve it by making better and more varied fur textures. Plus, refining the weights that affect fur length.

I also fixed the sampling problem from before. It wasn't a bug, it was user error.
>>
File: 1716721536237067.png (1.09 MB, 1594x560)
1.09 MB
1.09 MB PNG
>>1000693
Thanks for the feedback dood. This is what I'm going for. Line weights set by physically switching to different set width fineliners. A "reflective" style where the outside edges have lots of specular fresnel, always. I saw this style in a book on drawing and shading comic book super heroes

I am trying to translate this into a shader as a challenge. I tried the built in toon pipeline of prman to get the above, but now I feel like I have to explore the prmanOSL node
>>
>>1000708
looks like osl in prman has practically no features and blender is much more full featured
>>
File: sos yo.png (318 KB, 366x871)
318 KB
318 KB PNG
>>998180
I am this anon >>1000138
Made more clothes for a not-Fallout but smaller in scope (you have a safehouse and select locations a-la Hitman Freelancer, instead of there being an open world) and with anthros I'm creating with a coder man. We're using Unreal 4.
If any sound + music guy is interested to join our team feel free to share your portfolio, if your work suits our project I'll get back to you, cheers
>>
File: file.png (191 KB, 846x808)
191 KB
191 KB PNG
I was able to redo the petals with minimal effort yesterday. That's a lot more petals, redoing all the nice washouts will take a while. Whether it will look like a flower or not only after shading to know.

>>1000511
That's an interesting idea. In the end, I think I'll end up having to unwrap this and paint a base layer by hand.
>>
File: file.png (601 KB, 1920x1080)
601 KB
601 KB PNG
>>
File: file.png (697 KB, 1920x1080)
697 KB
697 KB PNG
>>1000774
>>
>>1000138
furry freak
>>
>>1000789
I can't deny it
>>
>>1000795
Based
>>
File: Projected circle.webm (1.22 MB, 426x380)
1.22 MB
1.22 MB WEBM
Reposting from /v/.
After identifying a problem with my eye projection method, I got to thinking on how to really make a circle that's impervious to edge distortion. After some experimentation, it seems like a way to do that, is to project everything onto the camera. This is something I've done in geometry nodes before. But shader nodes work a little differently. It wasn't as intuitive. Still, after some node bashing, I got it working.

The webm displays a circle getting distorted by edges at first, and then it's swapped out for the new nodes, which display a circle impervious to edges.

I'm still uncertain about how to make the circle stay in place. Because it's a projection back onto the camera, the circle is always centered on the screen. I need to figure out how to offset the position in such a way that it sticks to point in space.
>>
>>1000138
>>1000761
Oh shit nice. Do you have a wireframe?
>>
File: sos.png (491 KB, 1034x857)
491 KB
491 KB PNG
>>1000836
Thank you, and sure, here's a general view one from the front
>>
>>1000223
You can use rightclick>"Pin" in UV editor to lock in the unwrapped vertices. If you hit unwrap again it will unwrap every "un-pinned" face based on the pin vertices
>>
>>1000853
Sounds nice. I can't make it work though. I can pin the vertices just fine. but selecting unwrap makes all the unpinned vertices unwrap in full scale. taking up as much space as they can. Not following the flow of the pinned vertices at all.
>>
>>1000866
Did you ctrl A> Apply rotation and scale ? Also on the UV panel , press D or check bottom left for the Unwrap settings , try messing around with them
>>
>>1000874
>Did you ctrl A> Apply rotation and scale ?
They're applied and neutral.

>press D or check bottom left for the Unwrap settings
??? Pressing D gets zero reaction from blender. Literally nothing to see in the bottom left. What function did you think was assigned to D?

