[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: wippy.png (2.13 MB, 1500x882)
2.13 MB
2.13 MB PNG
/wip/ - Works In Progress - "You know you guys can make this thread yourselves, right?" Edition

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

Previous thread: >>980444

List of free resources: https://pastebin.com/cZLVnNtB (embed) (embed)
/3/ Discord: https://discord.gg/gbYCEBPuK2
>>
>>982474
yeah I'll probably make it thinner and make it just black for now. Have never really messed with grease pencil before now so just trying out whatever random thing comes to mind. long term plan with this model requires simplicity i think. idk yet.
>>
File: 1711495302935086.webm (1.02 MB, 1920x1080)
1.02 MB
1.02 MB WEBM
Spent all day figuring out how to redo the "crack".

Forgot to UV before I rendered, so my procedural shader that breaks up the spec didn't work.
>>
File: City.jpg (1.62 MB, 1920x1080)
1.62 MB
1.62 MB JPG
Started putting the full scene together with the car. Which has some preliminary animation on it.
It's extremely early stages. I'm hoping to have a bunch of canals, some bridges across the water to fly through/under, flying traffic in the distance, some highways, and probably some holograms and shit too.
Still, it's nice to finally get this shit in the scene.

It was a fucking nightmare trying to get it all to work though. Turns out my "solution" of randomly generated buildings wasn't conducive to an entire city (who knew). Verts were around 70m (had to array the buildings to make the animation loop), and RAM usage skyrocketed to like 60gigs on render.
I had to go back in and make low poly versions of each of the buildings, which didn't fix the problem. So I had to say "fuck it" for generating a random building in place, and just use the parts I had to make buildings manually (really just turning on the array modifier), and just instancing those with GN instead of generating. RAM use is under 9gigs now. I think I didn't need to make low poly versions of the buildings if I just did that in the first place, but I guess they're nice to have.
In hindsight, I think the main culprit was the "Realize instances", which I was trying to avoid using in the first place (for this exact reason), but with the way I set up materials for the random neon and accents I had to use it because without it they apparently count as separate objects. Blender really fucking needs a "real" randomness shader node, not the object based or island based one.

Still, no better way to figure out if something is actually usable in a project besides actually using it.
And that's it for my blogpost.
>>
>>982484
I still don't understand what are you trying to achieve?
>>
>>982484
Is this morbid obesity cellulitis fetishism?
>>
>>982484
Are 2 chicken legs fused together?
>>
>>982505
>I still don't understand what are you trying to achieve?

FEM and vellum simulation of characters I design doing animations I keyframe, not cop-outs in blender. No compromises. Right now, I am working on the stomach and limbs
>>
>>982484
you don't need to go to all this work to get a nut anon, just visit >>>/fit/fph
>>
>>982510
it doesn't feel like work, going through the steps lets me do it for any character faster later
>>
File: assymcgee.jpg (54 KB, 234x320)
54 KB
54 KB JPG
>>982484
Turn it into assy mcgee
>>
File: 0001-0500.webm (1016 KB, 640x480)
1016 KB
1016 KB WEBM
Still doing exercises with the animation system. I'm totally a noob at animating manually. But the technical stuff is working and I wanted to have a taste of what a longer animation themed like this would look like.
>>
>>982563
are you even making use of easing, anon?
>>
File: anim.jpg (311 KB, 1440x831)
311 KB
311 KB JPG
>>982585
No. For now I'm just interpolating poses, still looking for bugs in the C-side of things. I really really have no idea how this art thing works.
>>
>>982589
easing is like the bread and butter of animation, not only in 3d but in all animation in general, practice it a bit, first exercise of 2d animation is a ball moving on the x and y axis and bouncing, look up some 2d animation theory on the subject and you'll see some tricks there, like the deformation of the ball, keeping its overall space through transformation, acceleration, deceleration, exagirating, etc.
>>
>>982594
I see you're trying to give him helpful advice, but this dumbass is completely resistant to taking it.
He's hellbent on his weird ass "solution" that looks like trash to shortcut learning animation entirely (his own words).
>>
File: MainGuyTest3600001-0250.webm (237 KB, 1920x1080)
237 KB
237 KB WEBM
>>982476
first character, can a noob get some feedback? around 20k tris rn looking to polish it up and paint it in blender then port to UE because substance painter scares me and im not sure I want to spend the time baking normals when im going for a low-mid poly look.
>>
>>982636
You can get up and running with SP in under an hour anon. There's nothing to be scared of.
Meanwhile painting in Blender will take you hours to fumble together any kind of "working" solution to approach even a fraction of what SP does.
I can't speak for this exact playlist, but the official quickstart tuts used to be really good to get started with SP.
https://www.youtube.com/playlist?list=PLB0wXHrWAmCwnqWfKdGEmbtSKN2EzvLrY
There's a few things you pick up with experience, like workflows and stuff, and you get a bit more comfortable with the software, but it really is intuitive and easy to use.
>>
>>982594
>>982620
Try to think of it this way. Suppose I have a little storyboard where Character goes from point A to point B, then performs action C, pauses and goes to point D and so on.
What happens between checkpoints is not very important so long as I hit those checkpoints. Because that's what tells the story.
Machine-interpolated motions made with 20 years old algorithms are not going to look great, but that's ok to me because I'm not a AAA industry standard compliant studio in my bedroom.
The iTaSC thing has been part of Blender since 2.5x. It's poorly documented but as I'm researching into it, it seems to be the only readily available tool which could be useful in terms of what I've just described.
>>
File: 1.jpg (379 KB, 3840x2160)
379 KB
379 KB JPG
started this shit yesterday
my first time ever making a car
>>
>>982642
thanks for the tuts anon, ill check em out. perhaps I will pause after cleaning up the topology and UV unwrap the guy and do some tutorials with SP and some project files for practice - baking normals from high poly to low poly and rough/metallic maps
is my next challenge

Is there anything I can do better in terms of 3d modelling? I know the verts aren't shown so its hard to make out any topo mistakes but im always looking for feedback where i fucked up and can do better
>>
File: cylinder collapse 2.webm (497 KB, 488x420)
497 KB
497 KB WEBM
I got a mesh to stay within the boundaries of an edge. Conceptually, both endpoints of the edge have a sort of "plane" attached. Those two planes are infinite in scale. The angle of the planes are perpendicular to the angle of the edge. So I manipulate the edge using bones. Which in turn moves the planes. And the cylinder between the planes can't escape. The cylinder can move freely between the two planes, but when its tries to go outside the bounds, it squashes against the plane instead.
>>
File: Screenshot (3).png (1.12 MB, 1361x1045)
1.12 MB
1.12 MB PNG
>>
Is it possible to get so good at 3d art that girls will want to suck your dick
>>
>>982707
kys weirdo
>>
>>982707
It would be better to be a painter. 3dcg may have the opposite effect.
>>
>>982708
No.
>>
>>982707
Yeah. I mean girls like artist in general. If you're actually good that is.(sometimes if you're not good)

It helps if you actually make good money off your art. Every girl is attracted to a successful man.

