[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/3/ - 3DCG

View post   

File: 2.71 MB, 1500x882, wip 1.png [View same] [iqdb] [saucenao] [google]
980444 No.980444 [Reply] [Original]

/wip/ - Works In Progress - "We're going back to our roots" Edition

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

Previous thread: >>977313

List of free resources: https://pastebin.com/cZLVnNtB (embed) (embed)
/3/ Discord: https://discord.gg/gbYCEBPuK2

>> No.980446

(embed) (embed) (embed)(embed) (embed) (embed)(embed) (embed) (embed)(embed) (embed) (embed)

>> No.980448
File: 358 KB, 1476x868, nano chan.png [View same] [iqdb] [saucenao] [google]

just started. nichijou has those distinctive face shapes, idk how im gonna do it honestly. I think I'll see if I can hide how weird it looks at certain angles with the shader when i get there

>> No.980449
File: 3.77 MB, 2060x1025, Blender 4 1.png [View same] [iqdb] [saucenao] [google]

This took months.

>> No.980451

Scrap it and start over

>> No.980452

meant for >>980449

>> No.980465
File: 119 KB, 918x1135, hammer_06.jpg [View same] [iqdb] [saucenao] [google]

I finished up the low poly space dwarf hammer I posted in the last thread. At least as far as I'm gonna take it.

>> No.980484

What you need to keep in mind is that you're not trying to faithfully replicate a real 3D shape. You want a shape with a faithful silhouette from the most important angles. That might end up being a weird looking thing with the default shader attached to it, but will look right with a shader that flattens it or hides surface details somehow.

>> No.980491

Looks nice. How did you do the textures? Asking for a beg.

>> No.980492

It's really just default stuff from Substance Painter. The normals/AO/etc. were baked (in substance) from a high poly of course. So mostly just dragged and dropped on, and separated a little bit by the baked ID map.

>> No.980493
File: 473 KB, 1603x896, Unity_2024-04-13_22-26-10.jpg [View same] [iqdb] [saucenao] [google]

Saddly this is mostly downloaded assets, but I'm working on reconciling frame rate and appearance. I think I took a big step backwards because the very distant trees are bright when they should be the darkest trees,

>> No.980495

>sadly this is mostly downloaded assets
I mean if it wasn't downloaded even in production it would just be speedtree'd, so I wouldn't feel too bad about it.
What are you doing to improve performance?

>> No.980496
File: 417 KB, 1602x891, Unity_2024-04-13_22-34-06.jpg [View same] [iqdb] [saucenao] [google]

After extending the distance for the LOD models on the trees it looks better, and in both scenarios the FR is ineffective, meaning low. So maybe not an worth it optimization

>> No.980497

I am lowering the LOD and the culling distance. And seeing what increases frame rate. Unfortunately nothing's really been working except lowering the tree count till the forest isn't as dense as I really want it to be. It's already much sparse than I envisioned.

>> No.980499

My WIP for the night is I started this anime girl. I'm gonna start refining the topology a bit tomorrow. I've never managed to do a good job with the butt and groin before but I'll give it a shot tomorrow, along with refining the arms and shoulders, I didn't have much arm reference on hand so they're a bit janky.

>> No.980500
File: 140 KB, 819x972, anime_01.jpg [View same] [iqdb] [saucenao] [google]

Aaaaand, I forgot the picture.

>> No.980510
File: 772 KB, 1590x921, WIP_7.jpg [View same] [iqdb] [saucenao] [google]

I kinda feel some blender more and more Blender modifiers are getting obsolete with geometry nodes

>> No.980513
File: 140 KB, 1920x1080, Messenger_creation_88b4f5d3-ac1f-4b33-b1d3-c6abf26af4be.jpg [View same] [iqdb] [saucenao] [google]

More of this particular piece

>> No.980516
File: 179 KB, 635x898, longstudy.png [View same] [iqdb] [saucenao] [google]

Still a long way to go on this model, but happy to finally get some clothes made, even if it's just a simple shirt.

>> No.980534
File: 98 KB, 271x401, buttTopo.png [View same] [iqdb] [saucenao] [google]

I find having topo on buttocks like this pic to help more with giving them a proper deformation.

>> No.980535
File: 337 KB, 1603x898, Unity_2024-04-14_10-05-16.jpg [View same] [iqdb] [saucenao] [google]

I made some changes you wouldn't think would help with optimization, but in the end this scene maintained 15fps, when the previous scenes were at 9.
First I added a row of trees to the horizon of the skybox. So I didnt need as many trees to block out the light. Also I was previously using the 3 smallest tree model in my pack and now I am using the 3 largest. Bigger trunks, canopies, more shadow coverage and distant vision coverage with fewer trees. Although this is not fewer trees, it's the same rate of tree generation. The previous skybox also gave me a lot of trouble because it made every reflective surface appear to be glowing while in the shadows of the forest. So I disabled reflections on all the materials. Just now thought I turned it back on with the new skybox material and they all looked way more appropriate. The last thing is that there were major bald spots in my forest because trees were not set to spawn in certain areas for certain reasons. I want continuous forest, by default, and then clearings will be added purposefully for points of interest.
The last problem then is that distant trees are not receiving shadows until you get closer. One of the ways I want players to find these clearings is by seeing them from a distance. If the distant trees always look bright then you can't use it for hints.

>> No.980545
File: 1.80 MB, 1450x814, coffee.png [View same] [iqdb] [saucenao] [google]

Making some coffee-related props

>> No.980546
File: 546 KB, 649x838, coffee machine.png [View same] [iqdb] [saucenao] [google]

Really proud of this coffee machine, I designed it myself

>> No.980547


>> No.980555
File: 113 KB, 712x911, posing.png [View same] [iqdb] [saucenao] [google]

Hospital gown made and weight painting cleanup mostly done.
Now I can make clothes that turn on or off based on layer visibility, and the clothes inherit the weight painting from a base mesh.
Only 567 tris, right now, but that'll go up when the mouth and eye rigs are done, too.

>> No.980557

Thanks bruv

>> No.980591
File: 695 KB, 1920x1080, file.png [View same] [iqdb] [saucenao] [google]


>> No.980606
File: 154 KB, 1150x894, anime_02.jpg [View same] [iqdb] [saucenao] [google]

I only did a little bit of work on the hips, but I did improve the chest, tummy, shoulders and little bit of arms. The topology's a lot less consistent right now, but I'm trying to get the main forms in and then I'll retopologize after knowing what the minimum poly density needs to be. It's coming along though. Arms are complicated and I hate them.
I did move it a little bit in this direction, still needs more geo to look right I think but it's better despite some un-terminated edges at the moment.

>> No.980608

Do they fuck?

>> No.980609

Yeah big.improvement in butt topo. You want thise edge loops to open like an accordion when she bend her legs.

>> No.980613

I think hard edges for shadow would be better than gradients but up to you. Also really need some fold shadows.

>> No.980615

Oh, for sure. The vertex colors are just there for reference. I intend on getting proper textures, and then play around with either vertex shading or just shading the texture. Haven't decided yet.

>> No.980624

Nice, like the model otherwise!

>> No.980633


>> No.980639

Why not?

>> No.980643
File: 56 KB, 927x665, Screenshot 2024-04-15 032736.png [View same] [iqdb] [saucenao] [google]

I make the pew pew

>> No.980646

It's good.

>> No.980654

Hm. Good point. I'll get back to you.

>> No.980655
File: 615 KB, 922x675, 1703230122987448.png [View same] [iqdb] [saucenao] [google]

Is there a better way to do low twin braids or are they fine as they are for later cel-esque/simplistic rendering?

>> No.980657
File: 106 KB, 843x232, Screenshot_20240415_174415.png [View same] [iqdb] [saucenao] [google]

Just make a proper mesh for them. Why would you want to create something that looks like shit just to save a few dozen vertices?

>> No.980658

I used this video as a tutorial: https://www.youtube.com/watch?v=FB47bNPRQC0&t=3s
I think my bigger issue is how to attach them to the back hair, and more importantly if It's better to model this type of back hair directly as a mesh instead of some path bullshit (https://www.youtube.com/watch?v=Ng8FbOHvsLQ)) that's probably not working quite as well as it should.

>> No.980665

Nah the issue is that hair appears to flow from a common source, like water. So normally its ok to have these gradual transitions between hair chunks, but on a braid there are distinct and seperate lengths of hair which are woven around eachnother. That means sharp prominent transitions between them. You braids are too lumpy and smooth.

>> No.980666
File: 166 KB, 296x747, 1712541882349074.png [View same] [iqdb] [saucenao] [google]

Is it possible to unfuck it messing around with creases and the subdivision modifier? It looks good otherwise in low poly, but starts to look like mush subdivided and shaded smooth. I also prefer this style of larger-segmented, less ropey-looking braid better for this character but I do get what this image >>980657 means in making the component lengths of hair look more distinct.

>> No.980667

Having seperate geometry for each hair strand is the solution. Even if you subdiv the crease between seperate geometry is preserved no matter what.
Sire theres going to be some solution where you carefully apply special creases and normal, but that will take longer to do than starting over and jusy extruding new tubes.

>> No.980668

You would just have to poly model the hair segments to smooth appropriately with the modifier. I would recommend against doing this with creases, but if it gets you results then whatever. You only have 1 segment repeated, so doing it with proper topology once shouldn't be that big of a lift, though might take some time to figure out.

>> No.980676
File: 331 KB, 593x711, 1708440758952643.png [View same] [iqdb] [saucenao] [google]

Alright, this looks much better. Thanks

>> No.980708

Looks good anon.

