[ 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: 589 KB, 1024x1024, 1705375109850784.jpg [View same] [iqdb] [saucenao] [google]
972982 No.972982 [Reply] [Original]

/wip/ - Works In Progress - constructive criticism edition
Post your work-in-progress projects, recently finished projects, or things you'd like critiqued here.

Previous thread: >>968417

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

>> No.972986
File: 193 KB, 591x813, heater stove.png [View same] [iqdb] [saucenao] [google]
972986

Really proud of this wee heater stove I made today, designed it myself too

>> No.972988
File: 963 KB, 3036x1147, kris_sculpt.jpg [View same] [iqdb] [saucenao] [google]
972988

Okay, what do you think. I don't think it's very important to copy the shape of the reference 1 to 1, it's important that the face is appealing. Maybe eyes too big?

>> No.972991
File: 333 KB, 1280x960, 0001.png [View same] [iqdb] [saucenao] [google]
972991

https://files.catbox.moe/1mogh9.mp4

>> No.972997

>>972986
Looks nice anon, glad to see that you put sharp edges on your low poly model
>>972988
Yes the eye needs to be slightly smaller
>>972988
Nice

>> No.973009

>>972982
What better way to show the downfall of /3/ than this OP. Not that it had any "golden age" to begin with.

>> No.973023
File: 346 KB, 620x468, surface projection test 01.webm [View same] [iqdb] [saucenao] [google]
973023

Working on a way to project all points of a mesh onto to all faces of the same mesh using geometry nodes. Technically, I can already do it. But I don't want to merely sample the center of every face. I want an actual projection.

My math here is still off. I have two balls that are supposed to be projecting on each other, but they aren't. For some reason, the math only manages to hit the plane I set up as a base. Which is to be expected. The balls *should* project onto the plane. But they should be projecting onto each other AND the plane.

Anyway, I'm finally beginning to see expected results. Note how when the projection leaves the plane, the points disappear. That's what the difficult math is doing. It's detecting if a point is inside or outside of any given face, and only showing the points that are inside. Deleting points that are outside. And note how when the projection crosses from one face to another, it keeps shape? Technically, the points projected to on face are getting deleted, and the points projected to the other face are coming out of deletion. But it happens so seamlessly, it's as if all the points are the same.

I still have to figure out what's wrong with the math. But just getting this far is exciting.
Some points are flickering. But I think that's just float point error. If a point calculates a teeny tiny decimal greater than the face area, it will get deleted. So I'll have to leave some room for error.

>> No.973024

>>973023
It’s called adding a mirror to the flat surface. Something Blender can’t do and requires plugin from some rando that might have viruses.

>> No.973025

>>973024
You got any more details for that? I'd search it up myself, but I don't even know where to begin.

>> No.973028

>>973023
Don't you have somebody near you in real life who could perhaps tutor you on some math at least to a middle school level?

>> No.973034
File: 26 KB, 711x640, area within polygon.png [View same] [iqdb] [saucenao] [google]
973034

>>973028
Rude. I'm learning what's necessary to learn. Some stuff I knew and had forgotten, because I didn't have a use for math until recently. Other stuff my school didn't even teach.

Just yesterday, I figured out how to test the area inside of a polygon on my own. And then googled it, just to confirm it was possible. I can't find the link to the site now. But I had to re-learn how to find the area of a triangle. Which I hadn't needed to know in FOREVER. Then I had to figure out how to tell stupid blender how to do it in nodes. Which apparently, is still wrong. I have the idea down on paper.(pic related) I just don't know how to tell blender to do it right.

>> No.973035

>>973034
You've got to find somebody in real life to tutor you in math to at least middle school level. I don't care if you think it's rude. It is what it is.

>> No.973036

>>973035
Whatever, man. I guess /3/ has its fair share of trolls too. There's really no escaping it on the chans.
I don't want to hear another word from you, unless you can accurately identify the problem I'm having.

>> No.973037

>>973036
>I figured out how to test the area inside of a polygon on my own. And then googled it, just to confirm it was possible.
Why didn't you google the correct solution directly?

>> No.973041

>>973037
I wasn't entirely sure how the point needed to be tested against the polygon. I only had a vague idea that perhaps if I get some kind of relationship between the normal and the position, then something might happen. After some thought, I figured that because the vertices of the polygon could make any shape, the point would need to be tested against every vertex. After some more though, the area method came to me.

I worked it out on paper first, because it was stimulating. And because searching google comes up with a lot of fluff results. Most solutions involved ray tracing, and then counting how many intersections the rays has. An odd amount of intersections proves it's inside. And even amount of intersections proves it's outside. As far as I'm aware, Blender's raycast node isn't capable of accumulating the number of intersections per ray. It only detects the first hit and stops there.

Some other answers determine whether a point is on a plane, but not within the bounds of a polygon. I didn't need to know that.

>> No.973042

>>973034
>I have the idea down on paper.(pic related) I just don't know how to tell blender to do it right.

https://www.youtube.com/watch?v=goayI0oaRg4

>> No.973043

>>973042
Thanks. I'll give it a watch later. I'm relaxing for now. Did as much as I could handle for one day.
I am using some of the nodes in the thumbnail to get face corners. I wasn't aware an offset face cord node exists. That might be just what I need. Perhaps it will bring some order to the randomness of the index.

>> No.973048

>>972997
>Looks nice anon, glad to see that you put sharp edges on your low poly model
Thanks man, sharp edges add a lot to a model so that was some good advice the anons gave me

>>973009
You're taking it too seriously anon-kun, it's just a funny pic

>> No.973051

>>973036
The problem is you, just make the flat model into a mirror. Stop using Blender, nothing in blender works as expected.

>> No.973053

>>973036
If you feel like you're dealing with a troll the last thing you should do is responding to him

>> No.973055 [DELETED] 

>>972982
Who the fuck doesn't use the trad image for wip thread. As the author of picrel I disavow this shit.
Fuck you op you are a faggot.

>> No.973058 [DELETED] 

>>972982
>>973055
Also nice picrel :)

>> No.973076

>>972991
Nice

>> No.973086 [DELETED] 
File: 1.80 MB, 600x450, ezgif-2-0145ed8604.gif [View same] [iqdb] [saucenao] [google]
973086

>>972991
>https://files.catbox.moe/1mogh9.mp4
Dude use a free gif converter for our convenience ffs.

>> No.973093
File: 411 KB, 1280x960, 1movie.webm [View same] [iqdb] [saucenao] [google]
973093

>>973076
>>973086
Thanks anon, the truth is that i felt that transcoding it would make it lose some fidelity or blur it. I think i can transcode it to webm to post it.

>> No.973094

>>973086
>pic
abslutely love it, except the hotel sign, that is -3.5 points

>> No.973095

SNEED

>> No.973096

>>973048
Not that guy but hobbit hole is board culture, anon. You don't just cast it away in favor of 'a funny pic'. That's not cool.

>> No.973103

>>973048
>You're taking it too seriously anon-kun, it's just a funny pic
Fuck off, tourist.

>> No.973109
File: 420 KB, 1441x907, drake_01.png [View same] [iqdb] [saucenao] [google]
973109

>>972793

>> No.973110
File: 545 KB, 1920x1080, 2024-02-05 22-21-36.webm [View same] [iqdb] [saucenao] [google]
973110

baked cloth sim Im making for a vrchat avatar.