Wait, I just figured out the problem. After I pinned the vertices, I was independently selecting the loose vertices and unwrapping them. The solution was to simply select the loose AND pinned vertices. I guess I should have tried that, but I didn't think of it.
>>
File: rrat.png (240 KB, 678x791)
240 KB
240 KB PNG
>>
>>1001034
this rat seems familiar, is it from an anime or something?
>>
File: scarab.png (665 KB, 1040x821)
665 KB
665 KB PNG
Sketched out a ref sheet to model from
Reference on the right
Beetles and bugs are icky as fuck though, was feeling queasy just looking at IRL references.
>>
>>1001058
Beetles are beautful though how could you possible be yucked by those? They're adorable, you should look at the sheer intent those little fuckers push balls of shit around like their life depended upon it, it's inspiring man
>>
File: GUUfJmWWUAAEhco.jpg (296 KB, 1685x2000)
296 KB
296 KB JPG
>>1001055
Friend's OC heavily inspired by Sakuya
>>
>>1000429
is it transparent or the normals are just flipped?
>>
File: f.png (339 KB, 598x796)
339 KB
339 KB PNG
Super super cute
>>
File: color.png (45 KB, 324x611)
45 KB
45 KB PNG
>>1001085
>>1001140
that's why it felt familiar, an aspie, it's a small world
try setting your color management to Standard when working on NPR models, Filmic is the default for Blender and it looks dark and washed out
>>
>>1001141
That's something I noticed as I was texturing, Blender's colors looked washed out and nothing like the colors I was putting down in the texture itself
By the way, what does NPR stand for?
>>
>>1001143
Non Photorealistic Rendering, as opposed to PBR which stands to Physically Based Rendering (realistic)
>>
File: fira.png (312 KB, 524x761)
312 KB
312 KB PNG
>>1001140
Texturing is so exhausting
Specially when I have to export to single layer > confirm overwrite on CSP every time I want to see my changes
>>
>>1001221
Right, don't texture 3d models in clip studio paint.
If texture painting in blender bothers you that much, use an external program that's designed for 3d texturing like substance painter.
But really texture painting in blender should be fine.
>>
>>1001231
>substance painter
>SaaS shit
>Adobe shit
No.
>>
>>1001232
Right, I mean it's your workflow, whatever, but I really don't see what about the texture on the character really requires you to open up an external program to paint onto a UV mapped texture instead of, you know, painting onto the model directly.
>>
>>1001231
How do you keep substance painter from shitting all over the entire texture map when you paint on model and happen to get a little too close to the edge of the UV? Its horrible. Shit scribbles all over the place.
>>
>>1001236
I don't use substance painter, I manage with blender's built in features.
It's admittedly not the best but it beats manually editing in CSP.
Texture paint has that problem too sometimes, just make sure you're only painting on selected faces and if there's a problem with seams use your clone tool and have a large enough margin between UV islands.
>>
>>1001231
Blender's built-in texture painter leaves a LOT to be desired
CSP's brush engine alone makes it my go-to for anything 2D. Plus, I find it much more convenient to just export the UV layout and overlay it on top of my texture in CSP and work that way. Realtime updating is literally the only thing missing from my workflow
>>
File: 0001-0500.webm (2.4 MB, 640x480)
2.4 MB
2.4 MB WEBM
Skin stretching test
>>
>>1001406
mogged by the cute lil rat maid a few posts above
>>
File: link_head.png (2.1 MB, 2556x1411)
2.1 MB
2.1 MB PNG
Hey guys, I spent 3 hours modeling ears then when I was finished turns out Link has elf ears.
>>
>>1001451
im mean just go into scuplt mode and drag out those ears until they look elvish
>>
File: dubbie.png (961 KB, 1301x847)
961 KB
961 KB PNG
>>998180
Modeled, rigged and textured my double barrel shotgun, for the most part anyway, I only need to add some damage detail to the Diffuse and generate height + roughness + metallic maps. Might make some color changes too
>>
File: fira_thumbs_up.png (343 KB, 437x813)
343 KB
343 KB PNG
:thumbs_up:
>>
File: sos.png (1.14 MB, 1446x831)
1.14 MB
1.14 MB PNG
>>1001515
It's done, a double barrel shotgun and a sawn off variant I easily derived from it
>>
File: sos.png (905 KB, 1363x734)
905 KB
905 KB PNG
>>1001579
Textured the shells as well, NOW it's done
>>
File: 0001.png (2.11 MB, 1080x1920)
2.11 MB
2.11 MB PNG
frogga
>>
I wish /3/ was as active as /ic/
>>
>>1001653
I don't.
>>
File: Head_sculpt_female.png (1 MB, 1920x1080)
1 MB
1 MB PNG
What does /3/ think of my third ever character head sculpt attempt?
>>
File: Head_sculpt_girl.png (1.06 MB, 1920x1080)
1.06 MB
1.06 MB PNG
>>1001685
Here it is with added hair, I didn't make the hair, I just added it to see how it would look.
>>
>>1000134
>2m verts in a point
Sounds like you're the idiot.
>>
>>999134
>To be reborn as my fursona.
>>
>>1001515
>>1001579
>>1001601
this looks awesome, good job
>>
File: 2024-11-15_T1820_0002.png (160 KB, 745x767)
160 KB
160 KB PNG
Here's a sculpt I made today. I'm still learning how to make animal faces.
>>
File: firagun.png (338 KB, 631x754)
338 KB
338 KB PNG
(untextured) gun
>>
File: point.png (534 KB, 965x858)
534 KB
534 KB PNG
pew pew
>>
>>1001761
>>1001779
It's a nice, tight, low poly model. And yet, I still think it would look better if you spent more polygons rounding out the ears.
>>
File: firagun_1.png (677 KB, 985x844)
677 KB
677 KB PNG
>>1001848
It's not low poly, I tried the low poly style for this model initially but I'm too used to quads and good topology and every time I made a triangle I felt like I was being stabbed. So instead I pulled out every 3D modeling technique I know and didn't particularly care about the poly count