>> No.980733
File: 202 KB, 1283x1138, anime_03.jpg [View same] [iqdb] [saucenao] [google]

Some more incremental improvements
>improved the butt a little more
>slightly better shoulder geo and actually integrated the arms properly now feel like they might be a tad thin)
>Added a belly button because tummy
On the to-do list:
>Add 2, maybe 3 more edge loops to the arms, see if I can block those shapes out better
>add some blocked in hands and improve blocked in feet
>start to work on doubling the density of the mesh to get the rest of the main details I want and refine any curves
>as that happens, improve elbow, knee, and back topology

There is one triangle on the side of the thigh that is driving me crazy, but I just can't figure out how to terminate it without adding in too much extra geometry. I should be able to do it when I don't have the bare minimum polygons though.

>> No.980765

>There is one triangle on the side of the thigh that is driving me crazy,
It's not that big of a deal. It's often better to have triangles than poles or high variance in density anyway.

>> No.980815
File: 800 KB, 1236x886, Tummy and Tush.png [View same] [iqdb] [saucenao] [google]

This is probably my 50th time retopologizing the butt, and it may very well be my final time, as I think I've cracked the code. Well first, it started with me retopologizing the stomach and hip area. I finally tried something I saw in other people's work. Where all the edges sort of converge on the belly button in this radial pattern. That allows for some good edgeflow where the thighs meet the groin. I tried something like it before, but it didn't look quite right. This time, I fully committed to the concept and got it working. Now the hips and stomach are nice and shapely and deform really well.

From there, I wasn't sure how I was going to handle the butt. Moving the hips as I did made a lot of diagonal edges. Well I figured it out. It's a similar concept to the belly button technique. Setting up this radial cleft in the crack of the button, And then looping outward makes for this really clean edgeflow. that contours to the creases of the butt.

The cool thing about this "radial" topology, is that the n-poles are tucked away out of sight. In the butt's case, the n-poles are deep in between the crack. And for the stomach and hips, the n-poles are deep inside the belly button. With the n-poles completely hidden, it ensures the shading will remain very smooth over the entire surface without interruption.

>> No.980822

Idk bro. You have horizontal edges that transition into the vertical loops of the leg. I dont think it will deform right.

>> No.980833

With 2-3 posts a day is this thread going to be entirely people trying to figure out ass topology for the next 50 days?
/3/ needs a monthly challenge model or something like polycount has/had.

>> No.980836
File: 437 KB, 640x360, Butt Deform.webm [View same] [iqdb] [saucenao] [google]

It will work. Here. My armature is still trash. But I threw this animation to show it deforming.
Don't think of it as horizontal lines turning into vertical lines. It's difficult to explain it. But it's more like the back and the butt are the same plane. And then the leg is sort of "under" it. Or "in front" of it. So to speak. Like the butt is kind of a shelf.

It's similar to the way they teach people how to construct shoulders. The top of the shoulders wrap with itself. And then the arm is sort up "under" the shoulder shelf. Similar construction.

>> No.980837

That would be cool. Challenges and competitions and stuff are one of the things I really miss about old forums.

>> No.980839

That Bender thread was pretty succesful.

>> No.980840

It's not that it looks bad necessarily and if that's the way you like it then, it's ok. However it does not resemble anything human.

>> No.980843

I feel like this comment is just meant to antagonize me. What am I supposed to make of it? What exactly isn't human looking about it? I can probably guess. But you're so vague, I might guess wrong.

>> No.980845
File: 165 KB, 1280x1158, 3-modelling-challenge-apr15.jpg [View same] [iqdb] [saucenao] [google]

Okay folks. Here's the challenge this week.
Model this flash hider (no texturing required, just make a high poly model of it).
Just for practice to break up the monotony of whatever you're normally working on.
Encourage people asking for help as long as they also post what they've already tried themselves.

>> No.980846

I doesn't look bad. The problem is that there no pelvic bone. But then again who says that there has to be one. It's a fantasy creation.

>> No.980848
File: 57 KB, 800x600, 0001.jpg [View same] [iqdb] [saucenao] [google]

Where do I collect my prize?

>> No.980849

I see. But also I don't see. Because I did give her a pelvic bone. You can even see the little bump of it in the webm.

>> No.980850

Not bad. The front should have a bit more of a curve to it though.
I forgot to mention that the prize is a virtual high-five, and getting to think you're better than all the people who didn't participate.

>> No.980851
File: 908 KB, 1653x1255, fortunate son intensifies.jpg [View same] [iqdb] [saucenao] [google]


>> No.980852

Maybe there's too much Corrective Smooth. To me it's like it's pretending too much to be an ass. But if you can finish a model and keep that same style consistent everywhere, then it could be fine.

>> No.980853

I wonder how many people think this is funny.
I do, I'm just thinking it's gotta be a pretty small number of people.

>> No.980854

I've kind of cheated because it's an SDF and I've done it in 30 seconds. You probably were expecting a subdivision model.

>> No.980856
File: 57 KB, 800x600, 0001.jpg [View same] [iqdb] [saucenao] [google]

I've noticed that just after I pressed the Render button. I should have changed the shape from cylinder to capsule. I'm not going to re-do it but I'll give you this instead.

>> No.980863

It's a real "you had to be there" type thing. Granted a lot of people were in the thread, but how many of them are still around, that's the question.

Still (and this goes for you >>980845), artificial, "challenge threads" really never take off. Organic ones definitely do. People always want to show someone up when it comes to the more organic "challenges". Like one anon can't model something seemingly simple and asks for help, and anons pour out of the woodwork and take over the thread to get super autistic about how to make this one thing "correctly".
Meanwhile a genuine challenge thread shows up and no one gives a fuck because their time would better be spent elsewhere than modelling some shitty household object some /beg/ anon came up with.

Point is, people don't give a shit about internet high-fives or "pRaCtiCe", they DO give a shit about being right.

>> No.980864

Well obviously it's about being right. Right about my topology being better than everybody elses'

>> No.980865
File: 439 KB, 927x806, Tummy and Tush 2.png [View same] [iqdb] [saucenao] [google]

I don't remember if corrective smooth as on for that webm. I turned it off and on about a dozen times before finally rendering. But honestly, corrective smooth isn't make a huge difference here. the deformation is primarily done by the armature, and preserve volume is off. corrective smooth just smooths a little like advertised.
Anyway, the deformation isn't to my liking yet. That's why I said my armature is trash. The butt isn't butting how I want it to butt. But I figured even a crusty demonstration of deformation should get the idea across that the edges are going to deform well. Oh, and I guess it should be noted that I didn't give it any shapekeys. So the but is just moving by some helper bones.

It's true I'm not going for 100% realism. But I still want a butt that looks nice. I guess I'll work on improving deformation. Which I was planning on doing anyway.

>> No.980866
File: 375 KB, 1591x2027, week1.jpg [View same] [iqdb] [saucenao] [google]

It's not perfect but it's pretty close. Unfortunately I don't know how to make a decent looking clay render in Blender. I'll find out another time.

>> No.980867

Those are very difficult to photograph indoors even in the real world anon, they hide the flash.

>> No.980868
File: 202 KB, 639x1021, a jokejpg.jpg [View same] [iqdb] [saucenao] [google]


>> No.980871
File: 48 KB, 637x715, side_bent.jpg [View same] [iqdb] [saucenao] [google]

The border of the costume is where the pelvic bone is nearest to the skin. The area under the influence of the pelvic bone doesn't deform when the legs move. But your model is your interpretation and one detail in isolation is always going to appear weak.

>> No.980872
File: 27 KB, 436x547, side_bent_2.jpg [View same] [iqdb] [saucenao] [google]

The costume rests on the pelvic bone which is part of the spine and it's rigid because it's a container for whatever nonsense it's in there. For more details you could study anatomy but in my personal experience this alone is enough in order to make something credible.

>> No.980876

>Make a model
>Needs a little background to it I think
>Wait this might be better as a little animation in a scene
>Wait it would be even better if I made it a longer animation inside of a bigger scene with lots of cool camera cuts
>Lunge into it headfirst
>Never made an animation longer than about 3 seconds with one camera angle

Bros am I biting off more than I can chew or is this how you get better? I'm going to make it (kinda) specifically for something but it doesn't need to be done for another 5 or 6 months. To anyone else 5 or 6 months is a long time but I'm a boomer and got into 3D late so I can never really git gud like you young bucks with your whole lives ahead of you.

>> No.980880

Animation in the sense of just a static model, or animation like rigging and animating a character?
Either way yeah probably, obviously the latter will take more time but you could definitely still do it to some decent degree in 5 or 6 months.

>> No.980881

It's a combination of the two. There's only one character and all he does is walk across the room and sit down. I just downloaded a rigged model instead of making my own, but I'm too retarded to even know how to adjust that so I'm making something around his animation cycle instead.
There's only one other really animated part, but it's so simple even I can do it because it doesn't need rigged.
Other than that it's only things like closeup pans of objects and some camera movement. I've given it a maximum length of 30 seconds, but really I think it'll be closer to 15-20 and even that is probably pushing it.
It's almost done with it just being one small room. I'm at the stage of throwing some props in there so the room doesn't seem as empty and then working out the lighting and how to use multiple cameras.

>> No.980883

>It's a real "you had to be there" type thing. Granted a lot of people were in the thread, but how many of them are still around, that's the question.
I wasn't in the thread either, THAT SHAPE is just kind of a hardsurface meme for anybody who's learned subdivision modelling.

>> No.980891
File: 2.38 MB, 1920x1080, gvhvgvhghm.png [View same] [iqdb] [saucenao] [google]

i photo scanned a ceramic slab i made

>> No.980892
File: 538 KB, 740x733, Screenshot_20240417_214533.png [View same] [iqdb] [saucenao] [google]

Curved circular holes are kind of tricky.