>> No.973112
File: 356 KB, 828x1149, image_2024-02-06_013911953.png [View same] [iqdb] [saucenao] [google]
973112

Working on a Dakota Johnson likeness sculpt as practice.
First time doing a likeness sculpt, any tips or tricks people got?

>> No.973129
File: 1.83 MB, 640x360, surface projection test 02.webm [View same] [iqdb] [saucenao] [google]
973129

>>973042
>>973043
I'm stuck. See, I can distribute points onto all face corners, and then offset them with their neighbors. But I can't figure out how to get only 1 point per face. I think if I could do that, then I could figure out the rest. I just need 1 point per face, in coordination with face corners. Bah, I'm so lost, I don't even know how to explain it properly.

>> No.973130
File: 529 KB, 729x871, kn7.png [View same] [iqdb] [saucenao] [google]
973130

Working on a Dwarf Fortress-inspired goblin kidnapper model that I plan to 3D print, paint, sell STLs, and possibly cast in metal as part of a pack of several DF-related creatures. Currently working on his skirts, which are proving really tricky to sculpt folds into. Everything is still very rough, still need to make hands, sleeves, finish the arms, tweak the pose, so on and so on.

Question for those with /tg/ hobbies: does it look, so far, like it would be fun to paint? Could it take edge highlights okay?

>> No.973131

>>973130
I'm trying to hide that lowercase g in plain sight on the model. It might be a bit obvious right on the forehead, but I kind of like it as a weird primitive bit of jewelry or something. In the ascii version of the game, goblins are shown as just "g".

>> No.973132
File: 545 KB, 753x866, kn9.png [View same] [iqdb] [saucenao] [google]
973132

>>973131
Back view, sorry for spam fellas, I'll stop here. The skull on the bag is getting replaced, btw.

>> No.973133

>>973131
Shame he doesn't have enough skin exposed to where you could have it be a tattoo/brand on his bicep or forearm or something. Though that might be too subtle.
Or just change the letter from a clear as fuck modern font, to something more "runic" where you make the lowercase "g" with straight lines.

>> No.973137
File: 16 KB, 1125x553, IMG_1088.jpg [View same] [iqdb] [saucenao] [google]
973137

>>973133
I like the idea of using it as a brand too, and I was even considering using it as a boolean to make g-shaped wrinkles for his elbow. Thanks for the idea, anon :)
Pic rel is the exact font from the game, I tried to recreate it, and though the runic idea is awesome I think I'm still going to try and keep it as close to this as possible somehow.

>> No.973138 [DELETED] 
File: 1.34 MB, 1465x963, GreatUncleanOneRotigus.png [View same] [iqdb] [saucenao] [google]
973138

>>973130
it looks good, the letter could be wider too maybe, I'll be working on Rotigus next, I know I can find a fuckton of these online but this is mine, I finished it a couple of years ago but I lost the progress in a computer reset.
And you can post on >>>/tg/91814425 too. I've posted progress there lots of times, the op says
>Share your printed minis, terrain, print fails, 3d modeling

>> No.973139

>>973130
I hope tarn and zach sue you

>> No.973140
File: 3.61 MB, 600x338, turntable.gif [View same] [iqdb] [saucenao] [google]
973140

>>973133
cool concept, could be made of twigs or something but that would be tedious

>>973093
i like the lighting on the wall of the hotel and on the building interiors

architecture student, work mainly in rhino with some CAD and BIM. trying to make low detail rhino models to use for graphic design purposes. wondering if anything looks odd or jumps out as awkward, not in the graphic design (posted in /gd/) but rather the models themselves

>> No.973142
File: 1.06 MB, 1862x1007, file.png [View same] [iqdb] [saucenao] [google]
973142

>>972986
Nice comfy. Now colour it.

Figuring out how to rig stuff. This is gonna look really cool when it can actually move.

>> No.973144
File: 2.45 MB, 3210x1080, 動物1.png [View same] [iqdb] [saucenao] [google]
973144

>>973142

>> No.973145

>>973139
Right, because that's the way to promote their game and establish good will for an extremely niche title. Something tells me they're not as much of a faggot as you are.
Besides, it's not like they own the rights to the letter "g" on a command prompt. Even basing it off the Steam version, the graphic for the snatchers is pretty different. Without the dude mentioning its origins, you'd have no clue what (if anything) the dude's goblin was from.

>> No.973146

>>973145
I will be sure to report you to them once I see your model for sale, I will be sure to include this thread of you saying slurs

>> No.973147

>>973139
>>973146
NTA, but why does this anger you? Why are you even acting as a janny? Actually, BELOW a Janny because you weren't even asked.

>> No.973148
File: 149 KB, 825x671, calc-point-on-face.png [View same] [iqdb] [saucenao] [google]
973148

>>973129
>I just need 1 point per face, in coordination with face corners.

... pic related, calculate the vector A to b, B to c and C to a, take this three vectors to find the point where they intersect, thats what you looking for, i guess. when you do it on the UV level, its easier because blender does the recalculation into the 3d space - when you move a, b and c along the vectors A B, B C and C A, you can define any position on the face.

https://www.youtube.com/watch?v=Nj5a4blpcAk

this vid shows you how you can handle the UV stuff in geonodes ... i made this in python some time ago for a tool i made, its a little bit tricky!

>> No.973156
File: 2.12 MB, 640x360, surface projection test 03.webm [View same] [iqdb] [saucenao] [google]
973156

>>973148
Thanks for the help. Though, I don't really understand the UV stuff. And I'm not sure you understood my issue. A failure to explain properly on my part.

But I woke up, and after a few minutes, I was suddenly able to figure it out on me own. Sleep really does work wonders. The faces are split to make all the points visible without overlapping each other. This *should* help in doing the next step. Maybe. Hopefully.

>> No.973169
File: 477 KB, 700x865, kn8.png [View same] [iqdb] [saucenao] [google]
973169

>>973146
A couple of things for you to understand. First of all, >>973145 is not me. Everything he said other than the slur is verifiably correct, though. No one owns the rights to the concept of a goblin or to the letter g, thank goodness. Second, this project will be pretty much unrecognizable when it's finished, and I am not so desperate or so stupid that I will slap a Dwarf Fortress title and logo on it when it's done. I've seen a lot of blatant copyright infringement on STL sites, and this, even as it stands, is a far cry from even the least dubious of those. If you do manage to get a hold of Tarn and Zach, be sure to pass along my well wishes and let them know I'm super excited for April and for the myth and magic update further down the road.

>>973138
Your model looks totally amazing, and a lot cleaner than mine currently is too. I have been learning 3DCG since 2020, and I still often have trouble taking my projects to those last steps that make it look complete. Well done chum, and thanks for the heads up about /tg/, I asked about it there and was told to gtfo.

>> No.973170

>>973138
Oh shit I'm stupid, I just reread your post and realized what you said about losing your progress. I'm so sorry to hear that, anon, it looks like it was coming along super well. That shit is devastating. I back up everything in two different places because I'm so paranoid about this happening.