It's around 12k polys I think (and 2k of those are the frilly skirt under the outside skirt) but I agree, it would've come out better if I had spent more polygons rounding out the ears

I also set it up for VRM for funsies. This was the initial prototype, I fixed the hair and added a blink shapekey and my friend already debuted it on his stream :)
https://files.catbox.moe/crfv6m.mp4
>>
>>1001221
there are a couple tricks, you can save your clip project file as a psd instead, Blender can read photoshop files
I've just heard about a plugin that auto reload files but I haven't tested it yet
https://extensions.blender.org/add-ons/auto-reload/
>>
>>1001852
>12k polys
Thats still considered low poly. It just isn't what people think of when they hear low poly, theyre thinking ps2 or n64 looking ass shit.
>>
File: f2 uv preserve.png (65 KB, 661x551)
65 KB
65 KB PNG
>>1001852
>but I agree, it would've come out better if I had spent more polygons rounding out the ears
You should do it, if it's not too much trouble. Looks like you use blender. So go to your addons/extensions, and go to F2, and enable the function that preserves UVs as you edit.(pic related)
It shouldn't interfere with the UVs you already have. It should only place new geometry where it thinks it should go, in relation to the UVs that are already there. Still, it would be prudent to create a new UVmap over in object data properties. Because when you create a new uvmap, it copies your current uvmap. So in case something goes sideways, you will have a copy preserved.

Alternatively, you could just billboard the ears. Using Alpha to cut out a perfectly round shape. And then sticking the billboard in the front of the rest of the ear shape.

>>1001878
12k is higher than your average PS2 game model. Check it out: https://polycount.com/discussion/126662/triangle-counts-for-assets-from-various-videogames

For example: In the first Jak and Daxter, Jak's model was 4,000 tris. In the second game, it was bumped up to 10,000.

For Zelda: Wind Waker Link was 2,800 tris. twilight Princess Link was 6,900

12k seems to be right on the cusp, where a PS2 game might push that high for a complex model. Or a PS3 game might go that low for an economical model. For example, Alan Wake's model is 13,437 tris. And Alice from Madness Returns is 11,748 tris. Those are on the low end of the PS3/Xbox36 era. That era seems to average about 18k.
>>
>>1001686
>>1001685
Looks ok. I see the vision. You have to define the lids of her eyes. It's an intimidating task, but needs to be done. Best start learning now, so you're not 50 heads deep, and still struggling with eyes. Also the shape of her lips could be defined a little more too.
>>
>>1001852
>frilly skirt
GIWTWM
>>
File: chibi2.png (1002 KB, 934x843)
1002 KB
1002 KB PNG
>>1001886
Ah, thank you anon! Yes, you are right, I hadn't made anything like that yet. But where can I get textures for it? How does texturing work?
Currently working on this btw
>>
>>1001888
You're asking the wrong guy about texturing. But I know that you need to retopologize before you think about texturing. Or, you can say "fuck texturing", and just paint on top of your sculpts. Some people do that, but it's not recommended if you're trying to create a fully poseable character that can be viewed at any angle.