>> No.980893
File: 341 KB, 2448x3264, IMG_20240416_000528.jpg [View same] [iqdb] [saucenao] [google]

About to finish
I'm going to upload better screensht later

>> No.980894

>'curved circular holes are kind of tricky'
>Does them perfectly without TOO much extra geometry

Some crits though:
>I'd like it more if you managed to integrate the screw geo with the model, but terminating screw ends is a huge pain so I get it.
>The cylinder uses more geometry than is necessary
>the base of the cutout on the prongs is more curved in the reference than the sharp 90d angle you have there
>You need to add the supporting loops for the little chamfer on the prongs so so that it's sharp, I can see you added it but it got smoothed out
>The high poly mesh and some awkward terminations at the rounded edge look like they could be cleaner and easier to deal with, I can't tell if that artifact on the far right tip is caused by the smoothing or just an HDR lighting setup

The result does look really good though.
Are you using subdivisions or are you modelling it in high poly and beveling edges?

>> No.980898
File: 187 KB, 640x360, Tummy and Tush leotard line.png [View same] [iqdb] [saucenao] [google]

I feel like I'm talking to the riddler over here. You're going to have to be more direct. Redline me, bro.

>> No.980907

Follow the yellow lines throughout your animation here >>980836 You'll see that they change shape quite a lot, but in humans they can't because there's the rim of the pelvic bone right underneath.

>> No.980915
File: 107 KB, 929x706, imagine.jpg [View same] [iqdb] [saucenao] [google]

Imagine sperging out about a shape you can make with 42 active polygons and some symmetry modifiers.

>> No.980923

I'd say you're the one sperging out. Considering I was making a joke about the thread where most of the anons couldn't make a "seemingly simple shape".
Instead of taking a joke in stride, it triggered your autism.

>> No.980930
File: 1.15 MB, 1500x1500, Intrusive advertising.jpg [View same] [iqdb] [saucenao] [google]

Got some ads on the buildings. They're all placeholders until the dude sends me his promo stuff.
I've got a global slider that controls the probability a building has ads on it or not, right now it's turned up pretty high, but it'll be significantly lower in the final.
I've got some weird issues with the UVs on some of them (due to the UVs being mirrored), that I'll have to fix next time. Also I'm not too crazy about the layouts on some of them.

>> No.980931
File: 71 KB, 561x1016, anime_04.jpg [View same] [iqdb] [saucenao] [google]

This isn't a for real update but I started modelling the hands and was just kind of eyeballing it. Didn't really block out any size before starting to extrude the wrist. In my head I was just modelling them off the size of the wrist without thinking it was a super skinny model, and the arms might even be too thin.
Anyway the point is that I ended up giving her horrifying baby hands and thought it was funny.

>> No.980932

Fuck, now I want a Giga Burger.
Reminds me of the duke nukem 3d skyline at the very beginning of level 1. I'm sure yours looks better though.

>> No.980934
File: 2.61 MB, 1920x1080, test render.png [View same] [iqdb] [saucenao] [google]

does it look drawn/painted by hand?

>> No.980940

No. It looks pixelated like a video game without anti-aliasing.

>> No.980944
File: 128 KB, 500x370, damn those alien bastards.jpg [View same] [iqdb] [saucenao] [google]

>now I want a Giga Burger.
The ads are working.

>duke nukem 3d skyline
Yeah I can kinda see it. I think my city needs a bit more work to bring them together and add a bit more cohesion, and a bit more context in an actual scene rather than a bunch of buildings, but it's coming along. It'll be nice to have some flying cars and shit flying through it.

>> No.980947

It does not look hand painted, it looks kinda pixelated as previously mentioned. I like it though. Feels like Myst a little bit.

>> No.980953
File: 2.05 MB, 3000x1924, 1962.1_il-bucintoro.jpg [View same] [iqdb] [saucenao] [google]


>> No.980985
File: 473 KB, 402x287, multiJointButt.gif [View same] [iqdb] [saucenao] [google]

So look how all my horizontal edge loops stay horizontal and all my vertical loops stay vertical. This lets the butt meat open up and close like an accordian. Your butt is very small in comparison and the loss of volume isnt as big of a deal, but here I have an extra joint in the same location as the femur joint which is driving the motion of the verts on the parts of the butt between the pelvis/tailbone and the upper thigh.
So this riddle about the "pelvic bone". You have 2 joints here fighting for control over all these verts. you hip joint and the femur joint. There are parts of the body where skin and tendons are thoroughly affixed onto the bones underneath them. The part of the pelvis most relevant is the "ilium" and the "coccyx" (google em). Imagine (or model) an actual pelvis under your mesh. The verts contacting these parts of the bone should have a 100% envelopment by the hip joint. So they do not move at all when the femur moves. All that tissue is then tied into the big bump on the outside of the femur called the "great trochanter". When you move your leg around all those muscles and fat get pulled between the pelvis and the femur.
For the most part your rig is already doing this fairly well and unless the camera is going to be focussing on her ass cheecks while she does gymnastic moves it won't be an issue. Those 100% pelvis verts are important up around the hips to keep her from looking like she's made of rubber.

>> No.980988

Literally /atg/ - ass topology general.

>> No.980989

That gif is from a year ago and it looks as grotesque today as it did a year ago.

>> No.980991

So what? The topology is correct, and will continue to be correct for the rest of time.

>> No.980992

I'm sorry to see that one year later you're still stuck at the same level.

>> No.980993

Can't improve on perfection BAY BEEE

>> No.980996

lol dude that model is actually 4 years old. I just havent made another gif. I have posted new models of asses in a WIP thread recently.

>> No.980998

Nah man, not yet

>> No.981010
File: 136 KB, 1397x980, week_01.jpg [View same] [iqdb] [saucenao] [google]

I think this is a lot better of a render and my final entry.
There is a little bit more smoothing problems between the circular holes than I wanted, I should have upped the cylinder segments to have like 1 more per circular cut-out, but this is still fine imo.

>> No.981011
File: 2.60 MB, 720x370, big brain.webm [View same] [iqdb] [saucenao] [google]

>*Oppenheimer soundtrack plays in the background*

>> No.981012
File: 2.09 MB, 1920x1080, adsfffddd.png [View same] [iqdb] [saucenao] [google]

fun times

>> No.981017
File: 846 KB, 1920x1044, Screenshot_466.png [View same] [iqdb] [saucenao] [google]

finished modelling the arches in the doge's palace. GOD I HATE MODELLING ARCHES

>> No.981018

Why? Does the curve never look right?

>> No.981020

yeah and there are just way too many steps needed to create arches, especially when they have garnishments.

also having to incorporate the modelled arches into the main building is also a pain in the ass

>> No.981021
File: 317 KB, 400x877, Tummy and Tush 3.png [View same] [iqdb] [saucenao] [google]

I do *kinda* get what you guys mean. But I really wasn't expecting the deformation to look good at the time. My armature is still trash like I said. So getting told something doesn't look right on my trash armature is just a bother.

But regardless, your bothering is motivating me to work harder. I think the main problem here, is that I've been trying to find a bone set up that satisfies automatic weight painting. But automatic weight painting is good for very basic things, not good at dealing with finer details. For that, I'll have to go in and handle weights manually. Which I've been avoiding for I think 2 years now, because it's a paint in the ass.

Anyway, I began the tedious journey of manual weight assignment yesterday. Selecting specific points, and giving them specific weights. I was making some good progress around the hip/butt/leg zone, when suddenly, my vertex groups got scrambled. It's like all of the groups swapped traded names with each other. Which effectively reassigned all the bones to different weights. This is an absolute nightmare. I googled the problem, and the responses I'm reading just say "yeah, that happens..." Typical for looking up a problem in blender. I think I'll have to manually rename all the vertex groups to their proper bone name.

Well that issue aside, I still think my topology is better. I could easily recreate your lines. But I think that method makes the vertex density of the butt too sparse. Since the bulbous shape requires more vertices to cover the whole surface. The "radial" method I'm using creates the right amount of vertex coverage, without appearing stretched.I can also plump it up to much larger levels without too much stretching. You say my butt is very small. But it's really not. The bad deformation just isn't allowing it to move properly. It's getting squashed from poor weight paints. I plan to remedy that, as soon as I clean up my vertex groups.

>> No.981025
File: 1.44 MB, 1920x1080, doge-min.png [View same] [iqdb] [saucenao] [google]

thanks for this new photo bc of that ive now started to model the doge's palace

>> No.981026
File: 28 KB, 427x545, spine_wp.jpg [View same] [iqdb] [saucenao] [google]

You're making things up 100x as complicated as they actually are. Pic related is what the base of the spine weight paint looks like. Incidentally, it roughly looks like the costume in the reference images >>980871 >>980872 which incidentally roughly looks like the shape of the pelvic bone, which incidentally is also close as to what the Automatic Weights paint function does anyway.
You may be disappointed that there is not as much creativity involved as you may like, but that's what it is.

>> No.981027

That's what I'm trying to do now, more or less. Just a cleaner version of that. Which was working relatively well, until all the vertex group names got scrambled. In fact, they just scrambled again not an hour ago, and I'm about to pull my hair out. I *need* to figure out what the problem is, or else it's going to make this process impossible.

It might be caused by the fact that I'm editing the mesh while also doing weights.(I wanted to clean up the hand region) But I don't know why editing the mesh would cause the NAMES to change. It's like every name gets offset.
It might be caused by the fact that I'm deleting a lot of the helper bones I made for automatic weight paint manipulation. And somehow the missing bones are causing the vertex nodes to freak out? idk.
It might be something to do with the search function in the vertex group panel. I was searching out groups by name, and then deleting the excess. As I'm deleting, the next group on the list is highlighted. But when it comes to the end of the searched list, the highlight disappears. However, despite the cursor not being visible, I think it's still capable of deleting groups that aren't on the list. And somehow deleting unsearched groups causes the names to swap. But again, I don't really know if this is the case.