>>982709
If you're good enough, then your 3D work will compare to the artistry of a painting.
>>
>>982710
please?
>>
>>982705
Liminal
>>
>>982711
It doesnt impress women though. Sketching and acoustic guitar are the bj magnet art forms.
>>
>>982723
You have no idea. Women dig 3D. Just not the gun and survival gear autism you create. But a well sculpted fantasy character? Or a cute cartoon? Impressive.
>>
>>982712
Cope.
>>
>>982707
Nah, gotta do photography for that.
Free excuse to take photos of women in the nude, and get a bj on top.
Used to do it right after HS all the time.
Women like to look good, and they like looking at themselves. Photos are perfect for that.
>>
File: 1687836394904561.png (580 KB, 1431x706)
580 KB
580 KB PNG
The upper skeleton is mostly here...now to make the muscles
>>
>>982774
no s-curve in the spine.
>>
>>982789
wow anon, your eye for anatomical inaccuracies is 20-20.
>>
>>982774
are the legs meant to be as long as the arms?
>>
>>982636
Painting in blender is the worst thing you can do, it's literally a caveman experience
>>
>>982797
the guide im following says that for this head ratio (6 heads) this is correct, but if it looks off, i will change it
>>
>>982774
It's the skeleton of a Duende
>>
File: 0001-0400.webm (413 KB, 640x480)
413 KB
413 KB WEBM
The problem I think is that the people who wrote the iTaSC stuff intended it to be used to drive KUKA robots and I'm trying to use it off-label.
There should also be support for skeletal collision avoidance or something. I'll try it at some point.
>>
>>982774
It could just be the angle or something, but the torso seems too long in comparison to the legs. I'm not an anatomy expert though so I could just be talking out my tuchus.
>>
>>982774
This is so funny, you made my day anon
>>
>>982812
Anon, it's way off
Are you trying to make a lord of the rings dwarf?
>>
>>982824
I assumed yhis was the skeleton for some short stack coom character. Hence, its gonna need a major s-cirve to the spine more than it needs a better leg/torso ratio.
>>982812
For the record, on any normal human the pubic triangle points to the exsct midpoint of the human body. So the legs ought to be as ling as the torso and head combined. 3 heads on a 6 head figure. Very tall and very short people have legs that violate this proprotion and they look funny because of it.
>>
File: starter.jpg (94 KB, 1383x891)
94 KB
94 KB JPG
does this look like a good starting armor set for a game character? im still in the process of making it but how is it coming along?
>>
>>982842
It looks nice, like an old mortal Kombat character
>>
File: BigStretches.png (481 KB, 560x1120)
481 KB
481 KB PNG
I call it a Chi Wara
https://en.wikipedia.org/wiki/Chiwara
>>
File: ChiWaraWiP.png (1.85 MB, 1200x2000)
1.85 MB
1.85 MB PNG
>>982850
I will giver her clothes eventually.
>>
File: Carp.png (917 KB, 1200x1200)
917 KB
917 KB PNG
>>982851
I also made this fish (carp).
I wanted to practice animating and giving it a pixel texture so that I can incorporate those skills into my unique indie workflow and maybe compete with cris one day.
>>
>>982851
I don't know why but I really dig this.
>>
>>982851
>>982850
Really good abstraction of forms. On the side view, I kinda wish her horns had one more loop in the center to form more of an arc than a straight vertical (to tie together with the swooping curves of the body), but other than that, I think it looks really neat. It's kinda like an anime girl Kaminoan.
>>
File: City and bridge.jpg (1.62 MB, 1920x2160)
1.62 MB
1.62 MB JPG
>>982489
Some more work on the city. I added some highways in the background, and a bridge.
I know the bridge isn't too structurally sound, but the dude who wanted the project done, his logo is AYL, and I thought it'd be cool to take the design of the AY from his logo and use it as the supports for the bridge as kind of an Easter egg.
On the bright side, I've gotten a bit more comfortable with geo nodes. I was able to do the cables and structures in GN all by myself without help.

The bridge still needs materials and stuff, and I'm planning on adding some neon lights to it to accentuate the shapes of the logo a bit more. After that, it's onto some traffic, and finishing up the animation of the car so I can focus on building out the environment without fear of missing the deadline. Like as long as the animation is solid, if I'm still working on adding extra stuff to the environment until the last minute, I can just stop adding shit and render. That's the hope anyway.
>>
File: Screenshot.png (379 KB, 1207x1124)
379 KB
379 KB PNG
>>978944
I think the face looks better like this. also, some more texture work done
>>
File: untitled.jpg (2.82 MB, 3840x2160)
2.82 MB
2.82 MB JPG
>>982680
progress
>>
File: 1686161358694205.png (846 KB, 1090x1483)
846 KB
846 KB PNG
>>982822
>>982839

made some adjustments to the ratios of the bones. Also, here is my ref.

Now, I adjust the bindpose so the tets don't get stuck together and then start building the muscles
>>
>>982851
so this is the power of smooth shading
>>
>>982905
Miles ahead of my first car.
>>
>>982906
I'm not sure what you're trying to do here but rest assured I am following this with bafflement and awe.
>>
File: City2a.jpg (1.92 MB, 1920x1080)
1.92 MB
1.92 MB JPG
>>982884
Finished up the bridge, and I started working on some of the animations.
https://files.catbox.moe/ls3c1g.mp4
I've already got a regular anim of just cruising close to the water, so now I'm playing around with some extra bits. Got a little hop over the bridge, and a zero-g roll. For some reason I couldn't do the zero-g roll using just bezier curves and tilting them, so I just had to rotate the car with keyframes.

Even when you turn the resolution up on the curve and use as little points as possible, they're still jerky as fuck when you transition from one point to another. I've got no clue what Blender's problem is with curves but it's driving me up the wall.
I've tried smoothing them out as much as I can, but it's still somewhat visible at the top of the arcs. I'm using 3 points to make the arcs and it's still jerky. I guess I could try doing it with just 2 points so there's no midpoint to transition, but then you're just moving the issue to the points where it enters/exits the maneuvers.
I'll probably figure something out though.
>>
File: 1686821737675755.png (211 KB, 1435x690)
211 KB
211 KB PNG
using a signed distance field to generate the inner fat and outer fat (pictured is inner).

Next up is to get a breathing loop going with the abs / chest only with a subtissue
>>
>>983076
This is your brain when you don't just get a subdivided cube and sculpt from your mind and fix stuff that looks bad
>>
>>983076
Concerning.
>>
File: 0001-0500.webm (418 KB, 640x480)
418 KB
418 KB WEBM
So... The good news is that the iTaSC + Softbody system works as advertised. The bad news is that I really don't have the patience to make animations by hand and I want to find another way.
>>
>>983091
Yes ma'am
>>
>>982815
>>983091
are these really supposed to be mocap, i've never once seen any human being ever make any movement even remotely similar to this
>>
>>983107
They are in space
That's why the lack of gravity
>>
>>983091
>I really don't have the patience to make animations by hand

Based on those results it's really not 'patience' that is your limiting factor here.
>>
How's my topology?
>>
>>983076
>sdf
Aha thats why the butt cheeks are fusing
>>
>>983091
>really don't have the patience to make animations by hand and I want to find another way
You find a collaborator that wants to make animations but doesnt want to make rigs
>>
File: doublePole.jpg (48 KB, 774x723)
48 KB
48 KB JPG
>>983112
decent spacing and edgeflow but the marked 'double pole' you have here is one that can fuck you
and create pinching in the shading if you do facial animation.

You will have to have poles somewhere (vertex where more than 4 edges meet)
but having multiple poles on the same face is generally not a god idea
as the edge flow around that spot hill have tension in every direction.

Relocating poles to places where you have a natural break in how the shape flows is the more optimal approach.
I like to drop that pole that breaks how the eye loop and mouth loop lines up so it sits in the area
where you get a depression on your cheek if you suck it in.
>>
>>983116
Oh, so that's why I had shading issues near that area. Thanks! I still have trouble with moving poles around. I think if I tried to move the pole to the point you described I would fuck the whole geometry over.
>>
>>983107
That's not mocap. I've set up a number of poses by hand and I'm trying to interpolate between them. Like so >>982589 I've posted mocap based clips in previous threads.
>>983110
Because iTaSC does not support physics unfortunately. I wish they'd implemented a Bullet based ragdoll system instead, but that's what available. To add a Bullet based IK solver to Blender would be interesting, but it's too much work for me alone and the devs are busy with Geography Nodes and Grease Pencil.
>>983111
I've spent an hour or two on it and, for me, it's even too much. The point is to have something to fill in the gaps between mocap clips.
>>983115
That's not easy because 95% of people online seem to be stuck in the hell of the PUZZLE OF ALL QUADS RETOPOGUN TOPOLOGY and as such, they are completely useless.
I was planning to ask somebody I know IRL but then again, I don't think there's any money to be made and so I'll probably just continue to make experiments until I find something really good.
Specifically I would like to get out of Blender entirely and get into Unreal Engine because looking at the new features in 5.4, they may already have what I'm looking for.
>>
>it's not mocap, but it's mocap
more evidence of your retardation. let's say you somehow get this to work absolutely perfectly. it's still gonna look like absolute shit compared to a real animation
>>
>>983118
>That's not easy
Oh i know. Ive been looking for collaborators too.
>>
>>983122
https://www.youtube.com/watch?v=VjhOFXyscZ0
is what it looks like when I use mocap. Obviously in that case the movements are pre-recorded. I'm trying to find some sort of middle ground, if that makes sense to you.
>>
>>983116
I think I fixed it? I've had to sacrifice the backup mouth loop and a lot of the nose's geometry but from what I can see the shape hasn't been lost. It just feels so weird to remove so much geometry.
>>
File: 1684224355406855.webm (112 KB, 512x512)
112 KB
112 KB WEBM
Got basic breathing to work in simulation. Once I add the rest of the muscles on top of this they should all expand and contract nicely.
>>
File: an-animes.png (313 KB, 599x847)
313 KB
313 KB PNG
>>982905
listen...in the nicest way possible, it is worse thanmy first car..and i started using blender when i was 10...but at least you can handle fancy textures and materials.