So start looking up retopology and UV stuff first. Learn how to unwrap and texture a simple box first, to get the general idea, and then move up to more complex shapes.
>>
>>1001889
ah okay, thank you.
>>
>>1001731
Thank you anon-kun, I'm pleased to hear you like it
>>
File: wojakbody.png (157 KB, 886x830)
157 KB
157 KB PNG
Polish man wip
>>
>>1001852
How long have you been modeling for?
>>
File: link_progress.png (618 KB, 2560x1413)
618 KB
618 KB PNG
First attempt at body modeling, it feels like I'm just going through the motions. Gotta start somewhere right?
>>
File: hong.png (256 KB, 614x817)
256 KB
256 KB PNG
>>1001918
This was my very first on this board
https://warosu.org/3/thread/993026#p993227
>>
File: hospitalwip.png (1.09 MB, 1920x1080)
1.09 MB
1.09 MB PNG
Modelling a hospital room.
>>
File: I dunno lmao.webm (3.92 MB, 1920x1080)
3.92 MB
3.92 MB WEBM
I did this pretty much entirely in 3/4 persp, it doesn't shit itself in a basic turnaround, right?.

Anyway, i was an utter fool to think that i could go beyond donuts, i'm 2 gay 4 this.

Peace.
>>
>>1001942
wow amazing i love it
blog?
>>
File: thonk.png (262 KB, 423x744)
262 KB
262 KB PNG
Amost done, needs some rigging adjustments. Will probably sculpt details on the body one day
>>
File: thonk2.png (199 KB, 453x738)
199 KB
199 KB PNG
>>1001967
Doesn't look too bad what do you think?
>>
File: faith.png (641 KB, 1912x1056)
641 KB
641 KB PNG
how are we looking? struggling with arms and shoulders, I feel like they're always either too big or too small. plus idk what to do with the shoulder topology here to make it deform well without adding a shitload of extra geometry
>>
>>1001565
>>1001761
>>1001779
>>1001852
how do you texture?
>>
>>1001967
>>1001969
really cool anon :)
>>
>>1001981
hey I'm a retard and i just read the earlier posts, disregard this
>>
>>1001611
I like him
>>
File: hong1.png (253 KB, 515x820)
253 KB
253 KB PNG
>>
File: blockade_fortress.png (1.88 MB, 1694x666)
1.88 MB
1.88 MB PNG
>>998180
I think I'm starting to get the hang of it.
>>
File: bad faith hair.png (294 KB, 882x941)
294 KB
294 KB PNG
>>1001973
why hair so hard
>>
File: catalyst.png (531 KB, 1066x1077)
531 KB
531 KB PNG
>>1002101
adjustments
>>
File: hong2.png (440 KB, 691x861)
440 KB
440 KB PNG
beloved chinese daughter
>>
File: soos.png (1.52 MB, 1405x775)
1.52 MB
1.52 MB PNG
Made a whole bunch of furniture for our gaym project

>>1001969
That's pretty cool, it's stylized yet not caricaturized, the proportions and anatomy are believable, well done anon
>>
>>1002139
looks pretty good
>>
I joined the blendlets cos I refuse to keep paying maxon money, I finished modeling a shed for a project I'm working on just gotta UV and take it into substance
>>
File: xara_remesh.png (268 KB, 671x772)
268 KB
268 KB PNG
done with the retopo, 18500 tris. now to give ma'am some proper texture work...
>>
File: turnaround0001-0240.mp4 (3.82 MB, 1080x1080)
3.82 MB
3.82 MB MP4
donezo
for some reason the colors are super washed out in the final turnaround render even though I switched the display device in the color management tab to standard
>>
>>1002158
Must... Fuck...
>>
>>1002166
are you sure they don't look washed out because of the dark red background?
I tried color picking elements with >>1002116 and only the hair has very different colors
>>
>>1002178
Nah, not him but I had the same problem once. Some blender retardation about color space.
>>
File: _finger_.png (3.6 MB, 2304x1296)
3.6 MB
3.6 MB PNG
>>1001942
Trying to get the basic shape of this *happy camper*, needs looooots of retouching later on, Never used voronoi/gradient procedurals before, can't get my rusty ass image textures to work on the machinery either.
https://www.youtube.com/watch?v=UEEES8BeNCg
>>1001948
Pls no sarcasm ;-;
>>
File: S5678.png (112 KB, 678x833)
112 KB
112 KB PNG
>>998180
>>
>>1002153
Clean. Nice cuck shed.
>>
File: Kid named fingerSmaller.png (3.12 MB, 2074x1166)
3.12 MB
3.12 MB PNG
>>1002183
Eh, close enough, see ya later alligators.
>>
File: 1732315189489359 (1).mp4 (2.15 MB, 1280x720)
2.15 MB
2.15 MB MP4
>>1002176
she's quite picky.
>>
>>1002181
Did you ever figure out a solution?
>>
>>1002158
>>1002211
You're really good at this. I really wish I could make bodies that look that attractive. How many years have you been training to get good at this?
>>
>>1002211
Whoa, what? 4chan allows mp4s now? When did that happen?! Oh thank god. I can finally just upload without converting everything.
>>
>>1002216
I began modelling in 2019, but only started sculpting in early 2021.
her body's an anatomical disaster, lol. but thank you, i appreciate it
>>
>>1002217
yeah, i was equally surprised. honestly great, webm sux
>>
>>1002211
you did the easy part, the still model. The animation is the tricky part.
>>
File: still alive.png (533 KB, 1009x1022)
533 KB
533 KB PNG
>>1002105
idk if i want to get more complex with textures, i already fucking suck at it. have access to substance painter but i can barely figure it out so i just use kirta and blenders default texturing