>> No.973176
File: 635 KB, 523x694, incredibly sad.png [View same] [iqdb] [saucenao] [google]
973176

I think I'll have to bite the bullet and learn Substance Painter to texture my characters. I can comfortably make texes in Photoshop for simple props, but with characters it ain't that simple, I mean I could but it's cumbersome, I'm better off using SP there rather than Blender or Photoshop.
I just don't want my characters to have that generic ultra perfect look that most characters textured with the software seem to share, you know? The ArtStation portfolio look, i.e. technically impressive but very standardized, interchangeable with plenty of other similar artists and therefore forgettable. It's almost frustrating to see all those technically amazing works on ArtStation (when I used to go there, I dropped it after the AI thing) that despite the high level of skill are all so homogenized and indistinct. I dunno man, it's really paradoxical

>> No.973177
File: 1.57 MB, 1280x720, rig webm.webm [View same] [iqdb] [saucenao] [google]
973177

>>973142
Just a lil' bit more.

>>973176
>The ArtStation portfolio look, i.e. technically impressive but very standardized, interchangeable with plenty of other similar artists and therefore forgettable.
Aren't those done by artists wanting to be hired by Riot or some other video game company?? Are you surprised they are trying to be appealing as possible? Whether something ends up looking "generic" like you say depends on the art direction. Just... don't go towards that direction?

Though, take my opinion with a grain of salt I'm still a complete amateur and only really use Substance to hand paint.

>> No.973178

>>973176
You just have to make it different. The reason they all look the same is because they all use the preset materials. There is nothing stoping you from hand painting everything.

>> No.973179 [DELETED] 

>>973169
>>973170
Yours looks great too, yes the model was fine, it had all the weapons and all that but the skin wasn't that great, anyhow I sometimes revisit old sculpts so this time I'll go with him.
I sculpted the Magnus the Red the last thread too, I hope this doesn't take that much time.

>> No.973181

>>973023

... iam this guy >>973148 ... sry, havnt read your initial post, i gues you could archive your effect with the raycast node! no big math needed! you cast a ray from each point of your first ball into the direction of the plane, when the rays hits geometry it returns the positions, the "mirror ball", which you can use to raycast again to the center of the first ball, which gives you the projected points of the mirror ball on the first ball ...

https://www.youtube.com/watch?v=ZCCQXoJoIK4

>> No.973183

>>973176
>characters textured with the software seem to share
You are retarded, it's quite literally photoshop for 3d if you strip it to most basic functions, you can just handpaint with layers or do whatever the fuck you want. I can predict that your "original character donut steal" designs are dogshit btw

>> No.973184
File: 1.88 MB, 818x568, surface projection test 04.webm [View same] [iqdb] [saucenao] [google]
973184

>>973181
Thanks again. But the raycast node just won't serve my purposes. I've done a bunch of experimenting with it already. With raycasting, a single object struggles to cast onto itself. If the ball and plane were two separate objects, then what you propose would be perfectly reasonable. But because I'm making a single object cast onto itself, it takes a bunch of fiddling with the ray positions. Offsetting them slightly. And then that offset shows problems when the mesh is moved. As tight positions make the offset sometimes touch the face, or neighboring faces accidentally. That just doesn't sit right with me.

>>973156
Well anyway, I got it working finally. Not only was I failing the face corner matter, I was actually failing the math portion too. I thought I set up the math in nodes properly, but I didn't. So I reworked it. I was also able to solve the float point errors by setting the compare not to "not equal", which comes with a little epsilon option set to 0.001. That stopped the flickering.

I see why you guys call it mirroring now. Because every plane mirrors the vertices independently. For some reason, I didn't think it would work like that. It took me so long to figure this out. I forgot what I was going to do with it. Probably some physics stuff.

>> No.973186

>>973109
Is this the updated hair?

>> No.973187

>>973184
>With raycasting, a single object struggles to cast onto itself.
>But because I'm making a single object cast onto itself,

...kek, i guess you trapped yourself in the "technological irrational complicated" trap, and lost the focus on the final effect/result you want to archive, and when you snap out of it, it hits you like a bong, and trust me, i know this trap very well - the good thing about such traps is, once you freed yourself from it, you are smarter than before!

https://www.youtube.com/watch?v=NhloXEPL3Vc

>> No.973191

>>973187
Dunno what you mean by "technological irrational complicated trap". But I jested when I said I forgot what I was making this for. I still know what my desired effect is. I just don't know exactly the path required to get there. I'm learning that path through trial and error. I believe raycasting could be a valuable vehicle to get to my destination. But not as the geometry node currently functions. This mirror thing has a better chance at helping me.

>> No.973193

>>973191
>I still know what my desired effect is.

...can you write down the endeffect you want to archive in one sentence, ...the way its done, maybe scares you, isnt the one you forcefully wants to go!!!

>> No.973194

>>973193
>...can you write down the endeffect you want to archive in one sentence
Boobs go *bwomp bwomp* in real time.

>> No.973198

>>973146
Oh no, what will I do? You'll send my random post sticking up for an anon, to a couple of game devs? What a travesty. And you'll include me calling you a faggot, and double nigger?
I'll never recover from this.

>> No.973205

>>973198
I am going to send the screenshot to my uncle, he collects Warhammer minis so watch out, he can give you a bad review on his 200 plus follower YouTube channel

>> No.973208

Do we have a gamedev board on 4chan? It supposed to be on /g/, but i don't feel like discussing a dungeon level layout or question about quaternion rotation would be appreciated there. But it wouldn't appreciated here either. I am in a bit of confusion.
Also i do stl models and print them on resin 350$ printer, i sell them and gift to friends, fuck GW and Wizards of the coast

>> No.973210
File: 357 KB, 1024x576, Engines.png [View same] [iqdb] [saucenao] [google]
973210

>>973208
The /agdg/ general on /vg/ is supposed to be the place to go for gamedev, but at this point those threads mostly consist of shitposting, whining, and crabbing at a level that puts /ic/ to shame. I don’t know what happened to that general, but now it seems like a thread with 750 replies will only have 5-10 people actually posting their progress.

The gamedev threads on /v/ might be a better option even though there’s less activity, because at least people there actually make progress. Finding a thread to post in can be a problem though, because people don’t seem to make these threads as often anymore, and the only one who does it regularly is a schizo who’s obsessed with 9-year-olds and boogers and his threads die quickly. The threads usually have OP pics like the one I have attached, but if you want to get one of these threads started, you will have to keep your thread on life support for a few hours to make sure it doesn’t die too soon. People don’t want to post in the thread if it doesn’t seem to have enough activity.

>> No.973211

>>973208
>quaternion rotation
/g/ doesn't know what quaternion rotation means /3/ doesn't either
Try/v/

>> No.973217

>>973208
There is the general on /vg/ like the above anon said but it's mostly shitposting so you have to read the entire thread to find progress. Occasionally, you will spot a gamedev thread pop up on /v/ which is of higher quality but when it gets created is relatively random. Though I suspect some anons made a commitment to make one every week or month.

>> No.973231

>>973208
try >>>/g/gedg/ it's not super popular but there are decent people there and the discussions are usually technically oriented

>> No.973235

how do i start.
Im familiar with Blender enough for my 3d Printing, but I wanna make pictures with 3D models, and then animations. What links in the resource list would help me? I'm at a lost to start.