this is not exatly wip anymore as i thinki ts complete..but my first time making an animes and i was only called the p-word by my friends. does it look like a decent animes character? i used the blue archive character as reference. i used automatic rig with rigify for it and it worked good enough
>>
>>982900
Hey that's a pretty nice Ashe, do you have anywhere i can follow your progress?
>>
>>983142
Looks like a pretty good chibi style model. Maybe more work on the ear since the head is the biggest part so it attracts attention but then i'm just nitpicking
>>
File: 6.jpg (821 KB, 1920x1080)
821 KB
821 KB JPG
>>983142
I did this without any tutorial or even a reference
After watching some videos I realized my way of doing it is way wrong
Admittedly initially I was trying to test my texturing and lighting
>>
>>983142
>i was only called the p-word by my friends
It's probably because it looks like a child.
Of course there's nothing wrong with making art of children as long as it's not meant to be erotic, but I've seen many people on the internet with a "non-erotic" interest in children turn put to be pedos all along.
They could also just be projecting. It could be because they themselves are pedos, or because they watch to many of Turkey Tom's "Losers. Creeps. Pedos." videos and have pedophiles on the mind constantly.
>>
>>983148
>I did this without any tutorial or even a reference
Cool but why
>>
>>983151
*turn out
*too many
>>
File: 14J.jpg (737 KB, 1971x1620)
737 KB
737 KB JPG
>>983152
That's how I always learned
try it the caveman way then watch someone do it the right way
>>
File: untitled2.png (1.38 MB, 1920x1080)
1.38 MB
1.38 MB PNG
>>983146
thank you, it feels good to have some actual feedback for a change.. personally i think the ears look fine, but that may be a question of taste. i dont know what to change in that regard..
>>983148
>I did this without any tutorial or even a reference
and it shows. but at the same time, i can appreciate that. i did learn many things doing things that way., but i suggest you start over, using a reference. use a chrysler LeBaron, it is boxy but has a distinct shape to get down, which is a nice starter piece to chew on.. alternatively, a lincoln mark 5
>>983151
i think you may be reading too much into it, my friends are morans who dont know how much work goes into creating a 3D model so their feedback is "looks good" or "looks not good".
>>
Oh god how did I fuck up this bad....
>>
>>983164
I've had the same problem in the past. I think this whole region in unsalvageable - in order to fix this you have to retopo the whole leg because it's a one big twisting spiral.
You should also be using the smooth and slide relax tools to make your quads more uniform.
>>
>>983140
>>
>>983173
She looks depressed beyond comprehension
>>
File: retop.png (403 KB, 858x664)
403 KB
403 KB PNG
>>983164
i don't know if this will help or if this is even a good topology for deformation yet as i need to get better at rigging first, but if it will help you what i find is that its easiest to create a loop that goes around the crease where the legs connect to the torso and goes around the butt, and another that goes around the upper thigh below the butt and then connect the areas in a way where the horizontal edges in between end going in perpendicular directions one towards the torso and the other towards the back and meets in the middle

again im no expert in this im kind of noob myself
>>
>>983174
I mean... she kinda is in the anime
>>
>>983177
I didn't notice it was lain without the hat
The cheekbones are too exaggerated
>>
>>983177
I'm a fan of the anime and I didn't see that was her, maybe the strand of hair only.
there's a lot of references for animation you can use
>>
>>983179
>>983180
Because all anime characters look exactly the same you delusional weebs
>>
>>983181
>seconded
>>
>>983179
I won't beat around the bush: I don't know what the fuck I'm doing with the hair. It looks like a bunch of blobs compared to the anime. I will, however, stand by those cheek bones - they shouldn't be as visible in the front with cel shading, but they should provide proper shape when looking from the side.
>>
File: reflectionVector.jpg (339 KB, 716x832)
339 KB
339 KB JPG
Tinkering with a custom PBR shader, the reflection vector visualized looks pretty trippy. Custom bent edited normals to fake internal refraction realtime on sparkly stone.
>>
File: 1700726266299691.webm (1.21 MB, 1920x1080)
1.21 MB
1.21 MB WEBM
Different breathing rates and diaphragm expansion sizes in the simulation. I can have it only affect certain areas with masking. I feel like I went too deep and big on the final breath.

Went for the double implant because I've heard about that before, in certain situations.

>>983114
yes, but I can adjust it with poly editing tools.. Here I didn't bother to make a crack because i will just have to do it all over again when I complete the rest of the arm.

Now, on to completing the arm.
>>
>>983181
>>983183

Show me a character that looks like Lain right now you filthy plebeian normies
>>
>>983187
is that some kind of tumor monster?
>>
>>983187
Anon, I'm scared
>>
>>983187
pritty fkkn cool
>>
>>983184
I have the Yoshitoshi Abe art book, you should try to download a pdf or something because it's full of illustrations you can follow.
Visual experiment lain and omnipresence in wired.
Thank me later

>>983181
Happens with almost every base mesh anon, it's like saying that every Pixar character base mesh looks the same. No shit Sherlock
>>
>>983210
Holy shit
>>
>>983187
pretty cool stuff ngl
>>
>>983210
I mean he is right that all anime characters looks the same, but art does imitate life so I think that outcome was probably unavoidable
>>
>>983218
You are wrong, grab any JoJo character and any sailor moon character, do they look the same?
What about berserk and lucky star?
Pokémon and ergo proxy?
Admit you had a shit take
>>
>>983145
I have posted my more recent models on Pixiv: https://www.pixiv.net/users/45000168
Probably not the best website to set up a portfolio, but it'll do for now.
>>
>>983164
Don't panic. Spirals are easily solveable. In edit mode, toggle on face selection. Then, select a column of faces from one end of the spiral to the other. In this case, it's from knee to hip. You only need to select a straight column, as in a line of face.
Delete your selection. Now, select the loops of what remains. Because you broke all the loops in the spiral, the your loop selection should stop at your newly formed gap. Making it very easy to see which end should connect to which end.

Now that you can see it, swap to edge selection, and, and select 2 edges that are supposed to connect and fill with a new face. Then, just fill up the remaining gap. Which should be as simple as holding the F key down.

It probably sounds a little confusing when read out like that. But when you figure it out, you'll see it's a simple 3 step process.
>>
>>983219
Right
And meanwhile western animation is dominated by bean mouths, and wherever it isnt its dominated by disney 3 point eyes.
>>
>>983241
>western
Jewish*
>>
>>983246
Go back to /pol/, your Palestinian friends need you there
>>
>>983187
This is interesting, what are you using to do the FE solving and how are you integrating it into Blender?
>>
I don't know if anyone in /wip/ does them, but where do you get images for packaging from (for material usage)? For instance something like a Coke battle label.
I've seen people say
>Just google "[product] label"
Or
>Just google "[product] wrapper"
and they come up with them no problem. Even for things like metal cans that can't be easily photographed or scanned. I try that and get literally fuck all apart from the product itself or some fake blank template shit that is only barely related to the product.
I think I spend more time looking for this kind of thing than I do actually modeling.
Is there some kind of database or something? I know >pinterest does have things like that on there, but they're really hit and miss depending on what you're looking for.
>>
>>983263
well you can always remake the lebels yourself in photoshop or gimp. They're mostly text and logo. You can definetely get a product logo in isolation, copy the background which is probably a gradient at its most complex. generic or actually product accurate nutrition facts and ingredient lists.
If you're really hellbent on an accurate high quality label texture remaking it from scratch could take less time than trying to find it.
>>
>>983239
Omg, thank you so much, I wasn't looking forward to redoing the whole region...
>>
>>983266
Nice, glad you found the advice useful. I remember going crosseyed the first time I had to fix a spiral. I didn't mention how to clean up any excess geometry that may have been left behind after the fix. But I trust you can figure that out on your own.
>>
>>983265
I have tried making them in the past, but they always come out shitty. Like I can't get the measurements quite right. Even when I "trace" it they seem off. Maybe I just haven't had enough practice at it. The real problem is when there's some kind of pattern or graphic that's hard to replicate.
Obviously it's time consuming and I'd rather have somewhere I could just download them from, but if that's not feasible I guess making them is the only way to go.
My alternative idea (at least for this) is I found a can label while I'm doing the bottle version. The two labels are really similar and it's only like the ml amount and a slight color difference on a small part of it. My guess is I'll be able to edit those and turn it into the bottle version without too much hassle.
Thanks.
>>
>doing retopology in mirror mode
>starts to affect me IRL
>wash only one hand thinking the other will be affected the same
>>
>>983256
>what are you using to do the FE solving
unfortunately, the only thing that I can tell you is that there are far too few users using what i am using so i feel like i am in a very very very VERY small pond and that its not available for purchase

>how are you integrating it into Blender?
>sim
>export ABC
>import multiple instances and spread out
>offset start times
>export ABC
>open in blender
>render with cycles / AGX
>>
>>983301
Anon... your lead is on the phone, they're wondering why you've created so many simulations recently without submitting anything for approval?
>>
>>983308
i'm not in the industry, i work a regular job, i just happen to have access to this solver
>>
File: 0001-0300.webm (385 KB, 640x480)
385 KB
385 KB WEBM
Nothing special. Still doing exercises with the iTaSC solver.
>>
>>983365
Can you explain to me how animation like this happens? I understand people who can't make smooth animation
with good transitions etc because that is legit hard but how do you end up with animation that flows like this
without even having key-poses that look legit anywhere thru-out the animation.