>>1002211
mp4s? on 4chan? that's the one good change they've made since like... 2012.
>>
>>1002222
animating isn't really much of a tricky part for me atleast. rigging is actually the tricky part. it was also not really easy to model this
>>
File: still alive.png (374 KB, 833x1009)
374 KB
374 KB PNG
>>1002224
>>
File: Finger.png (3.73 MB, 2304x1296)
3.73 MB
3.73 MB PNG
>>1002210
My sister asked if this was supposed to be the crab from Moana, don't even know how to respond to that, 4real.
>>1002270
Are the arms supposed to taper at the elbow that way? genuinely asking, don't know if it's a topology thing for that number of polys.
Best of lucks buddy.
>>
File: faith tat top view.png (426 KB, 1311x538)
426 KB
426 KB PNG
>>1002314
>Are the arms supposed to taper at the elbow that way?
sorry what do you mean by that. should I change them?
>>
File: mirror's-edge-combat.png (1.87 MB, 1600x1200)
1.87 MB
1.87 MB PNG
>>1002317
I mean the curves going from the shoulder to the elbow, and from the elbow to the wrist, they look rather uniform to me at this point, kinda blending all together? on the other hand i don't know shit from shinola, so feel free to ignore me.

Talking about changes since 2012, don't know why the 900 sec posting timer is still a thing, election's over and so the number of psy-ops/shitposts should be decreasing, i miss the unique ip counter more than anything tho.
>>
>>1002321
ah i see what you mean, sorry. I'll try to tweak it a bit more

this place is just a fed site now anyways. all that matters is controlling /pol/ the other boards are mostly meaningless. /3/ and a lot of the slower boards should have been skipped for the 900 second thing but here we are. i wish moot had the foresight to spin off pol into its own website. so the rest of us wouldn't have to be affected by their retardation.
>>
File: facewip5.jpg (317 KB, 2502x1459)
317 KB
317 KB JPG
I'm working on this face rig. Please critique it so I can make it better. I've been looking at it for too long I need to get a reality check from somebody else to understand the problems with it.
>>
File: 1.gif (2.53 MB, 540x540)
2.53 MB
2.53 MB GIF
rachida wip need to texture
>>
>>1002348
I hate ugly furry stuff like this
>>
File: 8fx0xq99tu7d1.png (143 KB, 753x594)
143 KB
143 KB PNG
>>1002348
peaklog mentioned
>>
File: feth.png (316 KB, 916x763)
316 KB
316 KB PNG
>>1002323
i regret every decision that has led me to this point. i was gonna do several faces and set up the rig so i can animate them but idk at this point. maybe just 1-2 frames for the eyes to make blinks i need to wear a shock collar that jolts me every time I attempt to draw. either way, finna rig her soon
>>
File: AllLiesShouldDie.jpg (201 KB, 1821x1295)
201 KB
201 KB JPG
>>998180
i'm learning to texture low poly low res retro graphics, made the jacket but i think my colors suck.
The style is megaman legends inspired.
>>
one of my tweet is getting attention, around 400 likes, it's the first time that's ever happened
it's a weird feeling that people might like what I made
>>
>>1002348
Why not post it in the /snoot/ thread in trash?
>>
>>1002418
i started getting traction too, it was good for a while, but then i got shadow b&. Just saying, you could feel like you made it one day and have a rugpull the next, just because of awful moderation
>>
File: MOGGED.png (5 KB, 402x121)
5 KB
5 KB PNG
>>1002418
MOGGED