>> No.973300
File: 1.65 MB, 1701x910, image_2024-02-08_001013289.png [View same] [iqdb] [saucenao] [google]
973300

>>973112
I think it's getting closer, this is very very difficult.

>> No.973325
File: 976 KB, 1441x1931, drake_02.png [View same] [iqdb] [saucenao] [google]
973325

>>973109

>>973186
I don't know.

>> No.973329
File: 1.99 MB, 1920x1080, Hill_Wip3.png [View same] [iqdb] [saucenao] [google]
973329

Still early wip.
Cities are a fuckin pain.

>> No.973360 [DELETED] 
File: 1.69 MB, 1586x1026, GUOR-01.png [View same] [iqdb] [saucenao] [google]
973360

So I've picked the basic sculpt and I've blocked the face, I'll try to finish the basic shape of the body this weekend and to add the details after, should I make different objects for the fat, the muslces, the pustules and all that? I think I'll do the other variation only if I don't get tired of it.

>> No.973377
File: 201 KB, 2560x2560, bottle.jpg [View same] [iqdb] [saucenao] [google]
973377

How do my textures and lighting always look dogshit? I think glass is difficult to get right anyway, or for me it is. Mine always seem to look significantly worse than everyone else's.

>> No.973386
File: 54 KB, 1362x595, fighterpilotpewpewpew.jpg [View same] [iqdb] [saucenao] [google]
973386

>> No.973392

>>972986
split normals better

>> No.973397

>>973055
>>973096
>>973103
If the hobbit hole image were used routinely then it would lose its luster. By creating scarcity, i.e. by limiting its appearance, whenever it is used again it is appreciated more, instead of it being given for granted. The hobbit hole will return, and when it will you'll appreciate it more. You can't appreciate the spring without the winter, you can't appreciate the day without the night

>> No.973406

>>973397
>If the hobbit hole image were used routinely then it would lose its luster.
It's been used routinely for over a decade and people still call out its absence, then it hasn't "lost its luster".
This is just another case of a faggot zoomer tourist thinking he's doing us a favor, simply because he hasn't bothered to pick up on how things "function" (using that term extremely fuckin loosely) here.
It's not just an image, it's a tradition.

>> No.973423

>>973397
Not how it works.
When I look for the WIP thread I look for the hobbit hole image, it has been the WIP image for decades, I thought that we didn't had a new WIP thread until i search for it.

also you don't change the stop sign just because it's overused, That would be stupid. This is about conventions and not about how scarce and image is.
I don't care if you are a newfag, learn our traditions. If you want to "make it about you" because you are an attention whore just be a namefag like the rest.

>> No.973424

>>973377
Because your hdri and lights aren't that good, like I tell my students, a decent 3d model extremely well lit is magnitude times better than a perfect model with ok to generic lighting.
Remember that we don't see the polycount, we see the final render and it's our job to make the render as good as we can
What I used to teach to my students is to search for references and copy the light
You for example can chech product photography reference and copy the light try to match the temperature etc.

>> No.973433
File: 928 KB, 951x643, broyoujustpostedcringe.png [View same] [iqdb] [saucenao] [google]
973433

>>973406
>>973423
Sorry, but whatever is put in the OP image is up to the thread creator's discretion.
Next time you go make a new thread when the bump limit is reached rather than waiting for someone else to do it and then complain when he does

>> No.973438
File: 1.03 MB, 1280x720, Anim.webm [View same] [iqdb] [saucenao] [google]
973438

>>973329
Added some highways. Shit's a bit flickery due to the small size, sample count, and aggressive denoising.

>> No.973439

>>973433
>Sorry, but whatever is put in the OP image is up to the thread creator's discretion
That may be true, but if I make a fucking thread with the hobbit right now nobody will post here, because everyone would assume it's the WIP thread. Only a new fag like you and op would not know that.

>> No.973442

>>973438
Looks nice anon l, light up the car just a little bit

>> No.973444

>>973360
Do you 3d print them? So you have a final print of your work?
>>973392
They are decent
>>973433
I didn't even realized that it was a WIP thread until I searched it, i have to disagree the hobbit hole has a function, to make the thread visible quickly.

>> No.973445

>>973130
Looks good and I really want to paint it, sometimes new minis are full of details and look like shit when printed small, I like your level of detail
Are you going to sell the stl file?

>> No.973446

>>973442
Yeah that's the plan.
The environment up there isn't finished yet. I'll have some lamps and the like up there to brighten things up a tad, and to add some cool light to contrast the warm of the city.

>> No.973450 [DELETED] 

>>973444
>Do you 3d print them?
I print some of them, others are just digital sculpts, I think I bought the resin printer just as excuse to sculpt more, yes that one wasn't printed.

>> No.973466

>>973424
My hdri wasn't working properly and I don't know why so I just got rid of it. Usually, depending on the model, I'll try a few different ones to find which lighting fits it best. This time the lighting stayed exactly the same with every single hdri.
I don't know if it's the sample size I use or not but the viewport seems dull as fuck compared to when it's rendered so it's difficult to judge how close or far off I am with the lighting without rendering it too.
I'll see if I can improve it based on what you've said. Thanks.

>> No.973475
File: 727 KB, 1280x960, 1movie.webm [View same] [iqdb] [saucenao] [google]
973475

>>973093
Follow up of this one, save some details i would say i'm 90% done with it.

>> No.973478

>>973475
Mind giving a bit of an explanation on how to achieve this pixellated look? Is it just downscaling to nearest?
In >>973093 you can see some kind of dithering on the light the street lamps are emitting. Is that an encoding artifact, or intended?

>> No.973480

>>973478
It is the other way around, i render in low res, (this one is originally 640x480) and just upscale 2x with nearest neighbor. It's rendered in eevee at just one sample, because further samples add antialiasing and kills the pixellated look.

That dithering effect was intentional, using a checkered texture to influence a mix shader node with a emission/transparent shaders. I'm not sure if i'm gonna leave it flat like this or with the fake dithering.

>> No.973486 [DELETED] 

>>973475
Love the yellow light reflection. Its coming up together pretty nicely.

>> No.973528

>>973446
Looks nice I really like the 80 s feel
>>973450
If you print it post it here anon
>>973466
>My hdri wasn't working properly and I don't know why so I just got rid of it.
Are you using Maya?

>> No.973529

>>973377
you have no roughness variation and little in the way of normal information either.
'imperfection' maps are a category unto themselves e.g.:
https://quixel.com/megascans/home?category=imperfection

and you usually want at least one of these plugged into roughness. even if you're doing a nice product render and don't want it to look grimy, you need something like wiping residue, which you pipe into something that lets you control the range and you make it barely noticeable.

similar think for using imperfection maps in bump/normals. a little goes a very, very long way. you can also use these maps for other things. e.g. a scratch map plugged into normals will get you a scratch, but most of the time that same map should also affect roughness a bit and depending on the surface it might also affect the diffuse/albedo.

also glass doesn't blow perfectly so you should have a big, non-rough noise going into normals to create a bit of wavyness in the reflections

also, when lighting very shiny stuff the reflections and environment are just as important as the materials. you can get nice light maps for softboxes and things like that so you don't solid blocks of white light in the reflection (you can also use ramps/gradients).