It suggest to me you're not doing 'pose to pose' but animate using 'straight ahead action' like claymation
but then if that is the case how does this happen? How do you look at what you've done and go "ok this looks fair so far, onwards!"
>>
>>983373
This >>983365 in particular happened because I was trying to fix a performance issue. I've just started to read the code and I don't know yet what parameter does what exactly.
>How do you look at what you've done and go "ok this looks fair so far, onwards!"
Software development doesn't happen all at once. There are stages to it and you're looking at the first stage where the system is expected to hit certain targets but doesn't care how. Every other additional behavior is disabled because I was trying to debug something in that stage.
>>
>>983373
He's not an animator or an artist, he's a programmer. They work differently to us.
>>983187
Let me guess, your dad makes car crashes simulations and you sneak out to his pc to make your weird Coomer simulations.
I've done some fun shit with Radioss but not this schizo
>>
>>983379
So this is animated via script? You're trying to do something akin to 'robotics' via a manually composed traditional AI type state machine?
>>
File: 0001-0300.webm (393 KB, 640x480)
393 KB
393 KB WEBM
>>983383
It's not a script. I'm trying to make the iTaSC solver work. It's in Blender but was abandoned long ago. It's an IK solver but different from the default one, it supports cyclic dependencies and other fancy stuff. You can do things like pull on a hand controller and the legs and shoulders may shift position automatically.
In the previous clip >>983365 I had disabled a parameter because reasons. Now I'm trying to re-enable it. I think it looks a little better but it's the same thing, just different interpolation.
Will it look good in the end? Research papers say I should be able to achieve Skibidi levels of animation eventually. Will I succeed? We'll see.
>>
Is there an easy way to flatten unwrapped uvs when the mesh isn't made up (almost) perfect squares? There's too much stretching when they get flattened but I don't like adding lots of cuts to make it more uniform because it increases the poly count massively when it's unnecessary other than for texturing. Is there no trick to this I can use?
>>
File: 0001-0300.webm (622 KB, 640x480)
622 KB
622 KB WEBM
So, the robotic motion thing was my fault apparently for not knowing anything about animation in Blender. To avoid it, the keyframes for the control armature (the blue segments) have to be set to Bezier interpolation with automatic handles and it goes away.
>>
File: 1686317618899867.png (29 KB, 154x430)
29 KB
29 KB PNG
>>983382
>Let me guess, your dad makes car crashes simulations and you sneak out to his pc to make your weird Coomer simulations.
>I've done some fun shit with Radioss but not this schizo

lol

You're way off and out of line and now just typing buzz words like schizo when your own hypothetical scenario is absurd and ridiculous.

----

Somebody brought it to my attention that my previous approach to the skeleton bones driving the muscles in a absolute relationship where the skeleton bones don't move beyond their initial keyframes and cant react to gravity is not being used anymore and that you in fact want to duplicate the skeleton and turn it into a tissue that is relatively quite stiff and does react to gravity and other tissues and attach it to the input skeleton so you in effect get a simulated bone result that your real tissues then attach to, making the animation rig truly procedural. So, its very important that I implement this as soon as possible after I finish my modelling and simple skeleton rigging.
>>
And you can see the difference with the default IK solver here >>983442 because there are no pole targets. A limb tries to match the position and orientation of its control bone, you can specify priorities, limits, there's a rudimentary self-collision avoidance thing. You can specify multiple targets within the same chain and the effects can propagate to other chains. All sorts of stuff.
But like with the Softbody simulator, there are bugs and the quality of the documentation ranges from nonsensical to misleading, so the only way to know what a parameter really does is to read the source code.
I don't know why they where abandoned and vandalized over time. I don't know the story behind it, but for myself, when I'll put it all together it should come out quite nice.
>>
>>983187
Fucking disgusting, get some help
>>
>>983445
Dude, you are beyond retarded, it's because what you do looks like an amputee a car crash victim.
>>
>>983187
This looks like someone squeezing meat through a toothpaste tube.
>>
>>983457
A creature made of chicken thighs and processed meat for cats
>>
>>983442
Can you show me how to do this?
I know a bit of html and CSS, probably it's the same thing programming in the language of blender
some anon told me about a pithon but I'm a bit scared of snakes.
>>
>>983454
>Dude, you are beyond retarded, it's because what you do looks like an amputee a car crash victim.

you cant even comprehend that you are in the WIP thread where I build my animation bit by bit and make a skin each time with a procedural signed distance field. Hell, I bet you don't even know what that means.
>>
File: 0001-0160.webm (198 KB, 640x480)
198 KB
198 KB WEBM
Today's exercise. I'm surprised it worked at all and I've got an idea of something I could program into iTaSC, that is a collider constraint.
What it would do is let a joint avoid meshes in the same way that the Softbody simulator does. But that's just an idea for now.
>>
File: 1.jpg (1.5 MB, 1920x1920)
1.5 MB
1.5 MB JPG
fucking shite
>>
File: QueenOfShemakhaSmall.webm (3.84 MB, 480x270)
3.84 MB
3.84 MB WEBM
>>982867
Thanks, I’m glad that you do dig it, even though you don’t know why.
At least for me, I’ve developed a bit of a thing for “abstract” looking women. It’s hard to say what flicked that switch, but the queen from the 1967 animation of The Golden Cockerel comes to mind. It’s not exactly the hottest character design, but I find something attractive about how her proportions are exaggerated, her cloudy hair, and the way she moves around. For my character, I like to do worldbuilding for the game/series I will try to make eventually, and wanted to have a sort of an “ancient aliens” thing in it.

>>982871
I’ll try to play around with the horns a little and see what I can do. My goal was for them to look like the horns from the sculptures in the Wikipedia article I linked (though those are just a few examples). I also wanted to give her a pair of swords that would match the curvature of her horns, so I’ll try to get all of those things to have some harmony with each other.

>>982913
I learned of the power of smooth shading recently after playing games like FFVII and LSD: Dream Emulator for the first time at the age of 23. Most people my age associate the retro game look with pixel texturing because many of us had never played genuine old games and are mostly exposed to faux retro styles. I was surprised to see how many old games used smooth shading, flat shading, and vertex painting. I’m unsure if modern smooth shading and old smooth shading work the same way, but I thought it was an interesting design choice to take inspiration from.
>>
File: CarpSwim.webm (247 KB, 900x900)
247 KB
247 KB WEBM
>>982852
Here it is, babby’s first animation
>>
>>983462
OMG! everyone lets call the news, the government, the president! this anon knows the magical ancient secret of SDF
What a retard, do you think that voxels with stored oriented distance its not a concept 90% of anons understand here?
>>
>>983187
i want to fuck it for some reason
>>
>>983507
Nice
>>
File: 0001-0160.webm (406 KB, 640x480)
406 KB
406 KB WEBM
Another point of view of >>983501
>>
>>983507
That's a fine fish
>>
File: julius_3.png (1.19 MB, 2000x900)
1.19 MB
1.19 MB PNG
I sculpted Julius Belmont mid-grand cross from this fight (at around 2:15 here) : https://www.youtube.com/watch?v=Pjyku6HuUIg

>>982900
Really cool Ashe, it does look better than it did before.
>>
>>983518
Now I see what you are doing. I just needed another point of view.
>>
>>983545
lol that's cool.
Have you tried inverting the normals and making the aura mesh envelope the character? It could allow you to have the same effect but from any angle.
>>
>>983518
So what's going on under the hood? Did you just rotate a controller to make the character turn 90 degrees and the iks moved on their own to accommodate?
>>
>>983577
>the iks moved on their own to accommodate
The purpose of the system is to fill-in the blanks of an incomplete specification of a pose. It tries to figure it out based on a number of parameters such a rotation limits, velocity, stiffness and so on.
>>
>>983582
it looks really unnatural, no one would move like that.

Also, the hand is clipping through the chair
>>
>>983576
Hey, that's a neat idea. I didn't think of that. Thanks!
>>
File: 5jfk.jpg (713 KB, 3840x3840)
713 KB
713 KB JPG
I don't even know how I would go about texturing this
>>
>>983604
Procedural textures anon, it's really easy
>>
>>983604
Is that a building or a dildo with a building shape?
I have to ask because there's some weird anons in this thread
>>
>>983623
It's a buildo
>>
File: 1715657859819263.webm (895 KB, 1920x1080)
895 KB
895 KB WEBM
Worked on simulated hands today.