jokes aside, the algorithm is a fickle mistress. most social media will try to breadcrumb you with engagement to trick you into a dopamine cycle and get you hooked. numbers are meaningless, everyone is a bot, etc etc.
>>
>>1002460
honestly, those are bad numbers. But, honestly, after getting shadow b& and having my audience for my content be myself only, I realized that we are facing an impossible uphill climb
>>
>>1002468
better to be yourself than be a slave to the algorithm and the jews that run it. lest you end up like mr beast, or for a more /3/ related example, smeaf. eternally condemned to appeal to the lowest common denominator and produce slop tutorials on how to delete blender's default cube.
>>
>>1002469
its just all so tiresome. I feel like we all got rugpulled with this ai shit and they have stuff to make what we are doing better, not just in cg but in regular jobs, but they wont let us have it and we're all doing the same old shit
>>
>>1002471
the niggas in power hate creativity with a passion. if they could, no one would ever create ever again, just consume. literally just make anything, no matter how good or bad it is. don't even have to post it. if you get demoralized they win.
>>
>>1002348
Ugly furry design aside. Make the light brown portion cover more of her inner eye.
Pull fangs back a smidge.
Cel shading. Which will really help the hair.
Highlights for hair too.
>>
>>1002447
>then i got shadow b&
Post proof or you're coping.
>>
>>1002486
Here's the proof : when I post to the sub it doesnt show up in "new" or "hot", just in my profile. Nobody can see my topic except for me. When I message the mods about this they don't reply. I have been 100% shadow b&.

I'm done making content for subreddits, they will shadow b& you at a moments notice, despite your posting history. Same with all other social sites. Its a complete scam.
>>
File: clack clack pow.png (168 KB, 995x556)
168 KB
168 KB PNG
>>1001515
>>1001579
>>1001601
Making more guns, this time a pump action shottie
>>
File: wp.png (341 KB, 1110x843)
341 KB
341 KB PNG
>>1002389
>>
>>1002503
It seems like reddit will permanently (shadow) ban you for creating a topic vaguely similar to another topic you created in a dedicated topic oriented sub (and not tell you before hand that they will do this) while youtube encourages you to create hundreds if not thousands of similar content for your channel. Because they don't tell you this and they effectively ban you permanently with a shadow ban that will never be lifted is why people get pissed off.
>>
File: file.png (162 KB, 1098x1268)
162 KB
162 KB PNG
I followed a basic picture of a dude doing an A pose but I still feel like the proportions are off. Can someone give me a quick criticism?
>>
File: SDF fur 01.webm (2.06 MB, 850x486)
2.06 MB
2.06 MB WEBM
>>1000700
Been doing a bunch of noding. I'll skip the long winded explanations this time. Basically, I've been concocting various other ways to create fur. I want a more procedural method. So I've doubled down on SDF experiments. Trying to create SDFs of my own from scratch. It's a struggle.

In the webm, I've extruded the mesh, and then put an SDF tube inside of each face. This experiment gets shitty results. But I'm just glad that I was able to make the tubes at all. Took a lot of tinkering and problem solving. The problem is the boundaries of the face limit the visibility of its respective tube. So when you look at a face head-on, you can see the whole tube fine. But looking at a face from an angle, obscures the tube.

This is still pretty exciting. Because I can place a procedural object in space.