>> No.973530

>>973529
oh and i can't say this enough - keep those normal map values low especially when you're using offline renderers.

>> No.973537
File: 713 KB, 988x784, medkit n bandage.png [View same] [iqdb] [saucenao] [google]
973537

Modeled and textured my own custom medkit

>> No.973538
File: 266 KB, 1400x1000, 9.jpg [View same] [iqdb] [saucenao] [google]
973538

what now?

>> No.973540

>>973438
I think the cars in the streets could be improved, and the lights on the buildings feel too syncronized

>> No.973543

>>973528
>Are you using Maya?
Nope. Blender.

>>973529
>>973530
Thanks I'll try adding those. I've had quixel megascans for a while and always forget to use it. Seems like I should break it out.
It felt like every reference I looked up was completely different too so that didn't help. Even that little circle thing took me fucking ages because it seemed different in every pic. Mine is way too small, but I had a hard enough job doing it as it and just left it. Pisses me off every time I look at it though. There was also supposed to be another sticker on there, but I couldn't get that right neither and left it out entirely.
I'll have a go at adding those variations tonight and see if I can make an improvement to it. I've looked up some other references, mainly for the lighting, and I think I can sort of see where I was going wrong. Or where I was mainly going wrong at least.
Thanks for the help dude.

>> No.973549
File: 381 KB, 844x1500, 00.jpg [View same] [iqdb] [saucenao] [google]
973549

>> No.973558
File: 111 KB, 1280x720, 16.jpg [View same] [iqdb] [saucenao] [google]
973558

Just modeled those three buildings, i haven't decided the composition or anything else beyond that.

>> No.973576
File: 1.67 MB, 6967x786, sc_150.jpg [View same] [iqdb] [saucenao] [google]
973576

After improving the head and taking a break to deal with other problems, I realized that I was not in a position to to take long time to polish the head to my satisfaction. So I decided to ask Machine God for guidance and passed picture of my sculpt through stable diffusion.

I think examples 1 and 4 are the best results, and soon I will start adjusting my sculpt according to example 1 to see how it turns out. What do you think?

>> No.973587

>>972988
>>973576
Welp, I'm not really a fan of the concept you're basing it on in the first place. But I'd say 4 is the best results. The original sculpt over emphasizes his nose bridge. 1 and 2 makes his nose too narrow. 4 keeps the width, while smoothing out the bridge region. Which helps a whole lot.

Your original sculpt is also too "realistic" to pull off the cartoony look. The lips are especially damaging to the cartoon style. You should just smooth out all detail in the lip region. Except perhaps the underlip to capture the under lip shadow. Even that should be shallow though.

But the other thing that's preventing overall appeal is the lack of cheek fat. His face is too sculpted. It does look like you changed that somewhat between now and before, so maybe it's not all that bad now, but perhaps a little fuller cheeks will help anyway.

And personally, I think the shorter hair is hella gay, yo. But that seems to be what you're shooting for, so I won't tell you to fix it. I just don't like it.

>> No.973591

>>973587

Thanks for feedback.

I have not yet decided exactly how the hairstyle will look, I will have to experiment when the head is ready.

>> No.973600

>>973540
Improved how? They're a noisy mess right now. Honestly surprised you can see them.
>lights on the buildings feel too syncronized
I 100% agree with you there. First thing I fixed after I saw this.

>Looks nice I really like the 80 s feel
Thanks! I've been trying to push it to that era a bit without doing the obvious things. I was hoping the orange glow of the city might do the trick.

>> No.973643

>>973028
Today, I learned what sin, cos, and tan are. Well, the basic concepts anyway. Using this site.
https://www.mathsisfun.com/sine-cosine-tangent.html
And doing this exercise: https://www.mathsisfun.com/sine-graph-exercise.html
That's something I never learned in school. I've always wondered what they were, because they're on every calculator. But just never looked into it until now. I think it's going to be important to know how this stuff works. Especially for the next step in my work. I got something planned.

>> No.973652 [DELETED] 
File: 1.58 MB, 1010x800, GUOR-02.webm [View same] [iqdb] [saucenao] [google]
973652

Basic shape done, I've gone with the mouth in the belly because it seems easier, no weapons or cloth yet.

>> No.973659

>>973300
more

>> No.973665
File: 220 KB, 754x868, ll.jpg [View same] [iqdb] [saucenao] [google]
973665

Started my second ever blender sculpt / character today. Any critique or advice welcome.

>> No.973687
File: 1.11 MB, 1100x1674, 3.png [View same] [iqdb] [saucenao] [google]
973687

Any feedback on this?

>> No.973693
File: 368 KB, 1302x1074, 1700295649424511.jpg [View same] [iqdb] [saucenao] [google]
973693

just wanna say, fusion360 is amazing
and all that is only 380 triangles, it literally does retopo for you
I'm never using anything else for hard surface again

>> No.973703

>>973665
Advice - use references

>> No.973716
File: 332 KB, 135x270, knife1_Spin.gif [View same] [iqdb] [saucenao] [google]
973716

Here's a kitchen knife. Now to make some bread or something for it to slice.

>> No.973720
File: 888 KB, 1600x1600, Knives.jpg [View same] [iqdb] [saucenao] [google]
973720

>>973716
That ain't a bread knife, gonna crush the bread with that. Give it some fish or meat to cut.
Still, lookin good. Knives are fun to make, though the pointy bit on the end confuses me sometimes.

>> No.973724
File: 82 KB, 1434x704, thigh.jpg [View same] [iqdb] [saucenao] [google]
973724

>>973386
Man, this going to be a bitch to texture.

>> No.973736

>>973724
Nah it'll be fine, low poly assets tend to be relatively easy to texture. Cool heli by the way, diggin' the overall design

>> No.973748
File: 161 KB, 1024x500, unnamed.jpg [View same] [iqdb] [saucenao] [google]
973748

>>973736
thanks, wish i could say i designed it.
i was upset there werent many good pictures of the ships so i made my own based on what little there was.

>> No.973752
File: 2.08 MB, 1219x827, sos.png [View same] [iqdb] [saucenao] [google]
973752

Today I textured my armchair for my unique and original post-apocalyptic setting

>> No.973753
File: 875 KB, 947x814, sos flat.png [View same] [iqdb] [saucenao] [google]
973753

>>973752
Flat version

>> No.973762 [DELETED] 
File: 205 KB, 1078x700, 15215.jpg [View same] [iqdb] [saucenao] [google]
973762

is she hot?

>> No.973766
File: 372 KB, 1600x859, 3_3.jpg [View same] [iqdb] [saucenao] [google]
973766

is she hot?

>> No.973772

>>973766
Kinda reminds me of Laura from Outgang

>> No.973777

>>973766
She can call me misogynistic White male how much she wants

>> No.973790

>>973766
She looks like the female elves from Morrowind if the game was made with modern technology

>> No.973799

>>973772
kek

>> No.973800

>>973687
>>973665
Looks like Cotton from King of the Hill. I don't even know what to say about it. Are you going for realism? It looks bizarre.

>>973752
I hope you plan to do a bump map for it.

>>973766
No. But she might be, if not for all the irregularities.