For this really quick test, I did a two step bone system where I had keyframed bones and then duplicated those bones but made them into stiff tissues that could slide and react with other tissues, meaning they were able to adjust themselves and more importantly not penetrate when I roughly made a fist resulting in a truly dynamic rig. Apparently this approach saves a ton of hand keying and I'm going to be using it a lot more, in all aspects of the rig.
>>
File: 1628072964725.jpg (643 KB, 1700x1913)
643 KB
643 KB JPG
>>983604
Nah I'm not into lude shit
I saw this pic and wanted to make a similar building
>>
>>983632
You really are into disgusting shit
>>
y'all are schtinky
>>
>>983632
That's some nice surface interaction, how long roughly was the solve time for this? Also what sort of scheme are using for the element sizing (i.e. are the surface elements fine with coarser elements in the interior)?
>>
File: MM_Straight_0164.jpg (1.84 MB, 1920x1080)
1.84 MB
1.84 MB JPG
Don't you just love spending 12 hours rendering (and suffering countless blue screens due to a CPU clinging to its last vestiges of life), just to find out that your looping animation doesn't loop?
https://files.catbox.moe/k4equb.mp4

It was an easy fix (just maxing out the probability for buildings to have ads/neons), but I sure wish I would have checked the first and last frames in a render. I figured if they were the same object in an array, they'd use the same properties for the ads and shit since it's re-using the same object/mesh data, but I guess not.
On the bright side, the next 12 hours should go by smoothly now that I've found the magical balance point of threads I can use on my CPU to keep it from crashing while rendering. It's 3.... of 8. Poor lad is barely hangin in there. I'm rendering on a 4070ti, but that dying CPU (i7 7700k) is killin me. Hopefully I'll get a new one soon.
>>
File: SEG_sample.gif (1.27 MB, 256x256)
1.27 MB
1.27 MB GIF
Does anyone have experience making synthetic data for computer vision tasks? Pic related is a project I'm working on for image segmentation.
>>
>>983639
95 seconds to solve.

> Also what sort of scheme are using for the element sizing (i.e. are the surface elements fine with coarser elements in the interior)?
my outer surface is fine but my setup overall is not optimal right now. Looking at it i feel like I made things in general too fine and its slowing down my solve which would be a big problem if i had a complete rig ie the rest of the body and not just one hand
>>
File: 1700362479618370.webm (459 KB, 1920x1080)
459 KB
459 KB WEBM
One more test on the hand before I continue, showing the tissues of the hand basically solving themselves in response to various movements by the cylinder with the nonlinear wave deformer made possible by the 2 stage bone rig
>>
>>983661
>95 seconds to solve
To solve what? One frame?
>>
>>983674
You are into degenerate stuff anon
>>
>>982636
You show promise, cool guy. To texture your characters you can use either Substance as that other anon suggested or you can use Photoshop or Krita or GIMP or other 2D texturing programs:
https://www.youtube.com/shorts/N9pMg45OyYk
Texturing stuff with 2D image editing software is the old school way of doing things before Substance became the industry standard circa 2014 onwards.
With the old school workflow, the way you set up UVs is important. First, UV unwrap your whole character, then, in Blender's UV Editing workspace export the UV Layout (this generates a image of the UV layout your model will use; if you can't find it look it up on Google, I know where it is by muscle memory, I think it's under the "Image" tab somewhere in the left side of the UV Editing workspace). Afterwards, open said UV layout in Photoshop or other and start texturing (on a separate layer, obviously). Refer to the video I linked in regards to how to reload in Blender the image you saved within Photoshop or other 2D software of your choosing. You should really model and texture other lesser props before tackling characters which are relatively complex, but if you're going straight for characters then I suggest you start with, say, that flashlight your cool Stalker guy has on his helmet, so you get a feel on how to texture shit. Good luck bro, I liked Metro and Killzone too if either or both of those games were your inspiration, and if neither have anything to do with your dude, well I still like him anyway (no homo)

>>982852
>>983507
Cool fish anon

>>983187
Horrifying

>>983504
Good work, keep it up
>>
>>982699
Substance has some nice baking, but there are other programs that also allow you to generate texture maps (height, metallic, roughness, even normal) from an image texture. Look into Materialize if you're interested, it's free too, there are more texture map generation softwares but that's the one I know best
>>
>>983678
95 seconds for 43 frames which includes 10 initialization frames which weren't included in the webm
>>
File: 8high.jpg (1.37 MB, 3840x3840)
1.37 MB
1.37 MB JPG
>>983604
>>
>>983736
>That dildo looks painful
Nice building btw
>>
>>983733
Do you last 95 seconds while looking at your tub meat paste waifu?
Try kegel exercises
>>
>>983632
This is pretty cool, just don't post the nightmare raw chicken leg meatball pls
>>
>>983797
>ust don't post the nightmare raw chicken leg meatball pls
i have no idea what you're talking about. Grow up. Its cg.
>>
>>983632
They remind me of Worms' hands, or the type I would expect them to have. Cute.
>>
>>983808
>I have no idea what you're talking about
>Grow up
Why are you telling me to grow up if you don't even know what I'm talking about? Especially after I said that your work is cool
>>
>>983811
Hes a bit angry because he cant last more than 95 seconds while touching himself to 3d chicken legs.
>>
>>983820
Who can?
>>
File: 1688077821069907.webm (685 KB, 1798x1080)
685 KB
685 KB WEBM
>>983811
no worries, dood. I just want to stay true to myself, within reason.

---

Started on my hardest challenge yet - the face. I am dealing with some high frequency noise in the tissue, which I haven't started to debug. I have seen two approaches - building only the outer skin and cutting it up and making a blendshape control / animation rig or modelling all the interior facial pieces. I somewhat started the latter, as I used SDF in a "mr potato head" style along with booleans.

What I want to do next after fixing the noise is to add eyelids and blink controls
>>
>>983832
It looks like me when I drink coffee in the morning
>>
>>983832
>>983875
It looks like me when I'm shitting my guts out after I drink coffee in the morning
>>
>>983744
neither clever, nor funny
>>
>>983515
cope
>>
>>983890
no really kegels are good for you and will make your dick go super sayajin
>>
File: 1705351168591053.png (431 KB, 1093x577)
431 KB
431 KB PNG
Never used Blender before. Started today and completed that tutorial from CG Fast Track about making a Nordic magical sword.

Man, this thing is fun.
>>
File: 1684801833241806.png (764 KB, 1250x773)
764 KB
764 KB PNG
>>983926
Picture without the other stuffs.
>>
File: 1699990162731715.png (918 KB, 1215x760)
918 KB
918 KB PNG
>>983926
>>983928
Found out that the certain things were clipping.
>>
>>983890
I wasn't trying to be funny, I think you are a filthy degenerate
>>
>>983926
Thank god someone posting some normal shit.
I would gauge my eyes if I see another raw chicken meat Coomer simulation.
>>
File: 1693742907053303.webm (34 KB, 512x512)
34 KB
34 KB WEBM
Found out that only one way of working with the face will work - cutting it into pieces, creating simulation driven blendshapes, hooking it up to a blend shape rig, and then taking this animated blend shape result and doing another sim off it by attaching it onto the outer fat layer to get the real result. Modelling individual facial muscles and trying to get them to fire to move the face will just not work. Now my next step is to build my facial blendshape rig
>>
>>983187
And we thought that Cris was mentally ill, those were the nice days.
Raw chicken Coomer is in a different level
>>
>>983943
>My psychiatrist told me that I can't rub raw chicken meat in my cock
>He never said that I can't simulate it
>>
File: MM_CityPass_0372.jpg (1.43 MB, 1920x1080)
1.43 MB
1.43 MB JPG
>>983648
https://files.catbox.moe/nyq2os.mp4
Well, I finished it up. At least kinda. It was a "soft opening" type thing. I have some issues with it, notably on the transitions (and there's a visual glitch on the "city pass" section), but I had to cheat with the loop points since I was extremely limited in time and had to get things out the door in time for the dude's show which was just a few hours ago.
Now that that's all wrapped up though, I can polish it as much as I want and fix those issues since I don't have to worry too much about rendering 3 mins of footage in a few days.

The biggest issue really was just the buildings' neon trims not looping. Which is kind of my fault for using the object info's "random" output to drive the colors. In a normal array modifier, they count as the same object so they'd get the same color, but I used geo nodes to array them to save on memory since you need to realize instances to use modifiers on them, but obviously, instances count as their own separate objects. So they get a random color which fucks the looping. I only realized that was the issue well into the rendering process though, and my computer was already balancing on the edge of working and crashing, so realizing instances and using an array modifier probably would have been too much to handle.
Blog post over.
>>
File: Toronto Drive.webm (3.61 MB, 576x324)
3.61 MB
3.61 MB WEBM
>>983946
For reference, it was meant as kind of a spiritual successor to webm related which I did for the dude a few years back.
>>
ARRRRRRGHHHHH. i'M SO CLOSE! jUST A LITTLE MORE AND I GOT IT! wHY ARE GEOMETRY NODES SO HARD?!
>>
>>983943
The tube chicken Coomer strikes again!

>>983946
Nice anon.
>>983982
Because they were made by mentally ill people.