>>1002858
Give him a neck.
>>
File: Armor1.png (1.73 MB, 624x965)
1.73 MB
1.73 MB PNG
This is taking forever
>>
File: qt.png (144 KB, 683x984)
144 KB
144 KB PNG
im making a cute brown girl!
>>
File: Black Girl 01.png (943 KB, 1214x853)
943 KB
943 KB PNG
>>1002996
Same!
Well not really. I just wanted to work on face topology, and decided to make her black.
I don't follow the conventional "mask" shape around the eyes. Because I think it doesn't look good at the corners of the eyes. So I formed it a little different there.
And the mysteries of the nostrils still elude me. But I think this nostril topology is an improvement over the previous.
>>
File: 1728691362259797.png (2.41 MB, 1520x858)
2.41 MB
2.41 MB PNG
>>998180
wip
>>
>>1002858
Biceps too long. Forearms too short. Hands might be a bit smaller than they should be but that might just be me
>>
File: qt2.png (136 KB, 672x874)
136 KB
136 KB PNG
>>1003007
omg shes so cute wish i was that good. what are you gonna do for her hair?
im trying to expand from low poly shit and make something actually semi-realistic but man it's hard.
>>
File: tradgirl.png (214 KB, 543x794)
214 KB
214 KB PNG
>>
File: 1732883128244345.jpg (73 KB, 1024x1024)
73 KB
73 KB JPG
>>1003166
You know...
>>
File: Black Girl 02.png (1.19 MB, 1419x881)
1.19 MB
1.19 MB PNG
>>1003124
I'm probably not going to do hair. Since it's too hard for me at the moment. But if I were, I would rely on Blender's hair curve system. I'm starting to get this crazy idea that hair can be done using SDFs. So I'm probably going to waste the next year of my life trying to figure out how SDFs work. Only to discover that it's too complex for a full hair system.

For now I'm just trying to get really solid topology. Working on the ear now. Trying to get clean topology that accounts for all the little twists and turns of the ear. It's looking like an ear now, but I think the topology can still be optimized.

My rule of thumb, is that each "side" has to be at least 2 faces. So if I wanted to make a cylinder shape, it would be 8 faces around. It can go higher as necessary, but only by intervals of 2. 8, 10, 12, etc. That way, the mesh can be un-subdivided cleanly. So conceptually, my mesh can be optimized down to 1/4th of the faces showing in the face counter. I use the 2 face per side minimum, because it's acceptably smooth to my eyes.
>>
File: Black Girl 03.mp4 (1.61 MB, 464x666)
1.61 MB
1.61 MB MP4
>>1003213
Wow, the node bug bit me today, and suddenly I was able to create procedural eyebrows. All you have to do is draw a curve on the surface, and it does the rest. There are a few parameters to play with. Hair density, hair thickness, scaling the inner and outer parts of the brow.
The way the hairs lay against the surface can even be adjusted.
But unfortunately, the rotation equation is a limited right now. I'll have to figure out how to reliably adjust the angle of the hairs depending on where they are along the curve.

Each hair is flat instead of round. I figured since they're so small, that it would be better to make them flat, because then I can bend the shape more smoothly. Normally, it would cost 10 faces to get a single bend. Which looks all jagged, even at a distance. More bends adds 4 more faces. And with so many hairs, the face count increases rapidly. With them being flat, I can spend 8 faces making them appear smooth.

At last, I clip-masked the tips of each hair, so the tips are rounded. The rounding done in shader nodes, so they're perfectly calculated half circles.
>>
File: Black Girl 04.png (1.51 MB, 1360x849)
1.51 MB
1.51 MB PNG
>>1003234
I've been on a blender binge. Trying out various procedural texturing methods.
Copied the brow node group and tweaked it to make lashes.

Figured out how to quickly get the boundaries of materials. Using the shortest path node, face group boundaries, and material index. I was able to isolate the edges that connect all of the materials, and then weight the rest of the mesh toward the boundary edges. This creates a clean gradient across the mesh, where the material boundaries are zero. From there, you store the results and take it into geometry nodes, and fiddle with the color using the mask range node, until you've isolated the transition between materials. That's how I was able to blend the lips with the rest of the skin.

Using a similar shortest path method, you can isolate any region you want creating a vertex group that selects the points you want to transition between. That's what I did to mask around the eyes and the upper lip. Blending in darker skin pigmentation there.

I learned that the white noise texture node is good for blurring. That helped soften the lines of her lips. I want to avoid UVs at all cost, so I'm trying to think of ways to create lines and creases without touching UV stuff. I'm not entirely sure how to do that. At least, not easily. Maybe there is a way to assign directions to each face, and then map the texture to that direction. Dunno.

Also they eyes are procedural as well. I improved it slightly by giving it the outer glassy layer. Complete with bump where the iris sits. It's not much yet.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.