>> No.973806
File: 2.13 MB, 800x450, 845.gif [View same] [iqdb] [saucenao] [google]
973806

>>973800
Not going for realism lol, but I'm really new to modeling so it would probably look like shit if I would.
Thanks for the feedback.

>> No.973821

>>973693
What's Fusion360? How much does it cost?

>> No.973826

>>973665
Do not use red matcap.
Red matcap gives permanent sanity damage.

>> No.973829

>>973821
it's a lightweight CAD application
and it's free, just need an autodesk account

>> No.973832

>>973826
I used a red matcap when I was sculpting knuckles from sonic and it didn't made me mad, it just made me a furry

>> No.973849
File: 930 KB, 1441x1938, drake_03.png [View same] [iqdb] [saucenao] [google]
973849

>>973325

>> No.973851

>>973826
How and why? Some youtube tutorial guy told me it's good...

Then what should I use?

>>973832
I don't wanna be a furry

>> No.973855

>>973800
>I hope you plan to do a bump map for it.
Sure, once the entire texture image is filled (that there is an atlas, other objects will use it too) I will run it through Materialize to generate those sweet sweet maps

>> No.973885
File: 3.21 MB, 8506x7127, Mech_overview_small.jpg [View same] [iqdb] [saucenao] [google]
973885

Reworked the Textures on my Mech so it does not look so blunt anymore, what do you guys thinK?

>> No.973892

>>973885
Can you repost the order one?

>> No.973893

>>973851
>I don't wanna be a furry
Then don't use the red matcap

>> No.973913

>>973885
Looks better to me. Nicely done, anon.

>> No.973918

>>973893
Which one then?

>> No.973925

>>973885
Good shit anon, looks good

>> No.973931

>>973885
I like it, but presentation could use a bit of work.
I'd recommend checking out noax.design on artstation and try to copy his lighting setups/quality target. really nice design though

>> No.973932

>>973885
Nice. Textures look way better.

>> No.973987

>>973918
Be a man and use the grey one

>> No.973994

I'll ask here since you guys are actually doing shit.
I learned modeling through the Blender official playlist teaching the tools and shortcuts but I haven't learned modeling by following along someone's project.
I have some time now to train game asset creation but when I open Blender, I stare at the screen for ten minutes and end up closing it. I have no idea what to do and how to do it. It's like the blank page syndrome, but the modeling version.
Any tips for someone in my situation?

>> No.973995 [DELETED] 

>>973994
Just download some references first and try to make that in 3d, guns, cars, robots whatever, idk the fuck are you interested in ....

>> No.973998
File: 324 KB, 991x720, 1703395513012549.png [View same] [iqdb] [saucenao] [google]
973998

>> No.974004

>>973994
Step 1:Donut
Step 2: Make an unguided project utilising things you don't learn making a donut(I'm making a scene with cloth, smoke and liquid sims atm)
>Step 3: Download reference images of Waifu/Husbando and model
Is my plan at least

>> No.974039

>>973994
best advice:
>>973995

Do this anon, what kind of games do you want to make? Try making environments with props? Then you have multiple smaller pieces that can all fit together into one broader scene.

>> No.974073
File: 855 KB, 1920x1080, Hill_Wip7.1.jpg [View same] [iqdb] [saucenao] [google]
974073

>>973438
Added some planes n shit (there's like 3 in the shot, but one is clear enough to see at this res/samples), some Hollywood spotlights, and a shitty, temporary, fill light for the car until I buckle down and add some lamps and stuff.
I think up next I'll try to fill out the environment on top of the hill a bit more. I've got quite a bit up there already, but none of it is visible from this angle (since there'll be multiple angles in the anim).

>> No.974130
File: 519 KB, 1893x1317, sc_174.jpg [View same] [iqdb] [saucenao] [google]
974130

I think i polished the turd to "good enough" so I can soon go to the next steps.

What do you think? How Is 6.5 heads body proportion looks?

>> No.974133

>>974130
Looks good anon, I think you should work on the ears and the sides profile, especially the neck

>> No.974182
File: 1.37 MB, 800x450, 667.gif [View same] [iqdb] [saucenao] [google]
974182

>> No.974185

>>974182
actual art, anon, keep going, good shit

>> No.974190

>>974182
Harrowing. Something about this draws me to it. Also the low frame rate adds a nice touch to it

>> No.974209

>>974182
Point and Click adventure cutscene/10

>> No.974210

>>974182
I don't really get it but the important thing is that at least you seem to be having fun with it which is good.

>> No.974225
File: 1.97 MB, 2882x1941, drake_04.png [View same] [iqdb] [saucenao] [google]
974225

>>973849

>> No.974297

>>974185
>>974190
Thanks for the feedback. I didn't make the animations myself, I used ready-made ones from mixamo. Just playing with the camera, lights and the model I made.

>>974209
Sounds cool, I wish some day I'd be able to make a game.

>>974210
Yes that is true lol

>> No.974362
File: 2.40 MB, 1920x1080, final.png [View same] [iqdb] [saucenao] [google]
974362

feedback appreciated thank you!

>> No.974372
File: 1.75 MB, 1000x1198, 9.9.png [View same] [iqdb] [saucenao] [google]
974372

>> No.974375
File: 667 KB, 1173x820, dogman.png [View same] [iqdb] [saucenao] [google]
974375

Very quick very dirty skinning test for my dog guy character

>> No.974393
File: 2.53 MB, 1000x1676, 9...png [View same] [iqdb] [saucenao] [google]
974393

First time making clothing for a character

>> No.974408

>>974393
If you're interested in making clothes you might like Marvelous Designer

>> No.974415

>>974408
I'll have to check that out, thanks.

>> No.974463

>>974408
Let him learn to walk first, jesus.

>> No.974475
File: 848 KB, 1235x697, hob3.png [View same] [iqdb] [saucenao] [google]
974475

critique plx

>> No.974476

>>973438
beautiful work, i love the low res flickering actually

>> No.974477

>>973885
impressive

>> No.974478

>>973177
what is your polycount for the entire spider?

>> No.974482

>>974475
More facets
More insets at the joints

>> No.974497

Did a cave environment using Unreal Engine 5, the main purpose was to learn Niagara to make the waterfall.

https://www.youtube.com/watch?v=yoy3EqBOo1c

>> No.974498
File: 868 KB, 1920x1080, Hill_Wip8.1.jpg [View same] [iqdb] [saucenao] [google]
974498

>>974476
Thanks! I'm happy it's finally coming together. It was one of those projects that just wasn't shaping up, and working on it was a slog.

As for the flickering, I definitely like it, but I'm gonna do a bit to make it look less like a low sample count and aggressive denoising.
Basically make it look a bit more like heat haze, so there's a tiny bit of shimmer, but not a ton. I think that's how it'd be irl anyway.

>> No.974530
File: 1.74 MB, 1920x1080, ontiteld.png [View same] [iqdb] [saucenao] [google]
974530

>> No.974531

>>974530
a quick low effort sketch on i made on blender and photoshop

>> No.974540

>>974362
compared to the car he seems a bit small.