>>983939
If that's your first time using blender it's pretty good.
>>
>>983832
I can't imagine what you do to your chicken MC nuggets.
>>
File: isht.png (824 KB, 1857x888)
824 KB
824 KB PNG
im obsessed with istanbul
>>
>>983982
>>983985
>Because they were made by mentally ill people.
I don't think I can blame them this time. It might be a me fucking up. But I'm very close to coming up with a method of creating a 3D voronoi in geometry nodes. Something is just "off", and I don't know what it is. It works in my mind. But not working in nodes.

Essentially, I start by creating a point cloud. And then every point creates a new point at every other point. And with some index manipulation, I can create a sort of "list" for all the points that related to each other.
From here, I have to do a sort of triangulation. By that, I mean I'm just going to eliminate the points that don't belong each each list. This is the part I'm fucking up. My equation SHOULD work, but it's not.
After eliminating the unneeded points, then you just put a convex hull around each list of points. Which is easy to do with the repeat node.

I figured I could do the triangulation by setting up a repeat node, checking the dotproduct of every origin point against the other points. And then eliminating the points less than 0 dot. The first calculation appears to delete the proper points. But subsequent iterations of the loop go wrong somehow.
>>
File: 1715507580232598.webm (1.77 MB, 860x1080)
1.77 MB
1.77 MB WEBM
>>983985
Just finished this, but something looks off. It is too bright. I like things with higher contrast.
>>
>>983943
I can smell that video
>>
>>984022
Un ironically good, I know it's a tutorial but nice.
>>
File: MM_OverBridge_0308.jpg (1.81 MB, 1920x1080)
1.81 MB
1.81 MB JPG
>>983985
>Nice anon.
Thanks, honestly I'm just happy to finally get some sleep.

>>984022
Really cool. Lightning has always been something that I've struggled with. Maybe "struggled" isn't the right word, but I've never found a way to do it that wasn't a pain in the ass. I haven't tried in a while though, so maybe that's changed and it's super easy with geo nodes.
As for your brightness issues, you might try changing your color space's contrast or adjusting the exposure/gamma settings for a really quick fix, if you haven't already. Maybe having it backlit a bit instead of from the front/side might make the runes more visible and have them pop a bit more. Turning down the emission settings might work too.
Still, keep it up anon, hope you end up enjoying the creative process of making things. Looking forward to seeing some non-tutorial projects when you start spreading out a bit more.
>>
File: 1707601644771425.webm (2.34 MB, 860x1080)
2.34 MB
2.34 MB WEBM
>>984032
I started playing today with Blender without much idea of what i were doing and honestly just because i was bored but i had a lot of fun learning and playing around. I finally finished the tutorial.
Here's the result following the tutorial.
>>
>>984042
yeah looks cool. I'd up the tileing on those stone steps though. It looks like the details are too big compared to the environment.
>>
File: 1703855865370234.webm (1.71 MB, 860x1080)
1.71 MB
1.71 MB WEBM
>>984045
The rocks are too big, the character is clipping through the floor, and originally he was a giant when i added it to the program. I've met so many small problems despite having tutorial. Now seeing i see a lot of things wrong but i need to hold on my compulsion to fix it. Lost the whole day to this already.

Tomorrow going to try some different tutorial.
>>
Can someone explain why my weight paint isn't showing up under the mask texture menu when I'm trying to apply them to hair curves?

I named the weight map "FurWeight" but it's not showing up so I can;t apply it.
>>
possible game idea.
>>
>>984052
aren't painted weights stored in vertex attributes instead of textures?
>>
File: notexture.png (2.56 MB, 3840x2160)
2.56 MB
2.56 MB PNG
>>984010
ive finished it! my favorite (maybe) work so far
>>
>>983632
ebin ninja grab
>>
>>984080
I came to realize this a few minutes later and now I feel like a straight up fool
>>
>>984101
Here again. Is there a way to prevent this weird white sheen on the fur after it renders?
>>
>>984106
Also, I can't seem to add or adjust modifiers to the hair curves that were added with the add brush...is there a way to fix this?
>>
>>984087
ask you're a shelf why this:
>>984010
looks way better than your current render, and act accordingly
>>
>>984106
can you point out what the weird white sheen is?
it would be easier to see if the fur wasn't white.
and what shader are you using? principled hair or just regular principled?
>>
>>984117
Thing is the fur isn'y supposed to be white. It should be plugged into a texture map as the defuse color, but the fur is rendering white for some reason. And I'm using principle hair, yes.
>>
>>984119
i'm not a blenderlad but are you trying to colour your hair based on the colour of the surface they're growing from?
>>
>>984121
Yep, which Is why the white color is puzzling, since I connected a texture map to the hair curves that should mirror the surface. However, I think part of my problem might be that most of the hair guides were added through the "Add" brush after I already used the quick fur modifiers. I also noticed that changing the parameters on the modifiers or adding a new modifier doesn't seem to affect anything.
>>
>>984124
ah, hopefully someone can help you out.
i'm still not sure what sheen you're referring to btw. might be easier to see if you just unplug the tex for the time being and just set the tint manually.
that said, probably best to leave it till you get the tex working.
>>
File: 7522024.png (130 KB, 1364x784)
130 KB
130 KB PNG
rate my work
>>
File: 4213a.png (160 KB, 617x832)
160 KB
160 KB PNG
different angle
>>
File: gripper.png (277 KB, 637x744)
277 KB
277 KB PNG
>>
>>984141
>>984140
um sir, this is a bluw board. My children don't need to see your animated boobies. Please keep it to yourself.

Mm-thanks
>>
>>984140
>>984141
It's a good looking man with fake tits.
>>
>>984144
Jawline =/= man
>>
File: Screenshot_1.png (119 KB, 430x667)
119 KB
119 KB PNG
>>984140
>>
>>984149
>>984153
It's not just the jawline or some random tip about breasts. It's that you've never observed a woman in real life, you have no idea what it's supposed to look like, therefore all you're able to make are odd looking men.
>>
>>984160
>you've never observed a woman in real life
You're right I'm a tibetan monk who's never seen a woman
Just say what's so masculine about the face. Sharp features? Proportions?
>>
>>984161
Nothing man, this board is infested with mtf and they start screaming every time they see big tits.
>>984106
Check if you don't have a second material, sometimes hair takes a second material from the mesh
>>
>>984164
I agree that this board is infested with jannies but those are not big tits. Those are 10 years old implants in need of a surgical revision.
>>
File: 0001-0190.webm (252 KB, 640x480)
252 KB
252 KB WEBM
First attempt at animating female_01 from Microsoft Rocketbox with the iTaSC stuff. It works and potentially there are controls for everything including facial expressions but I'm not very motivated to work on it.
>>
>>984164
Another thing about jannies is that they don't feel threatened by fake tits or artificial features. Because they can buy those. Jannies get really pissed when you show them a perfectly normal female body because that's something they want but they'll never be able to have.
>>
File: 1692067947577475.webm (657 KB, 1920x1080)
657 KB
657 KB WEBM
Trying desperately to get self collisions on the eyelids working. I may be nearing a solution, as I've tried a lot of stuff already. Turns out simulating a face is very very difficult
>>
File: No power.jpg (212 KB, 1000x1000)
212 KB
212 KB JPG
>>984166
>>984170
>anons think the board is "infested" with jannies when titty posts, single-question, and off topic threads are here for months.
The only time something is removed is because another butthurt anon reported it.
>>
File: 06_15.png (1.88 MB, 1920x1080)
1.88 MB
1.88 MB PNG
>>982476
Any idea why the grease pencil shows through the mesh?
>>
>>984183
Anon, Hes not referring to those jannies.
>>
File: 1709702941494725.webm (608 KB, 1024x1024)
608 KB
608 KB WEBM
>>984182
better results. I believe i was over-constrained before. I can fix the side of the skull showing, should not be difficult. Gearing up to start creating a Real animation test combining all these simulation pieces
>>
>>984019
>>983982
I think I found the problem. But I don't know how to solve it. I need the repeat node to repeat the operation 6 times. Once for all 6 total points. And on each individual iteration, it selects specific points for deletion. The problem, is that it's not doing the operation 6 times differently. Rather, it's doing the same 6 times the same. Which just selects the same points 6 times.

I'm still unsure what I'm doing wrong. To the best of my knowledge, I have the repeat zone set up properly. But it's just choosing not add to the indices upon loop. Even then I have a math node that tells it to add 1 every loop, it just doesn't. It's like the math node is inert. But I know it's not, because the resulting output of the indices are increased by 6, proving that the repeat zone is counting in some way. But it's not counting where it matters: in the sample index node.