None of these sound like practical reasons for all the groups to trade vertices or names. Yet, somehow, someway, blender is doing it.

>> No.981030

I think I found the problem. I'm pretty sure it's the search. I just did a test in a new document and reproduced the effect. If you're in edit mode, and then perform a search, and then delete a vertex group that isn't in the current search, and then undo while still in edit mode, blender freaks out and preserves the deleted group by offsetting it into another existing group. Which in turn offsets that, which offsets the next, and so on, until multiple groups are now wrong, and one of them is deleted.

Note to self and everyone: Do not undo vertex groups in edit mode. I should have guessed edit mode was the culprit, because undoing in edit mode has ALWAYS caused problems. Never rely on edit mode for undos. Never ever.

>> No.981032

What's to hate? Create one half of a pillar + arch, mirror it, add array modifier, and you're done.

It's getting better.

>> No.981034
File: 1.02 MB, 640x480, 0001-0800.webm [View same] [iqdb] [saucenao] [google]

A little animation which I'm probably not going to finish because the mocap data is noisy, the registration is not great and I feel like I need some RADICAL NEW TECHNOLOGY to properly deal with those issues.

>> No.981035

animating realistic movements is near impossible without a proper mocap rig. just do npr :3

but this looking cool tho

>> No.981037

>because the mocap data is noisy,
Sounds like a you need a Kalman filter in your life. I'm sure there are motion capture libraries already doing it for you, since this must be a pretty fundamental problem. A fourier transform and high-pass filter would probably work too. Or, even a moving average, although it might make the movements too smooth.

>> No.981041

There are big commercial repositories where you can buy mocap or even order them I guess, but I have an attraction for garbage data because it's an excuse to find ways to make it better.
I wonder if anybody is selling porn mocap files though. I bet Filian and Mizkief record themselves every time when they do it.

>Kalman filter
The mocap data should go to a rigid body simulator first, then to a ragdoll to take care of angular momentum and rotation limits, then to the softbody simulator. But I'm a solo developer and I'm doing this for hobby, so I'm cutting corners.

>> No.981042
File: 582 KB, 1967x1054, 1703220739533440.png [View same] [iqdb] [saucenao] [google]

I want to fix the hands into something better but I'm not sure how I would go about it while keeping the lowpoly aesthetic, they are always either too detailed or too robotic looking.

>> No.981050

>while keeping the lowpoly aesthetic,
Why can't you keep the lowpoply aesthetic while not actually being lowpoly?

>> No.981066

The arms need to be smaller and shorter too.
Make her a t-rex

>> No.981069

why is it called the doge palace? do they trade doge coin there?
I'm just kidding btw it's a joke

>> No.981087

They trade bitcoin

>> No.981093

yea exactly why the venetians were one of the richest for like 700 years

>> No.981142

what is the maya equivalent to holding shift to get precise movement in blender

>> No.981145
File: 499 KB, 1920x1080, Cheese_Sat-20Apr24_11.00.png [View same] [iqdb] [saucenao] [google]

Fiddling around with using blender to make a low poly 3d model in the style of megaman legends. Haven't texture anything yet because I don't know how to.
Here's a 3d model I tried making.

>> No.981151
File: 1.26 MB, 1920x1080, ea (2).png [View same] [iqdb] [saucenao] [google]

venice guy here, ive just started modelling the venetian library o7

>> No.981155
File: 151 KB, 877x924, nano chan.png [View same] [iqdb] [saucenao] [google]


>> No.981165
File: 46 KB, 994x545, m17_01.jpg [View same] [iqdb] [saucenao] [google]

What's a better feeling than seeing the slide of your high poly handgun starting to come together?

>> No.981175

sig p320? good choice that was my first ever gun model too

>> No.981176
File: 2.93 MB, 412x804, 2962437_neovalo_mim-v-3.webm [View same] [iqdb] [saucenao] [google]


it's a hard style to emulate, because it's all about the texture work.

do not get discouraged

>> No.981178
File: 1.63 MB, 1920x1080, file.png [View same] [iqdb] [saucenao] [google]


>> No.981179

Not sure if it's a response to >>980591 but I would like to see more.

>> No.981193
File: 434 KB, 1344x756, comped (1).jpg [View same] [iqdb] [saucenao] [google]


>> No.981199
File: 949 KB, 1920x1080, FACADE.png [View same] [iqdb] [saucenao] [google]

finished. very proud that i modelled 98 percent of it! (the statues and the ornaments on the top frieze are assets from sketchfab)

>> No.981204
File: 1.14 MB, 1823x1279, Scanner.jpg [View same] [iqdb] [saucenao] [google]

Makin a lil police scanner type thing.
Somewhat based off of a Regency M-100. I know fuck all about radios and scanners and stuff though.

>> No.981208


>> No.981211

thank you!

>> No.981213

It's good. Should those seams be there, though?

>> No.981214

in between the arches? no but i decided to keep it because it kinda shows how these things were built modulary

>> No.981216
File: 192 KB, 640x480, 0001-0120.webm [View same] [iqdb] [saucenao] [google]

After a million attempts I was able to set up an iTaSC rig. I wish it had supports for physics but some of the functionality seems interesting.

>> No.981219

Typing values into the channel box

>> No.981220


>> No.981221

Unfortunetaly the poster is not the creator of yhat model and they are trolling.

>> No.981222

Looks nice. Almost looks tangible. I liked the raised surface to the buttons. It's got that microwave keypad feel to it.

But I'm noticing two smaller dots in the very center of the scan and manual buttons. Is that purposeful? Or is the shading a little off there.

>> No.981223

oh my!! I really wonder what her armpits smell like

>> No.981240

Literally this >>973485

>> No.981241
File: 1.58 MB, 1920x1080, biblioteca marciana-min.png [View same] [iqdb] [saucenao] [google]

just finished the front facade im really proud of this! (sorry for the spam)

>> No.981248
File: 330 KB, 901x957, azumango.png [View same] [iqdb] [saucenao] [google]

i had an idea

>> No.981249

that's a neat effect. if you showed this to me without context I'd assume it was a drawing or painting.

>> No.981251
File: 1.55 MB, 1920x1080, LopPlaygirlNormalNSFWCycles.png [View same] [iqdb] [saucenao] [google]

I figured that's the case

>> No.981259
File: 291 KB, 640x480, 0001-0190.webm [View same] [iqdb] [saucenao] [google]

Another quick test with iTaSC. I think it can work for me for when I don't have appropriate mocap data. If it was easier to set up and was like 10x better documented I think people would use it.

>> No.981264

Yeah the raised buttons was what I was planning on making it feel tactile and stuff. Good to hear it worked.
And yeah the bump is fucked, it's just an issue with the gradient. Seems like the falloff of the gradient isn't rounding off. Probably because it's just a linear gradient from Illustrator where I made the keypad.
I'll see about fixing it, but I'm not too worried about it since it's part of a bigger scene/animation and it'll be in a dark car. Like the animation is finished, but I figured I'd do a few cutaways with the scanner for an extra bit of narrative.

>> No.981268

someone redpill me on this guy. who is he and why does he post this stuff in every thread I lurk

>> No.981271

If you like my stuff you can subscribe to my youtube channel:
And the reason I post stuff is because I make stuff.

>> No.981281
File: 147 KB, 1194x1396, m17_02.jpg [View same] [iqdb] [saucenao] [google]

After being a little annoyed by some un-planned for detail on the slide that messed with some of the planned topology, I'm going to properly planning out the rest of the model's parts before moving on with finishing the high poly versions. But at least the grip/lower is basically the rest of the model depending on how much effort I want to put into the parts that are only visible if the slide is back.


>> No.981285

Why would anyone want to convince you to like that faggot?

>> No.981288
File: 2.35 MB, 1920x1080, Scanner.webm [View same] [iqdb] [saucenao] [google]

Finished it up. Should look nice in the video.

>> No.981289
File: 149 KB, 1052x915, anime_05.jpg [View same] [iqdb] [saucenao] [google]

Was doing some quick tests to rig the mesh I have here. Actually kind of happy with how simple the basic workflow is despite no experience with rigging/animation/posing so far. I'm sure that's a rabbit hole When I get more into it.
Believe it or not the point here is not to have a coomer model, I just want to get the base body down well before I start turning her into a real character, and these were just copied from one of the first results form an 'anime poses' search.

>> No.981291

>the point here is not to have a coomer model
To a certain extent it's unavoidable. If you make anything remotely good looking, Cris is going to jerk off to it. Cris probably jerked off to >>981288 too already.

>> No.981293

If Cris can get off to my boxy radio, more power to him. He's made of stronger stuff than I to find eroticism in vintage electronics.

>> No.981298

ok newfag, just want to know what his deal is

>> No.981302
File: 894 KB, 640x480, 0001-0340.webm [View same] [iqdb] [saucenao] [google]

I'm not going to spend any more time on this. There are defects but I wanted to know how it could look like. Don't tell anybody.

>> No.981315
File: 214 KB, 986x1173, mech wip.jpg [View same] [iqdb] [saucenao] [google]

working on blocking out a front mission style mech right now. havent modeled in a few years, getting back into it.

if anyone's got a good tut for making guns i'd appreciate it.

>> No.981335

>Believe it or not the point here is not to have a coomer model
stop being a fucking retard

>> No.981340

Cris, you're into overdrive today.

>> No.981341

I've been here way longer than you. WAY fucking longer. I'm not proud of that fact at all.
As for his deal. He's trying to string together mocap animations by scripting them in notepad or something. With the intent that he's going to somehow shortcut the process of making actually good animation. Though he lacks absolutely ANY self awareness and spams his shit in every single thread. In the time it takes for him to cobble together some kind of shitty solution, he probably could have just learned to animate competently.
A hyper-autist, basically.
There's better hyper-autists here "worth" your time anon ("worth" being a highly subjective term), believe me.