>> No.974544
File: 645 KB, 1440x3088, Tower of babel wallpaper.jpg [View same] [iqdb] [saucenao] [google]
974544

I'm making an animated 'live wallpaper' for my phone.
Any advice on composition is appreciated.
I'm aiming for the tower of babel, I want to show the pinnacle of human achievement and the utmost societal order and purity. I think I'll do another wallpaper with everything in ruins. Maybe I'll have the wallpaper switch to that at night time or something.

>> No.974546

>>974540
its a yugeee merc so yeah

>> No.974579
File: 88 KB, 785x791, buff.jpg [View same] [iqdb] [saucenao] [google]
974579

>> No.974582
File: 705 KB, 1440x3088, Tower of babel wallpaper 2.jpg [View same] [iqdb] [saucenao] [google]
974582

>>974544
latest
I've got to do the cliffs, textures and clouds still.
I think the composition is good, some room for improvement around the sea on the left and the foreground building. Not sure what to put there though

>> No.974583

>>974582
nice and comfy. the green hills look a bit odd though

>> No.974592
File: 136 KB, 566x1176, Screenshot_20240218-145218.jpg [View same] [iqdb] [saucenao] [google]
974592

>>974579
Very masculine frame.
Women's bones a very different. In addition to a wider pelvis women's shins are angled so that they are straight while her femurs are pointed together.

>> No.974593
File: 150 KB, 700x817, differences_between_male_and_female_skeleton.jpg [View same] [iqdb] [saucenao] [google]
974593

>>974592
Maybe this will make yhe point a bit better

>> No.974594 [DELETED] 
File: 79 KB, 641x849, anitrastafaricolumn.jpg [View same] [iqdb] [saucenao] [google]
974594

>>974582
Nice. Arch here would be nice to use the 'would've been properly historical' assyrian columns.
>Not sure what to put there though
Some hanging gardens or an orgy would be proper.

>> No.974595 [DELETED] 

>>973423
Yep. Even I a narcissistic fuck, can't shit all over the ease of use of the Hobbit hole image.
>>973444
Yep now I have to look at my shitty render every day. But it's almost over.

>> No.974603
File: 374 KB, 1652x1262, reference.jpg [View same] [iqdb] [saucenao] [google]
974603

>>974592
>>974593
this was my reference. i guess lower body is the main problem.

>> No.974631
File: 1.56 MB, 1080x1080, 1677713135201338.png [View same] [iqdb] [saucenao] [google]
974631

>>972982
feedback on colors and composition, please. yet to add a bicycle and some stones and waves on the water

>> No.974655
File: 1.42 MB, 2882x1936, drake_05.png [View same] [iqdb] [saucenao] [google]
974655

>>974225
Done?

>> No.974661 [DELETED] 

>>974655
the mouth looks weird, other than that well done

>> No.974718
File: 57 KB, 1920x1080, cgc.jpg [View same] [iqdb] [saucenao] [google]
974718

Tried making a cute gas canister. Think my ring around the middle might be too "intense".

>> No.974721
File: 861 KB, 720x1544, its all an illusion.webm [View same] [iqdb] [saucenao] [google]
974721

>>974583
Thanks, shame it's all an illusion

>>974594
This is good, I can't manage the horse without it looking retarded but will add the spirals and other column shapes.

>> No.974722
File: 2.31 MB, 1568x1568, test render 01.png [View same] [iqdb] [saucenao] [google]
974722

Here is another version with more material and decal variety, also switched to blender for rendering, what do you think?

>> No.974723
File: 492 KB, 348x645, gas cylinder.png [View same] [iqdb] [saucenao] [google]
974723

>>974718
the ring looks fine
The holes at the bottom look odd though. In reality it's hard to define an ellipse in engineering drawings so everyone sticks to circles or capsule shapes.
The grab handles at the top look a little thin. Imagine trying to grab those with gloves on.

>> No.974734

>>974723
The holes will be easy enough to change so I'll do that next.
The grab handles I struggled with. In the ref they were really thin, but flared out really far so it made them nice and curved. I couldn't do that so I just invented some. It's not as thin as it seems though. I think it's just the angle because the ends of the circular part taper in. Even still I'll redo the handles because I wasn't happy with them anyway and just tried to bluff it.
Also wanted to put some weld marks around where the handle connects to the main body. Not even sure how noticeable it'll be so I think a normal map should be good enough for that.
Thanks btw. I always appreciate feedback.

>> No.974766
File: 59 KB, 742x849, project.jpg [View same] [iqdb] [saucenao] [google]
974766

Made the hips wider and the feet smaler for a more female appearance.

>> No.974808
File: 413 KB, 604x780, Screenshot_1.png [View same] [iqdb] [saucenao] [google]
974808

>>974718
Nah it's fine, for comparison here's my low poly rendition of a propane tank.
The other anon hit the nail on the head, I'd only change the bottom holes (change their shape to capsules) and the grab handle things at the top (make those "tongues" a bit shorter so that a person can realistically fit his hand in there)

>> No.975020
File: 437 KB, 1800x496, 1.jpg [View same] [iqdb] [saucenao] [google]
975020

>>973724
Got it skinned. Now ready to move onto adding textures.

>> No.975021

>>975020
wait, no, I'm jumping the gun. There are some huge changes I need to make to the wings connected to the cockpit. they're too fare off model from the original. Other than that, I'm happy where this is going.

>> No.975022
File: 57 KB, 1920x1080, cgcr.jpg [View same] [iqdb] [saucenao] [google]
975022

>>974808
That's cool are you the anon who does all of those low poly assets?
Made the changes to the bottom holes and I just got rid of those "tongues" entirely. I think it looks better at least.

>> No.975024
File: 560 KB, 720x400, 1movie.webm [View same] [iqdb] [saucenao] [google]
975024

Rough sketch for a future scene, still a lot to do yet.

>> No.975027 [DELETED] 

>>975024
your mom is a bitch otherwize nice render

>> No.975039

>>974766
I still think if your brought those knees in, just a smidge, it would dramatically improve the figure.

>> No.975046

>>975024
>sketch
You're not sketching.
Call it what it is. A blockout.
I agree with the other anon. s'alright, keep going.

>> No.975067
File: 267 KB, 701x613, lil guy.png [View same] [iqdb] [saucenao] [google]
975067

r8 my lil fella

>> No.975092
File: 3.04 MB, 1974x1149, Screenshot 2024-02-22 015947.png [View same] [iqdb] [saucenao] [google]
975092

>>973300
>>973659
Definitely getting closer and closer, first time fucking around with Blender's hair system and it's fairly rough to use.

Anyone spot anything major off?
The eyebrows, lashes and hair are basically all placeholders atm.

>> No.975095
File: 710 KB, 1974x1149, tweaks.jpg [View same] [iqdb] [saucenao] [google]
975095

>>975092
That's really good anon, rather than typing it out I made a paint over as to what stands out to my eye, just overlay your picture and go back and forth. Like pay more careful attention to the shading of surfaces and where your pinching in the shading is and how it shaped, I think you need to relax and flatten some areas a bit to better capture what's there in the reference.

>> No.975097

>>975092
>>975095

At some point like this I feel one almost need to transition into texture/color as it's just too hard to see the difference
in geometry looking at a monochrome representation of the surface.