This is frustrating.
>>
File: trannies.jpg (76 KB, 735x616)
76 KB
76 KB JPG
>>984189
ohhhhhhhhhhhhhh, I gotcha.
>>
>>984202
Is this the MC nuggets solver?
>>
>>984212
This is bullshit. I wrote out all the points on paper, and checked every one of them. to ensure that they're actually turning on and off like they're supposed to. It all checks out. The problem really is just the indices not incrementing on each loop.
>>
File: 1716375069164.jpg (528 KB, 1485x890)
528 KB
528 KB JPG
>>
File: 3d triangulation.webm (1.54 MB, 544x472)
1.54 MB
1.54 MB WEBM
>>984239
ok. Somehow the nodes fixed themselves. I don't know what happened. I essentially cluttered the repeat zone section trying all kinds of different configurations. It was getting messy, so I cleaned it up back to the previous state it was in. Somehow when it was all plugged back in order, it worked. No clue what I did differently.

So now that it's actually working, all the points are triangulated to each other by line of sight. Meaning, each point in the cloud, sends a sort of "seed" point to its neighbors. Then, all the origins do a test against their seeds to know which points are "behind", and which are "in front". The "behind" seeds are all deleted. I don't have the technical vocabulary for the concept. But it gets results the same as Delaunay Triangulation, without using the actual algorithms of Delaunay Triangulation. Because that math was way too hard for me.

After the repeat node started working, the processing time suddenly shot up. It takes damn near a whole second for the node to process. Which makes working in nodes a pain, because anytime you plug something in, or take something out, or delete, or undo, you have to wait for it to finish processing. So it feels like the nodes are lagging.

I placed the frame of a box around the other 5 points to help convey the triangulation better. Altogether, there are only 13 origin points. The initial point to point seed calculation only creates 169 points. For some reason, Blender is taxed by doing minor calculations on only 169 points.
>>
>>984247
Like I told you, geo nodes are made by retards.
>>
>>984247
>ok. Somehow the nodes fixed themselves. I don't know what happened. I essentially cluttered the repeat zone section trying all kinds of different configurations. It was getting messy, so I cleaned it up back to the previous state it was in. Somehow when it was all plugged back in order, it worked. No clue what I did differently.
use git, unironically
>t. web developer
>>
>>984264
What's git, and what does it have to do with blender?
>>
>>984267
it has nothing to do with blender at all, but it has all to do with programming.
git is code version control, which based on what you are doing you pretty much need. Things as "revert current changes" are a command away with git. things like going to x version of your program, saving progress, etc.
It's like having a time machine of your code you can go back in forth, as long as you "commit" which is like ok I'm saving the current state now. So you have those commits and you can go back and forth with them.
>>
>>984269
Oh. I don't need any of those features. It's not as if I needed to revert to fix my nodes. I never worked in the first place. And I must have changed something during the cleanup to get it working properly. If I reverted automatically, then I wouldn't have stumbled into the fix.

Coding looks too difficult to learn anyway. And I have a goal in mind with blender. It's just taking me a long time to get there.
I'm a little closer to having a voronoi I can actually control with my own points. That's exciting. I just have to figure out how to turn these lines into actual cells.
>>
>>984270
>If I reverted automatically, then I wouldn't have stumbled into the fix.
yes, this bit was obvious, and git won't get you nothing you can't achieve having multiple save files... except you won't need all those saved files in your computer you'll have them in the gitlab cloud, for free, and you'll only have one local file that you'll be able to sync to any of those stored in the git cloud
this is common practice in the programming world
>
so you're not programming, is that right? because if you're not programming then yeah, it doesn't even make sense
>>
>>984271
Geometry nodes is not programming. And I don't have coding experience. So no, I don't see the point in having a git. The files I have saved are a small amount. My entire blender folder is only 1.8gb.
Blender does have a revert option btw. For the current session.
And an incremental save option.(Which replaced one of the shortcuts for bones, so I accidentally saved one of my files 17 times over before removing the shortcut)

If I really need to preserve something, I just drop it into my dropbox manually. Which has an automatic file renaming feature to date your files, making incremental saves possible. It would probably be safer to allow dropbox to save my files automatically. But I just don't like dropbox working the background like that. Call it superstition.
>>
File: 3d cursor.png (811 KB, 1897x1174)
811 KB
811 KB PNG
have a little 3d cursor model here. I want the black axis thing to always render in front of everything else in the scene, then I want the red and white circle to always be in front of the axis, as well as everything else in the scene. How can I best accomplish that? I think I need to make a rudimentary rig for the whole thing as well to make it fluidly look like the real cursor when animating but we'll get there when we get there. for now the circle is just a torus constrained to face the camera.
>>
File: file.png (938 KB, 1585x940)
938 KB
938 KB PNG
>2 months into self learning 3D
>first humanoid model to actually look close to what i picture in my head
>still not great at modeling clothes
>>
"que gigantes se refire, señor?"
"aquellos!"
>>
>>984289
I'd just put it on its own render pass to be honest. If it's always gonna be from that angle, I probably wouldn't even bother with that and just animate the image of it over the video after it's rendered.
>>
>>984310
Me gusta mucho.
>>
>>984310
Very wonderful
>>
>>984140
Nice bimbo anon
>>
File: 20193927.png (224 KB, 1055x788)
224 KB
224 KB PNG
surely she doesn't look like a tranny anymore
>>
>>984337
She looks like someone who would sneer at u and call you pathetic when you cum without her permission.
>>
File: 421234f44123.png (270 KB, 797x784)
270 KB
270 KB PNG
new hair
>>
>>984310
basado y quixotepastillado
>>
>>984310
I saw this post on reddit
>>
>>984304
Very good progress anon. Clothes are hella hard to model. The real reason we're all modelling nudes.
>>
>>984260
>>984247
Small update: I eventually discovered that the reason why the node was causing trouble, was for the fact that it was enabled for multiple meshes. One of which being a high density mesh. I didn't realize it before, because apparently, sometimes modifiers on objects can still be active, even when the object is hidden. So it's not enough to merely turn an object off. You have to also turn the modifier off in the stack. That, combined with the fact that I tend to copy objects a lot when I'm making major changes, so as to preserve the previous work, in case I want to return to it. So the node I was working on, was attached to an older object I had forgotten about.

This doesn't totally absolve Blender. But disabling the offending modifier fixed the lag in a huge way. Now the delay is like 3 frames, instead of a whole second. I almost don't notice it.
>>
I'm starting to think that blender is a program made by the devil.
>>
File: 0001-0500.webm (2.36 MB, 640x480)
2.36 MB
2.36 MB WEBM
Another little exercise with iTaSC and Rocketbox. Next year I'll get good at this.
>>
>>984572
would certainly explain a lot
>>
>>984585
this looks like something you'd see in the 70s
>>
Goblin man im making for something, dont know what for though.
>>
>>984614
an interactive desktop assistant
>>
File: 1705324530954194.webm (255 KB, 1024x1024)
255 KB
255 KB WEBM
Working on an eye animation system with the FEM solver. Basically it was not possible to do what i was doing before to any degree of quality, which is animating on the final rig. That just doesn't work.

Now I have the eye look left / look right / look up / look down animation from which I will grab the corresponding shapes at their apex and add them to my blendshape control rig. Next, I will add upper lid up / upper lid down, and lower lid the same shapes and make a control rig. In theory this should let me adjust the lids while looking in any direction such as UL. From that animation made with those controls I believe I can simulate again to get the final.
>>
File: ray march 01.webm (850 KB, 982x680)
850 KB
850 KB WEBM
Something in my brain clicked recently, and I'm able to comprehend nodes and vectors a little better than before. With that newfound insight, I made a raymarching node. It's still inaccurate, because I'm a little lost on how to properly check front and backs and adjust accordingly. But I'm getting there.

This is raymarching, not to be confused with raycasting. It's doable, because the sample nearest surface node is very fast. And the repeat node can do iterations of that. All had to do was make sure the ray was scaling properly.
>>
File: softbody 01.webm (390 KB, 782x676)
390 KB
390 KB WEBM
>>984680
I also managed to get started on a physics sim. It can't do anything yet though, so it looks very silly.
Just falling and smooshing. as it collides with the plane. The sliding is completely incidental, as I haven't implemented friction yet. The goal is to create something of a softbody.
I figured I needed some kind of raycasting to do it properly. But I didn't know how to raycast. Ray marching however seemed feasible.
>>
>>984682
Try a high poly prism.
>>
File: 1702216649515579.webm (301 KB, 1024x1024)
301 KB
301 KB WEBM
Its working, so far. Now I feel better moving forward.