>> No.981343

I just wanted to say that I'm the guy who you're bad-mouthing and I believe you.

>> No.981344

nta but I'm confused as to what the point of this comment is supposed to be

>> No.981345

That's like as if you were asking what's the point of this >>981316 thread and why the jannies won't delete it.

>> No.981346

Does this board even have jannies?
I wouldn't be surprised if they just pop in like once a month to make sure it's not entirely on fire, and maybe to post another donut thread just to fuck with people.

>> No.981348

Yeah, that's still not making sense to me. Maybe I'm just stupid and I need you to explain everything in detail to me, because I don't understand what
>Believe it or not the point here is not to have a coomer model
>stop being a fucking retard
have to do with one another.

It has me thinking >>981340 is on to something, because the things you say have no cohesion with what you react to. It's like the things you say are simply a result of your synapses bumping into each other on accident.

>> No.981349

44 words on expressing the joy of making something.
49 words on making excuses for why it looks nice.
More than half of that post is a clumsy apology for doing things right.

>> No.981351

>I've been here way longer than you. WAY fucking longer.
tell me about soviet russia and I'll believe you.

>> No.981353

One thing I can tell you for sure it that this board is a ghost town because if I look at the analytics on my youtube channel I can see that less than 1% of the views I get come come from "forum links" and I've ever only linked it here.
If you look at the front page on slow days like this you can take a whiff that most of the threads are made by a single person replying to himself. And maybe that's why the jannies keep Cris "parked" in here.
Or maybe it's because the jannies are part of that category that I'm not going to mention but that starts with "tr" and ends in "annies" and they enjoy this situation. Who knows.

>> No.981354
File: 121 KB, 680x1178, 1655993805230.jpg [View same] [iqdb] [saucenao] [google]

plz no more bully

>> No.981356

>anon is using his youtube channel to honeypot the 3 anons on 4chan who try to 3d model
Has weird
>doesn't matter you're in my folder now

>> No.981357

thank you!

>> No.981358
File: 3.58 MB, 2880x1620, Bush did nine eleven.png [View same] [iqdb] [saucenao] [google]

Ello' gents, i tried making some proper cartoon character models after the donut tut but i'm really lacking in fundies, making appealing stuff is quite a pickle for me since i got a bad case of the 'tism and bit of the gay.
Soooo i would appreciate any *pointers on this kind of thingies since most of the stuff in the open is of the furrier, pornier kind, or more on the realistic side which is fine but i lack the IQ for that.

>> No.981359

Cris, why are you worried about honeypots?

>> No.981361

27 hours into blending

>> No.981362

There's nothing there

>> No.981363

IDK it's got a familiarity to it. A palatable appearance I can stand to look at.
The secret sauce to very silly 3d characters is to basically treat them like they're 2D.
You make multiple models which, from a specific perspective, look very much like the original drawings. What you need to do to achieve this will make everything look super weird from another angle, but in that specific one it looks perfect.

>> No.981364

hehe it's blocking content because of the adblocker. It got me too.

>> No.981366

Why are you lying?
Dude just post it on the fucking IMAGE board if you want to post something.

>> No.981367
File: 1.69 MB, 494x406, circle points.webm [View same] [iqdb] [saucenao] [google]

Made this from scratch. I fell down a rabbit hole. First I wanted to work on the hair of my character. Then I wanted to figure out a way to convert a mesh into hair. Figuring I can use only a few edges, and then interpolate the rest. The I wanted more control over interpolation, so I tried to make my own child particle thing. And this is as far as I got today.

Yeah, it's not impressive in the slightest. But I did learn some things. Like how to actually arrange the points around a center using math. My understanding of math is seriously below what's required to do such a thing. So I pretty much just got halfway there before looking up the equation and copying it. I'm not entirely sure how it works, but it does.

The part that's really stumping me, is how to align the point-circles with the edge. I looked up the equations for that too, and got results that are very close to correct. But when I push the edge way off to the side, the points fail to stay aligned by a small degree. I just don't understand rotation nodes at all. They ask me to input angles to work. But how am I supposed to *get* angles in the first place? If I have one line that goes one way, and another line that goes another way. How do you get the angle between the lines?

This is all probably pointless in the end. I should just take the hair curve pill and use the preset node groups they provided.

>> No.981368

ctrl shift N for private tab, maybe it'll work there.
my adblockers(vivaldi + ublock origin) don't block it for me. I'm rangebanned so no local images.

>> No.981369

catbox stopped being shit

>> No.981371

In Soviet Russia, Blender model you.

>> No.981372

I'm sure you know this by now, but you know there's a hair object, right?
You just select an object, then go to the add menu like you're adding another primitive and add a hair object. Why it's like that I've got no clue, but it's retarded and nothing else functions like that.

Still, neato shower heads anon.

>> No.981387
File: 1.10 MB, 3837x1114, body.jpg [View same] [iqdb] [saucenao] [google]

thoughts anons? i want some opinions on if the proportions are correct and if the pose/stance is ok for further work. im afraid of moving forward because my last attempt i got all the way to rigging and only then realized how bad my proportions were

>> No.981392
File: 639 KB, 640x480, 0001-0250.webm [View same] [iqdb] [saucenao] [google]

Quick test with a mech. This model has a lot of articulation including doors on the front that open up and there's the pilot cabin, a turret. But I'll look into that later.

I don't know if it's intentional but it's a mix of male and female parts. It doesn't look bad but it's also very odd.

>> No.981399
File: 1.81 MB, 406x502, ostrich.gif [View same] [iqdb] [saucenao] [google]

I'll tell you a secret about balance, In order to not fall over your center of mass must be placed above or vault over the area formed by your contact points.
Like a human standing on two legs wont fall over as long as their center of mass is anywhere inside the imaginary skateboard square between their two feet.

If you stand on on one foot, your center mass must be held above your footprint or you will tip over.

Or imagine a indestructible round table with 4 legs, a massive weight can be placed on-top of the table anywhere as long as it is inside the square of the 4 legs
but if it's placed outside this square along one of the round edges it will flip the table.

Your robot should walk like an ostrich placing each leg below it's center of mass if you want it to travel efficiently and linear like a human walker would
To walk with a stance more than shoulder width apart you would need to shift your weight side to side and travel in big waddle arches.

>> No.981402

>If I have one line that goes one way, and another line that goes another way. How do you get the angle between the lines?
I don't know what you're trying to do, but the angle between two vectors = acos((a dot b)/(|a||b|)).

>> No.981408
File: 663 KB, 640x480, 0001-0250.webm [View same] [iqdb] [saucenao] [google]

Double checked some parameters and now it makes more sense but I've noticed the shields on the side have big sliding joints. I'm not sure if/how I can make those with the iTaSC solver.

For now I'm just working on the IK constraints.

>> No.981413

slow, off-center bipedal walkers piss me off because the only way it would stay upright is either a gigantic gyroscope inside it or straight up magic

>> No.981415
File: 130 KB, 1296x514, angle math.png [View same] [iqdb] [saucenao] [google]

>I'm sure you know this by now, but you know there's a hair object, right?
Yeah. But I still wanted to try to do it on my own.
I'm starting to appreciate the curve nodes even more, because in searching for solutions of my own, I'm understanding a little more why curves work the way they do, and why the hair system relies on curves.

Like, if you convert a bunch of mesh lines into curves, then you can think of the mesh lines like the handles of curves. So it would probably be better to just set up the curves right away, instead of wasting time transforming the mesh into curves. I'm guessing that's why hair is its own object. Because if you were going to create hair by converting mesh, then you would create a new mesh object to start the process anyway. So why not just skip that part and get directly into creating an object that is specifically designed to handle curves.

I think that's the same equation I found when I looked it up. It's the equation I tried to implement. But I probably got wrong, since I can only find cosine not acos(arccos) in vector math nodes. Float math as it, but not vector math.

The results look ok at first, but when I rotate the edges certain ways, the point-circles don't always align with the edges properly. It's like there's this space where they struggle to cross, and then suddenly *fwip* they catch up again.

I'm also no sure how to get arccos to work. Every resource I read sounds confusing. Like:
>Arccos is the inverse of cos, so you swap values.
What values?
>Arccos knows the angle of cos. And cos knows the magnitudes of arccos
ok, but how does arccos ACTUALLY get the angle? Like, how are we converting these two lines into degrees?
Everything sounds so vague and groundless. It's like arccos defines itself. It just comes from the ether knowing angles. And you have to give an offering to arcoss for it to share its knowledge.

Anyway, I'd appreciate some insight here.

>> No.981423
File: 295 KB, 556x396, walker.gif [View same] [iqdb] [saucenao] [google]

Like it might move like this. Both feet on the ground, move the center of mass over the stationary foot, move the back leg forward, repeat.

>> No.981428

legs are way too far apart, anon
wrists and hands too thin as well
boota might be a bit too low

>> No.981429

When you import the .fbx into Blender, look at the import options. Go into the Armature tab, select Ignore Leaf Bones, Force Connect Children and then set Primary Bone Axis to X and Secondary to Y.

>> No.981436

ty for inputs, i did just now see the legs could be too far apart, ill remeasure the ratios of the hands but they seemed to be correct from references.

>> No.981449
File: 1.98 MB, 2048x2048, swd (1).png [View same] [iqdb] [saucenao] [google]

modestly and humbly speaking... i have peaked

>> No.981450
File: 649 KB, 1280x1984, 3-modelling-challenge-apr22.jpg [View same] [iqdb] [saucenao] [google]

Model a fire hydrant.

>> No.981452

Well, it doesn't pass as a drawing. But it still looks nice regardless. Very good work.

>> No.981453

You can't directly rig hair nodes, that stuff is for pubes and beards.