The face rig is controlled by blendshapes that were extracted polys from the base mesh, modelled with simulation tools, simmed again to make the actual blendshape, hooked up to a control rig where animation takes place and then simmed one final time.
>>
>>984614
breeding big titty elves
>>
Damn this board is a scary place all I can say is Fuck Jannies.
>>
>>984694
I don't know why you want to see that. But ok. I created this ridiculously high poly cone.
I didn't work on the sim node since yesterday, so it's just mindlessly squishing things into the plane. I recorded it without frame skips, so you can see how slow it truly is.
>>
File: 11.jpg (1.18 MB, 3840x3840)
1.18 MB
1.18 MB JPG
>>983736
getting there
>>
>>984700
if you're aiming for realism, record yourself doing those movements from a few camera angles
and act accordingly
that's good, but that's not realism
>>
>>984717
pls make the railings tallers so I won't fall
>>
>>984700
>>984656
You are almost ready for your chicken mcnugget porn scene
>>
File: 1690325473432047.png (1.21 MB, 1580x944)
1.21 MB
1.21 MB PNG
>>982476
to product animators, how do you guys references for animations? Pinterest? i am struggling with getting any coherent ideas
>>984310
hella sick, are you that guy who uses da vinci for this paint effect or is a blender comp thing?
>>
>>984142
Feet are hell to make, you're a brave anon for giving them a try
>>
>>984776
Feet are easy [spoiler]as long as you're modeling furries[/spoiler]
>>
File: 1710547537921022.webm (300 KB, 1024x1024)
300 KB
300 KB WEBM
>>984719
I'm aiming for "good enough to be somewhat believable, but not absolute realism"

>>984731
changed settings. Turns out there is a different type of attachment that i almost never use for some reason that makes sense here, sliding type. I put some strength into that attachment between the eyeball and the skin in the eye area. Its subtle, but I think it makes a difference
>>
>>984802
It's definitely because of the cartoonish skull, but somehow it's slowly crossing the line from shitty programmer "art" into something that could actually be considered real art. It's starting to feel like some practical effects and makeup like you'd see in a David Cronenberg movie, in a good way. I think the waxy, almost vinyl/latex material for the skin is what's selling it.
I wasn't invested before, but it's grown on me. Interested to see how this saga ends.
>>
>>984802
What solver are you using?
>>
File: Cuts.jpg (131 KB, 862x522)
131 KB
131 KB JPG
This might sound stupid but if I subdivide a cube, and let's say the dimensions are X=1, Y=1, Z=0.5, am I better off dissolving every other loop that runs horizontally along the Z faces so all of the faces will have similar size squares or will it not really matter?
That's harder to explain than I thought it would be. Should I do B or should I just leave it as A? You'll have to forgive the pixelization btw. It got all fucked up when I resized it in based paint. I think you can get the basic gist still.
>>
>>984814
You need acompañamiento psicologico as much as he does.
>>
>>984814
>It's starting to feel like some practical effects and makeup like you'd see in a David Cronenberg movie, in a good way.
Holy shit CGI practical effects is kino, you are onto something
>>
>>984822
It really depends on what you need the extra geometry for. If you're hard surface modelling, then I would say to delete all excess loops no matter the side. This way or that way, if the loops aren't defining a new angle, or shoring up an edge, then they're not needed. But if you need the extra vertices for say physics simulations, and the block is going to twist and wobble and rotate, then I would say go with B. Your image doesn't show the proper dimensions. But I'm imagining a squat box. No need to load it up with excessive loop cuts. Get it to match roughly the same density as the other sides. So yeah, B makes sense.
>>
>>984815
direct FEM solver, non Houdini
>>
>>984852
I don't really need a lot of it tbf. I thought there was more detail there than there is, but it's just a few buttons. I should probably really be doing it as three separate sections, but there's a hole that overlaps two of the sections and trying to put that in on two separate parts that line up would be a nightmare.
It's for a walkman. I didn't post the exact dimensions (which I have) and just kept it simple instead with the 1,1,0.5. I always think things will be easy or fun to do and every time I get fucking filtered as soon as I start. I've seen things people have made that look realistic and have a minimal amount of geometry. When I try the same thing I end up with mesh more dense than a blackhole and it still looks shit.
Didn't mean to start ranting like a faggot there btw, thanks, I'll keep in mind what you've said for the future.
>>
>>984835
I passed on Spanish class in HS, and took German instead, so I don't speak cris.
>>
File: 12.jpg (1.86 MB, 3840x3840)
1.86 MB
1.86 MB JPG
>>984717
I'm done for the day.
I thought I'd add some ivy and other little stuff but I'm too tired.
It's also freaking hot and my PC has been rendering all day
>>
>>984863
Cute!
>>
>>984863
>mfw I'm at the top floor and I'm gonna fall and die
>;_;
>>
>>983946
This is sick. Looks great, would love to see a fly through/into the city
>>
>>983946
bruuuh give that shit volumetric trail light
>>
>>982905
I dig the hyper-blocky aesthetic. It evokes minecraft, but without being ugly as sin. Like a good cartoon, it doesn't look realistic but it looks very attractive. Badass, anon.
>>
>>983141
What engine?
>>
>>984202
You're pulling skin from areas that aren't connected in a real body. Its causing you to underestimate how much skin is actually present. I think your collision will work just fine when you have tendon/muscle constraints worked in. (Particularly the triangle between the muscles that move the eyebrows and the connective tissue around the eyes - they work together to store the eyelid like a garage door between blinks.)
>>
>>984802
At least put a band of thick semi-rigid tissue at the base of each eyelid... You have a duct that runs from outside to inside that irrigates the eye. That's why you can look around WITHOUT dragging your eyelid with the eye. Its lubricated, and your tear ducts are semi-rigid so the whole eyelid is supported. Your simulations are already spot on, you just need to continue adding biological details. Fucking great work mate.
>>
>>984873
>would love to see a fly through/into the city
I'm working on that now, actually. While I had "finished it up" for the concert deadline, the plan was always to polish and add extra shots and stuff.
I'm working on a transitional shot where it flies through the A's down the bridge, and up into the city, and then making a few loops in the city of the car flying around.

>>984878
>volumetric trail light
I had that in there before, with trails coming from the jets using the Mesh Trails addon, but because the car is moving so fast, any setup I used for it just ended up making the trail comically long and coming off the frame. I didn't have the time to get something working in time, but it's also on the list of things to add in this second pass.
>>
File: loftarea.png (916 KB, 1871x1211)
916 KB
916 KB PNG
the loft area of a room i'm working on
i'm going to simplify the ropes near the bed
the room is designed for a fairly small character, which is why some of the furniture seems a little disproportionate
>>
>>984904
Clean geometry. Nice style. Probably going to look great when lit. I dig it.
I'm not even sure how you can simplified the rope. It already looks fairly optimal. I guess you could combine the knots in a simplified ball, and then texture them. But idk, seems unnecessary.
>>
File: loftarea2.png (953 KB, 1924x1116)
953 KB
953 KB PNG
>>984907
Thanks. i want to do something similar in style to little witch academia, ill probably mess around with some of the mesh's a bit later on to give them a bit more character.
As for the ropes, i'm thinking of reducing their number and routing them a little differently so that they aren't as cluttered. they are just there to stop my character from falling 3 meters if she were to roll out of bed
>>
gahdayum why is it so difficult to make a model look as close to a drawing as possible
>>
File: one and.jpg (22 KB, 743x799)
22 KB
22 KB JPG
No textures yet, but so far so good.
It's a half giant barbarian character.
I don't think I like the hair but I don't know how to do hair cards without completely redoing the toon shader I'm using.
>>
>>984922
Cuz the drawing is out of perspective. You could make your mesh loom exsctly like the drawing from the one angle though
>>
almost done highpolys, then onto baking normals
>substance license expired
fuggg, marmoset it is then
>>
>>984961
why is the piece above the trigger flat shadded?
>>
>>985140
That particular part is in low poly/lowest subdivision. Zbrush doesn't display normal smoothing (or normals for that matter) so you will see tesselation until the geo is made more complex. After subdividing a few times, more geometry is added between preexisting geo, allowing curved surfaces to smooth out while edges that I wish to keep sharp have been creased (adding edgeloops very close to those pre-existing edges)
>>
>>985145
wait you sculpted a hard surface gun? wtf
>>
test for a collision map I'm working on

les vrais savent
>>
File: mute-city-Recovered-min.png (2.32 MB, 6880x3488)
2.32 MB
2.32 MB PNG
>>
>>985158
Naw that'd be crazy. I modeled a lowpoly in maya and then brought it into Zbrush after. However, there are absolutely modelers out there who do hard surface in Zbrush, all the way from primitives to sculpting to retopology, but alot of their stuff is more exploratory or conceptual, rather than modeling to spec
>>
>>985320
what did you sculpt in zbrush then?
>>
>>985347
I used zmodeler to crease and uncrease edgeloops on the fly while I moved the model up in subdivisions, allowing me to not only account for edges/corners/curves that have different levels of softness from how the gun is forged and then milled in certain areas after, but also to keep my low poly and high poly close enough to each other to minimize retopologizing afterwards. I also manually softened up some areas afterwards with smoothing, specifically where pinching happened due to subdividing areas with triangles
>>
>>985364
zmodeler is cringe



[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.