>> No.981458
File: 5 KB, 128x128, 1691957919152453.png [View same] [iqdb] [saucenao] [google]

Didn't want to kill a thread for this question.

Anons, in the near future I need to commission a 3D artist for a custom model, but I don't want to drop $1000 on a passion project.
I want to cut costs by just buying a ready made 3D model of a dude and hiring someone to put a custom head on that model, pose it, and separate it into pieces for 3D printing.

Is this something that 3D artists normally do, or is editing a model like that (after the custom work) more labor intensive than I think it is?
How much should I expect to pay for a job like that?
I was planning on just hiring someone off CGtrader, but I don't know if there are better places to shop.

>> No.981459

I don't think you could directly rig hair particles either. I'm not really seeing the difference.
In either case you'd rig the mesh underneath the hair that it comes from. In the event it's something that's supposed to be all hair, you'd just stick an invisible bit of mesh underneath the hair anyway.

>Didn't want to kill a thread for this question.
Lucky for you there's a bunch of generals and questions threads for you to ask in. I'm sure you'll get an answer here though.
Like from me. Yes, people do headswaps. In fact, go onto any 3d modelling discord server, say a single word and you'll get like a dozen retards messaging you about it, because they can't model but can do headswaps.
Now I don't know if it's something you can search up, but any modeller with basic experience can manage it. Fuck anon, you could manage it yourself after a week. Save your money.

>> No.981468
File: 660 KB, 640x480, 0001-0250.webm [View same] [iqdb] [saucenao] [google]

I usually don't like line renderings but on this model it looks good. Internal renderer still has its mojo.

>> No.981470

>Anyway, I'd appreciate some insight here.
The equation gives you the minimum angle between two vectors. Two vectors actually define a plane, and there are two ways to rotate one to get to the other, the short way and the long way. That's probably why you're getting some kind of a popping effect. To fix it, you also need to know the normal vector of the plane, and rotate around that. The normal vector could possibly be defined by the hair curve you're using. It probably has some direction. In that case, you can do something like this:
If a X b . dir > 0, angle = acos((a dot b)/(|a||b|))
else, angle = 2pi - acos((a dot b)/(|a||b|))

Geonodes probably also have some "rotate around a direction"-function, which you could use to do all of this for you. And the whole problem you're working on is probably just solvable with matrix base change. I don't know, though, because I don't really know what you're trying to do.

>> No.981471

There are tools you can use to generate a human for you, e.g. Daz and MakeHuman. There are probably many others too. So maybe you're just looking for custom clothes, really.

>> No.981487
File: 1.47 MB, 1920x1080, LopPlaygirlNormalNSFWCycles.png [View same] [iqdb] [saucenao] [google]


>> No.981492
File: 93 KB, 1232x1613, ngons.jpg [View same] [iqdb] [saucenao] [google]

What's the best way to fill these faces bros? Looking back I probably should have just used more geometry, but that ship has sailed now. I did try adding more loopcuts to join them that way, but then it tightens everything up too much. I also tried just making irregular quads with the corner verts and the beveled verts. You can imagine how that turned out.
I've tried the google but I can only find results about beveling a edges and full corners, nothing to do with beveling vertices on a face and dealing with the cleanup.
Note: When I originally beveled the verts it did it that usual way turning them into a diamond shape which worked, but also made the shading look much much worse. Not sure if that's relevant.
Also the face in the middle is a separate mesh entirely so there's no issue with it.

>> No.981496

You should try the no-topology workflow.

>> No.981497

What's wrong with it as is?

>> No.981499

I don't know what this is but it sounds like bullying.

It gives weird shading with them being ngons.

>> No.981500

if you want it subdivide cleanly i think you're just going to have to redo it with more topo.

>> No.981501

Fuck. I'll try that instead.

>> No.981503
File: 486 KB, 854x700, Screenshot_20240424_211614.png [View same] [iqdb] [saucenao] [google]

You could do something like this, if it's just a shading problem, and you're not too autistic about it. If you want to subdivide that, you probably need to do some work with sharp edges and add some extra geometry.

>> No.981504
File: 1.24 MB, 1000x1000, Low poly fire hydrant.png [View same] [iqdb] [saucenao] [google]


I recently did a 3D model optimisation course for rendering large scenes real time. As practice I made a low poly fire hydrant to be viewed from distances over 200 metres.


Sent from my iPhone

>> No.981505

Thanks I'll try that first. I don't think the shading is even that noticeableof a problem, but it's one of those things that because I know it's there it's aggravating me.

>> No.981506
File: 126 KB, 557x467, topology.jpg [View same] [iqdb] [saucenao] [google]

You could try something like this. It works well for me. It helps me keep my vertices organized.

>> No.981507

What is that? The whole thing triangulated?

>> No.981508

Are you making a fire hydrant? I'll help you if you're making a fire hydrant.

>> No.981510

Nice. I'm sure it looks great from hundreds of meters away.

>> No.981511
File: 507 KB, 640x480, 0001-0250.webm [View same] [iqdb] [saucenao] [google]

Jokes aside it does have its uses for when you want to run some sort of FEM simulation like heat diffusion or you can bake AO into vertex color and then selectively decimate. If you notice, the triangles are chosen to be as equilateral as possible. It's an experiment anyway. I've used https://www.cgal.org/ to do it.

>> No.981513
File: 81 KB, 1535x688, WIP-16a1.png [View same] [iqdb] [saucenao] [google]

fucked up my fire hydrant

>> No.981515

It's kind of low poly and aliased, and also not a fire hydrant. Good try though. It would be decent if it were an m16.

>> No.981519
File: 150 KB, 1525x838, isolated bevel test.jpg [View same] [iqdb] [saucenao] [google]

Maybe something like this? I don't know. But it shades well, keeps quads, and isolates.
The idea is to inset, delete corner faces to clear space, bevel by 4, because it will create 4 new vertices, which will make quads possible. Then fill corner with grid. Set to a span of 2, and offset until it aligns properly. The select the center vertex of your new created faces, and smooth about 20 times, to force it into the center.

I skipped some minutia, but you'll figure it out.

>> No.981520

Maybe a lighting issue but it looks dirty, like an used faulty teddy bear

>> No.981523

It looks ok to me but don't show "the wires" to anybody who asks because you're going to get jumped on by strangers who're going to tell you that "you haven't solved the puzzle of topology". Yet, somehow, if you keep it to yourself nobody is probably ever going to notice anything.

>> No.981529
File: 406 KB, 832x840, whatif.png [View same] [iqdb] [saucenao] [google]

i dont know much about hard surface models topoplogy yet I've only done characters but wouldnt yo want to get rid of that weird shape?, even if its 3 vertices more?

>> No.981530

(That guy used way too much topology instead of letting the subdivision do the work for him)
The shape's not THAT bad but it's not necessary, oh and also I just realized they fucked up the distribution of polygons on the cylinder with those cuts. wtf is this supposed to be a low poly model or something?

>> No.981533

I wish somebody would critique my topology solution. I wish my models became more of a nightmare to deal with the second I wanted to make a change to them.

My recommendation is to not do what you're doing.

>> No.981534
File: 138 KB, 1410x1090, Let the subdivision handle the curves.jpg [View same] [iqdb] [saucenao] [google]

I may be a smug topology asshole, but that doesn't mean I remember to include images in my posts..

>> No.981535

so how do you get a sharper crease? just mark the edge as sharp?

>> No.981536

I tried that configuration before posting. Didn't notice a difference, so I deleted the excess. But I suppose that would be better if you wanted to add more supporting loops around the rim for whatever reason.

My first idea was to make the subdivision handle the curves. But after pulling loops back and forth for a minute, I realized I wasn't able to get the shape the anon was going for. So I figured I had to bevel to do it. Can you get a similar shape as his?

>> No.981538
File: 325 KB, 220x155, angry-cat.gif [View same] [iqdb] [saucenao] [google]

>just mark the edge as sharp?

>> No.981545

For pricing I'm more concerned about how much a custom head is likely to run me.
I can't just generate one because they're wearing a fictional design of gas mask that would need to be modeled. The clothing isn't anything special so I can just use a premade model for the body.
I appreciate the tips though.

>> No.981546
File: 723 KB, 2180x3669, Are sharp edges even good.jpg [View same] [iqdb] [saucenao] [google]

If you really need the edges to be super sharp then yeah there might not be an option other than to use more geometry, I still wouldn't say as much as the original had. You'll want to plan out the number of cylinder sides so that they generally line up effectively with the details you want to add to maintain the curves, and once you start adding more geometry you open yourself up to more/worse instances of pinching/smoothing-artifacts that are also sharper.

It's something I would caution against though - or at least do on purpose and not as a crutch. I spent literally years trying to get absolutely perfect subdivision geometry on my meshes while zoomed in 1 million % until I started to see that if you looked close enough you could see the same artifacts on some of my favorite works from artists I really respect in the hardsurface space if you look close enough.

>consider how close you are to the object in practice and don't analyze your mesh for sharpness or artifacting closer than that
Unless the edge needs to hold up as sharp when very close, it's probably going to read better (As well as being easier). If you just subsurface all your edges into sub-pixel 90d angles then what was the point. You can always find shit wrong with subdivision when you zoom into every single junction and it will never look sharp enough.

Meshes in picrel have tradeoffs, and can look basically the same depending on how close/far you are. I'd recommend against right always just because it's hard to work with, but even the low poly one looks fine at most distances imo and if you're baking it into a normal map the softer edge will bake better. You can get sharper edges with closer edge loops but to minimize artifacts it does need to be a little higher poly.

>> No.981551

Ahhh, you needed the head modeled. I figured you had a head already and wanted it put on another body. Yeah that's a bit above the skill level of the Discord spammers.
I think $100-$150 would be a fair estimate for something like that depending on how complex and in depth you wanted the person to go. Though they'd probably charge you to put it on a body if you asked for it (though again, it's something you probably could manage yourself if you found a body you liked).

>> No.981552

Remember what I said about "organic" challenges?
3 or so people did the "real challenge" that some anon came up with, yet people are pouring out of the woodwork to solve some random hardsurface problem.
Challenges are fuckin gay.

>> No.981554

I get what you're saying but it's easier to explain or demo 1 specific shape than actually model something in totality.

>> No.981556

Thanks, Anon. If I can get the job done for $150 I'd be pleased. I'll try to shop around a bit once I have all my reference material together.

>> No.981563

>3 or so people did the "real challenge" that some anon came up with,
What's wrong with making 3 or so people happy? That's like half of the population of this board already.

>> No.981565
File: 348 KB, 1686x876, 1694430906330083.png [View same] [iqdb] [saucenao] [google]

how did i do? they gave me an ai reference pic and only the front view as well. how do i get better at modelling from just 1 reference?

>> No.981566

Your curves are not quite as smooth, and the profile of the legs is not round. That chair would also probbaly tip over if you actually placed your back against the backrest.

>> No.981567

>Your curves are not quite as smooth
how do i go about making it more smoother? should i have used curves instead of box modelling it ?

>> No.981578
File: 52 KB, 498x280, chairs.jpg [View same] [iqdb] [saucenao] [google]

>how do i get better at modelling from just 1 reference?
Use ergonomical measures. Nice model but that chair is unseatable.

>> No.981583
File: 132 KB, 1754x826, 3.jpg [View same] [iqdb] [saucenao] [google]

learn about custom normals and data transfer modifier

>> No.981584

I have done tons of this type of works, my advice is to bring a character and try to sit him on the chair. That would give you an idea of how usable it is

>> No.981592

>learn about custom normals and data transfer modifier
nta but does it have something to do with baking?

>> No.981593
File: 196 KB, 1280x720, maxresdefault (36).jpg [View same] [iqdb] [saucenao] [google]

Idk why he's bringing it up for a basic hardsurface shape. But my familiarity with it is in the context of tweaking toon shaders like in picrel.

>> No.981600
File: 1.43 MB, 1013x869, cswork.png [View same] [iqdb] [saucenao] [google]

working on lowpoly render using CS1.6 textures.
I kind of love simple things like this.

>> No.981601

Isn't it common to use custom normals for nice mid-poly bevels on hardsurface objects?

>> No.981603

Not that I know of, but what do I know. I'm just a 30yo moodler. I see shapes, I moodle them. Some people might find it easier to do weird fancy custom normals by mapping a cylinder's normals onto a weird shape to get smooth shading. I'll stick to my day long moodling sessions.

>> No.981611

very cosy, got some more?

>> No.981615

Currently not i need to check out a CT model, rig it probably if it doesn't have bones, and UV map new stuff. Im new in blender but since i like low poly old games this is enough for me.

>> No.981624
File: 3.06 MB, 1920x1080, chair and fireplace.png [View same] [iqdb] [saucenao] [google]

added chair and fireplace...

>> No.981626

Add a steaming coffee mug, and a foxgirl gf.

>> No.981627

Steaming coffee mug is a great idea.
Fox girl im a noob to do so but as a wallpaper picture it would suffice, soft furry that is.

>> No.981629

Also definitely needs a rug with a folded up corner.

>> No.981635
File: 3.70 MB, 1920x1080, mug,foxgf,rug.png [View same] [iqdb] [saucenao] [google]


i added your ideas.

>> No.981636 [DELETED] 
File: 957 KB, 960x540, 1696402084309007.webm [View same] [iqdb] [saucenao] [google]

Turns out that my previous effort at making a sim from scratch had reversed normals on a massive, key area, rendering my results uncontrollable as the pressure being applied from the tissue to the bones was going in the wrong direction meaning it had no affect on muscle defintion in either the positive or negative direction.

Now I realize that I should have started in a different bindpose so the tets between the legs which were very close together on bind don't stick together which I was unable to work around even with selectively painting increasing the tet resolution to near maximum and applying a delta mush deformer. Regardless, I am curious what anon's opinion is.

>> No.981638
File: 950 KB, 960x540, 1704101809892897.webm [View same] [iqdb] [saucenao] [google]

Turns out that my previous effort at making a sim from scratch had reversed normals on the entire interior fat mesh rendering my results uncontrollable as the pressure being applied from the tissue to the bones was going in the wrong direction meaning it had no affect on muscle definition in either the positive or negative direction.

Now I realize that I should have started in a different bindpose so the tets between the legs don't stick together which I was unable to work around even with selectively painting increasing the tet resolution to near maximum and applying a delta mush deformer.

>> No.981641

Imagine taking all that time and skill to make something so disgusting.

>> No.981650

That's why you should say no to Position Based Dynamics and Tetrahedralization.

>> No.981651
File: 94 KB, 1272x761, m17_03.jpg [View same] [iqdb] [saucenao] [google]

We're almost done with the high poly here. I had a bit of procrastination not wanting to work on the smaller bits of the gun but we're close now, obviously it'll look much more complete when the lower is made high poly.

>> No.981658
File: 2.98 MB, 1920x1080, 1683746813232563.png [View same] [iqdb] [saucenao] [google]

i keep forgetting about this project, what all shit can i add in the platform other than humans to make it feel more lived in? also anyone knows whats up with the bg? is it because of the denoiser?

>> No.981660

Panties on the clothesline. And A-cup bras.

>> No.981687

most importantly think about why things are the way they are. for instance, why are the boxes neatly stacked in the corners, outside, not near anything? if i were a lazy bastard i would the boxes right outside the door. if i were less lazy i'd keep it as one stack, probably under the shade of that roof over the window. what happens at night, do we need lights here? we're outside, do people smoke? is there a leak in the roof above? will people damage the paint on the floor where they walk? "lived in" is really just a lot of tiny stories that add up to a rich single picture. i also really want to see what that room is for through the window. i think it's an elevator (i don't know why an elevator would have a window) in which case it'd be nice to see the button or something for a more definite purpose

++ add a ramp in front of that door or something, someone's gotta have tripped on it before

>> No.981689

Rate my bad boy
Am I good to rig and learn some animating?
I noticed the arms are too long.

>> No.981690

you should go ahead and rig this so you can realise why your topology is bad

>> No.981693
File: 213 KB, 2254x1268, 1701955375915891.jpg [View same] [iqdb] [saucenao] [google]

I'm having the same problem as you atm and I even have a human in there. The issue is I don't know how to fill up the empty space. There's some placeholders there I haven't done yet like the two seats, the little table, and a laptop, but it still feels empty. I'll just throw a sheet on those unfinished cardboard boxes in the corner because I'm a lazy animal.
I can't even find a decent reference of what a basement is like and I have to go purely off my memory because the only search results I get are about turning them into "games rooms" and that kind of thing.
Sorry I didn't mean to hijack your post without offering something. Maybe you could add some garbage? That always reminds me people are somewhere. Or graffiti even, that's a common human trait. Unless you're wanting it to be more clean and tidy. Maybe some signs? Aged or otherwise.

>> No.981723

External AC unit, light bulbs, light tubes or lamps, hanging plants
Maybe bottles?

>> No.981726

debirs and liter, like paper wrappings, some grime in the crevices, and little bits of dried grasses like the pine needle sized ones birds drop everywhere

>> No.981727
File: 654 KB, 1080x705, a89db731ac46003b100516f88d583797.jpg [View same] [iqdb] [saucenao] [google]

Start amassing some reference images, anon.

>> No.981729
File: 964 KB, 4684x2148, lowork11.jpg [View same] [iqdb] [saucenao] [google]

not trolling, just coincidence
I was sculping loppuny to make a figi
but I coldnt pose yet and forget to update the wip,
also Im planing normal/huge breast.

>> No.981733
File: 144 KB, 1600x1067, new-gas-furnace-installation.jpg [View same] [iqdb] [saucenao] [google]

Needs a stack of shitty plastic storage containers.
Some old pillows from an old ikea couch that fell out of commission, but you haven't thrown away yet.
Some half used paint buckets, because you'll need the same color later to make repairs.
An exposed furnace, with that tubing that goes to various places in the ceiling.
The laundry machines.
Skinnier floor rafters, with more wires going this way and that.

Just google "unfinished basement" to see the innards of a basement, and not the converted front.

>> No.981734

So a slutty bunny to rival Warframe's.

>> No.981741

The people who owned my uncle's house before he bought it put urinals in the basement, and the garage. Not gonna lie, they were pretty useful.

>> No.981755
File: 475 KB, 1107x757, state of war blend.png [View same] [iqdb] [saucenao] [google]

Since i like a lot old game called State Of War im making kind of models from that game they are not 100 accurate.

>> No.981760
File: 995 KB, 4020x1106, all.jpg [View same] [iqdb] [saucenao] [google]

made some small changes
>thinner torso
>abs smoothed but middle groove deepened
>increased hand size
>reduced leg spread
>ass moved up
>widened hips
>longer thighs
>rotated feet outwards
>changed the collarbone region so there is no longer an edge from the chest to the neck because looking at thin female models there doesnt appear to be one for most thin women
are the changes in the right direction? might need to change the head length, might be too long at the back, cant tell how long the head of the model im referencing is so i keep overcompensating i think

>> No.981761

reminds me of C&C, makes me want to play again but already went through the series once

>> No.981767
File: 1.58 MB, 849x1200, 1695820417598609.jpg [View same] [iqdb] [saucenao] [google]

no way thtats too lewd
>i also really want to see what that room is for through the window.
pic rel was the main ref i used for the foreground, the room was supposed to be some sort of mini super market where workers come and bug cigs and shit hence the carboard boxes but i wasnt sure how to portray it in a shady kind of way so i just kept it empty until i figure out a way.
oh yeah light bulbs make the most sense but its hard integrate it in a day scene while making it sublte, pic rel has lightbubls but the aesthetic it gives off isnt what i was going for.
>and little bits of dried grasses like the pine needle sized ones birds drop everywhere
can you show me a pic? i cant imagine grass in industrial settings like this
thats the difficult part, i think i dont know how to properly utilize reference images because i am looking for something between what you sent and something more cleaner and industrial

>> No.981783

State of War is close to that.

>> No.981792
File: 30 KB, 612x377, istockphoto-488940003-612x612.jpg [View same] [iqdb] [saucenao] [google]

there are types of birds that will make nests in buildings like that under the roofs, they tend to carry all sorts of little debris everywhere as they fly in and out and around so i always see these tiny needles in crevices mixed with dirt and grime from peoples shoes

>> No.981806

There's some cool ideas there I hadn't thought of so I'll include them. I was going to do a furnace or boiler, but I thought it might be too bulky. And I was going to do the pipes, but I wasn't sure what kind of pipes a basement actually has or which way they run.
I'll google unfinished basement and see what comes up.

>> No.981810

I tried to use custom normals in maya scenes imported into unity, and it caused a lot of problems and was a pain in the ass to fix. If your project has to transfer from different platforms it can be unpredictable how normals survive.

>> No.981811

take it as a compliment then

>> No.981847
File: 2.08 MB, 2749x1607, Tokyo Extreme intensifies.jpg [View same] [iqdb] [saucenao] [google]

Made a bunch of modular highway segments today.
I don't really need em for any kind of project right now, but I figured they'd be useful in the future. And weekends are my time off from client projects to fuck around or do "housekeeping" like getting my assets in order.

>> No.981890

How do you cope with looking at all these abominations every day?

>> No.981899


>> No.981905
File: 273 KB, 640x480, 0001-0300.webm [View same] [iqdb] [saucenao] [google]

I'm still learning how to manuever the iTaSC solver. I'm trying to do an uncommon walk cycle. I'm also not very motivated today but maybe I'll make it work eventually.

>> No.981906

this >>981899 is not >>981638

>> No.981907
File: 2.78 MB, 790x676, eyerigtest.gif [View same] [iqdb] [saucenao] [google]

Successfully reverse engineered an eye rig. Want to make a mouth rig using similar techniques, but a little more expressive.
Still needs textures.

>> No.981910

Good job including movement in the eye lids when the iris moves

>> No.981932
File: 2.56 MB, 2880x1620, TopoGigiology.png [View same] [iqdb] [saucenao] [google]

Ah cool, i was getting a bit worried of making people feel like Sgt Hartman finding that donut in FMJ.

Topo is driving me crazy tho. those fangs have way more polys than they deserve, silly me thinking grabbing some cones and putting a crapload of segments on them was a good idea, well, if people think it's fixable i'll finish the mouth and apply mirror + subdiv/subsurf.

Getting some Muppet vibes from this, maybe i'll go in that direction, giving it some wiry boneless, bendy arms, i'm more of a M.C Hammer than M.C Escher guy so i don't know if i can deal with perspective stuff, trying to align 3d collision boxes with isometric prerendered models in Godot gave me a bit of PTSD, i'll suck it up if push comes to shove tho.

>> No.981934
File: 2.32 MB, 1920x1040, file.png [View same] [iqdb] [saucenao] [google]

>> No.981954
File: 416 KB, 729x624, 1693583920987661.png [View same] [iqdb] [saucenao] [google]

I was right in thinking that the legs needed to be spread further apart the fatter the legs get or else the tets merge together.

However, when inflating the mesh and changing the fat/skin I've found that the skin sags a lot and I dont necessarily want the bones to change size, because ive created scripts that set up basic human skeleton to real world scale, but i potentially need to warp the muscles.

In renders, I've found that my final denoiser is drastically different than either rendering with IPR denoiser or no denoise

>> No.981958

Stop punishing yourself.

>> No.982008
File: 383 KB, 741x653, 1687187640642654.png [View same] [iqdb] [saucenao] [google]

Please. I'm learning how to create.

I increased the muscle size and created new inner fat and outer fat meshes and then simulated. It would take me too long to render them out for now, but the unfortunate part is my renderer is having some issues and the gpu path looks worse than the cpu and still as pictured here, the cpu shaders require uvs, which i didnt want to do at this time, so artifacts are visible. Regardless, comments on the form are welcome.

I realize that now the time has come to create the rest of the main body, including finally the feet and the hands and get back to animation

>> No.982018

Continue punishing yourself.

>> No.982019

He likes this type of shit, he simulates with one hand if you don't what I mean

>> No.982030

Why is the jam making thos

>> No.982032

Why there are so many comments cut in ha

>> No.982036

It looks nice so fa

>> No.982046

parked my ca in avard ya

>> No.982052

Wtf is happening there is so many com

>> No.982078


>> No.982079

What is wrong with this we

>> No.982121

1 2 3 4 5 6 7 8 9 0

>> No.982188

Does anyone know why I can no longer box select in blender 3.6? It still works in the viewport but won't work in the uv editor. I have to press b every time to box select in it.

>> No.982189

press t

>> No.982190

Thanks. I thought I might have pressed something by mistake but wasn't sure what.

>> No.982196
File: 368 KB, 933x984, osaker.png [View same] [iqdb] [saucenao] [google]

who up sata'ing they andagi rn???

>> No.982227
File: 2.49 MB, 4298x2262, body.png [View same] [iqdb] [saucenao] [google]

another few changes after downloading a free model from open 3D lab, studying it realized the pubic area was way too low, and i matched the arm placement and rotation because i felt it looked more natural, the previous version messes with the feminine look i think, looking at references i also made the side of the buttocks a little flat for the way the muscles and fat connect there, if anyone sees something wrong or off please tell i feel like i could be missing things that are obvious but not seeing it

>> No.982241
File: 1.14 MB, 1920x1080, RegLowPolyBurger.png [View same] [iqdb] [saucenao] [google]

Made today this low poly burger.

>> No.982259
File: 336 KB, 640x480, 1686517375840578.webm [View same] [iqdb] [saucenao] [google]

Instead of creating the rest of the body I decided to see if I could get wave propagation between meshes. It was a lot of work. Basically, the bones are there, the muscles are there, the fat is there and the paddle is there and they are all simulating at the same time.

Now I know, the shaders are a little suspect. The gpu path is incorrect and the cpu path takes so long i get fewer iterations

But, in the end, i think this prepares me for greater things because "paddle-like" objects interacting is the future.

>> No.982276

This is so disgu

>> No.982281

why, though?

>> No.982291

The blue hot spots make it look diseased
The post processing and frame rate make it look real enough that it looks like this is a real thing that's gross and weird looking.

>> No.982293

Why are you asking? It's plain as day. The blue-green coloring looks rotten. The fat is all lump and uneven. The buttcheeks are fused together. The flesh sticks to the paddle. Now, I'm tolerable of a little bit of cellulite, but this is an absolute catastrophe of an ass. Do you really have to make it look so gross while you work on the technical stuff? Can you really not make it look less appealing? Or is this some kind of gross-out fetish you have? Some people are really into gross shit.

>> No.982303

you should take it as a compliment that your artwork can stir such powerful emotions

>> No.982307
File: 513 KB, 1920x1080, 1712118406891390.webm [View same] [iqdb] [saucenao] [google]

Another take, rendered in different software on gpu. Its really really hard to make the paddle and the fat collide and then not stick together for a frame or two when they pull apart. I did all I can to minimize it. It used to be a lot worse.

Now what do you mean "the buttcheeks are fused together" ? Looking at this new webm, can you clarify that statement?

>> No.982308

Are you one of those perverts who like roping other people into your fetish by asking inane questions? "DAHUUURRRR, HOW ARE THE BUTTCHEEKS FUSED?!" Have you ever seen a but in your life?

>> No.982317
File: 1.00 MB, 1920x1080, swordtest.png [View same] [iqdb] [saucenao] [google]

Made this piece of crap i might texture it at least.

>> No.982340

not everyone is a 90lb twig. Liking heavier set figures doesn't make you perverted, and no, to answer your ?, I am not baiting you. I put a lot of work into this and will continue to do so.

I made a simple mistake with my prior paddle animation render, my renderer simply could not handle it, despite me subscribing to it. Now I switched renderers and need feedback so I don't make the same mistake.

Also, "how are the buttcheeks fused" is a legitimate question.

The underlying left and right glutes are separate and attached to their corresponding side on the left and right pelvis. Ofc there is a fat and skin layer above that that is one mesh.

>> No.982375

>The underlying left and right glutes are separate and attached to their corresponding side on the left and right pelvis.
Should have just said that earlier.
Still looks gross though. Butts are supposed to have cracks, because the two glutes touch in the middle. But that's just 1 issue among many. Look man, you're making man-made horrors beyond comprehension. So I don't know why you're acting coy about it. Just own it. You have this lab grown lump of flesh, which may one day amount to a pretty good body mass simulation. It already shows a lot of promise. But right now, its form is nightmarish.

>> No.982471
File: 769 KB, 1080x1080, outline osaker.png [View same] [iqdb] [saucenao] [google]

how do we feel about a double outline.

>> No.982474

I think when it works, it works. In your case, I don't really think it does. I'm not really sure what it's adding to the model except making the outline more visible. Maybe if it was a solid black outline, plus a white outline that moved like a scribble, then it could be interesting.

>> No.982478

Threw nead

>> No.983704

There are multiple ways to go about it (curves being one) but I would get the shape of the inner frame blocked out with planes, make the curvature look good, then solidify them outwards. Then you can add bevels and stuff between the solidified edges without worrying about the shape.

>> No.983850

Hey bro can you send this model? I need it for very important non-sexual purposes.