I realized this many years ago viewing a headscan of Olivia Wilde for Tron: Legacy and realized how different my interpretation of what she looked like
was from what her accurate untextured face actually looked like.

Like we experience shapes as being sized different depending on contrasting color with adjacent areas and things like that.

>> No.975125
File: 1.11 MB, 1920x1080, Hill env.jpg [View same] [iqdb] [saucenao] [google]
975125

>>974498
In the beginning stages of finalizing the rest of the mountaintop. The extra stuff up there doesn't need to be super duper detailed, but hopefully with the darkness your brain would fill in some stuff. They'll be pretty far from the camera, but I'd like it to be there for when I do a few extra camera angles.
All that's left for the detail up there is to finish up the details on large building on the front side of the mountain, the small one on the back side, fix some of the floating radio dishes, and place some extra light poles and stuff around the area to light some of it up. After that I think I'm done with it, and it'll be time to render after I set up some different camera angles.

Pic is brightened quite a bit to actually see the shit right now, but values will be closer to >>974498 but lamps and stuff around the area should brighten it up more naturally.

>> No.975142

>>975067
slem :D

>> No.975178

>>974478
23k tris.

>> No.975179
File: 395 KB, 600x338, BOING BOING BOING.gif [View same] [iqdb] [saucenao] [google]
975179

>>975067
Was this you?

>> No.975191

>>975022
Why yes that's me! Finally some fucking recognition!
Your propane tank looks good, good work

>> No.975192
File: 2.25 MB, 1365x811, sos.png [View same] [iqdb] [saucenao] [google]
975192

Today I textured my mail box. I also technically textured my bench, but all I did was just arrange UVs and slap on some generic wood + metal textures, it's such a simple object that I'm not even gonna bother doing any actual texture work for it

>> No.975199

>>975192
it looks nice, but why is it so low poly?

>> No.975203
File: 3.17 MB, 2048x1600, test render 01.png [View same] [iqdb] [saucenao] [google]
975203

I reworked the textures yet again, sorry guys for spamming this

>> No.975217

>>975179
no but i'm mirin that lil fella
look at him go. what a fun loving guy. what a silly little dude.

>> No.975236

>>975203
good, the different materials are looking more distinct

>> No.975253

>>975203
Lookin better and better anon. Much better than the previous iterations.
Would love to see some scorch marks and stuff from bullets or rockets that were deflected by the armor. Make that fucker look battle-worn, not just worn. You know, assuming it's a combat mech. If it's more industrial/gen purpose though, you're right on the money.

>> No.975254

>>975203
The brown parts and the axis on the frontal grabbers look wet, not metallic. Also, please don't render in a vacuum. Dark backgrounds and void environments don't help with judging how a model looks.

>> No.975266
File: 104 KB, 766x762, water green.png [View same] [iqdb] [saucenao] [google]
975266

>>975199
I just like it that way: faster and easier to make, it's easier on one's hardware for gaming purposes, and I just like how it looks.

>>975203
A E S T H E T I C
Great choice of colors anon-kun.
If for whatever reason you need to add more colors, I suggest picking a water green as it should play well with the others. I don't think your model needs any more colors though, it looks great as is

>> No.975271

>>975192
It would look good if you put some Sharp edges on the models.

>> No.975279
File: 485 KB, 1440x3088, Tower of babel latest 23022024.jpg [View same] [iqdb] [saucenao] [google]
975279

>>975203
>sorry guys for spamming this
anything to liven up this dead board is appreciated

>> No.975280

>>975203
A quick question
How does it see?
There's no obvious camera and nothing looks like glass apart from the window on the crotch area.

>> No.975283

>>975203
>sorry guys for spamming this
>only posted ot 3 times
The sad state of this board

>> No.975294

>>975191
Based. I used to look forward to your updates. They always reminded me of the PC graphics from the early 2000s and they always makes me feel nostalgic.
Thanks! I've came up with the idea for a meme theme thanks to the propane tank so I'm sure I'll spam the thread with that as it progresses.

>> No.975310

>>975283
I only ask for feedback when it is needed, constantly asking for help is more of a beginner thing wouldn´t you agree?

>> No.975311

>>975280
idk...gotta ask the concept artist. it´s future tech anyway...

>> No.975335
File: 70 KB, 793x273, opaint.jpg [View same] [iqdb] [saucenao] [google]
975335

>>975020
How do I paint on multiple uv maps? I can only paint on the first range at the moment.

>> No.975337

>>975335
>tl;dr
I don't have the knowledge to answer your question but to compensate I have other advice, i.e. I suggest to use only 1 image texture / UV map for your object, especially of the kind you're working with now

---

I don't mean to shoot you down but you appear to be using multiple texture maps (3 from the looks of it) for only one object. That seems overkill, especially if your object is only using flat colors in which case you could theoretically get away with a single extremely low resolution image for all of your model.
A single 2K (2048x2048 pixel resolution) image should be more than enough for your object, and if you wanna go lower res you can easily downscale. I highly recommend getting all your UVs into a single image texture / square

>> No.975338 [DELETED] 
File: 213 KB, 940x512, Color Palette Atlas.png [View same] [iqdb] [saucenao] [google]
975338

>>975337
>your object is only using flat colors in which case you could theoretically get away with a single extremely low resolution image for all of your model
You can test this right now if you want: download my image, use it as your texture, then move all of your green colored UVs to a green box, all of your red colored bits to a red box, and so forth, you'll soon understand what I'm trying to tell you

>> No.975346

>>975337
thanks for the feedback.

I spread it out because i didnt want to cram all that in there, but that seems to be the way to go. Should be easy to scale now that I have them laid out

im not painting the whole thing via maya. I was going to block out the colored shapes with a black line, take it into photoshop and start painting it in there. Saving there then reloading the texture in Maya to see how it looks. My laptop cant handle substance painter(something with its graphics card) but can juggle those programs just fine. lol.

Anyways, thanks.

>> No.975362 [DELETED] 
File: 1.52 MB, 1080x1080, GreatUncleanOneRotigusFinal.webm [View same] [iqdb] [saucenao] [google]
975362

And it is ""done"", it needs more details in the legs, the arms and the back to match the reference, the small guys are barely blocked and I'm not happy with the procedural thing that I used for the scars but I don't think I'll work more on this, at least now, 1.2k objects, 4M tris.

>> No.975443
File: 172 KB, 1920x1080, test-render.jpg [View same] [iqdb] [saucenao] [google]
975443

test render of an eva fan art im doing. will add more bg elements. what do yall think so far

>> No.975537

>>975443
Not bad.
Might want to do more with the water to make it actually come across as water. For like a good 15 seconds my brain was trying to parse it as a bridge made of a big red log. Pretty sure the shading of bright in the distance, and dark closer to the foreground is what's doing it.
Not really digging the flat cutoff from the horizon to the clouds either. Definitely seems like there should be some atmosphere or something to blend them better.

Also, yer waifu isn't meshing well with the 2d-ness of the rest of the scene. She still comes off as 3d and low poly at that.
Still, keep it up. For what it's worth the grass and stuff is really nice.

>> No.975539

>>975362
Looks great. It'd look a little better with all the limbs properly dynameshed together and with some polypaint. But what's there is pretty good.

>>975443
Best girl looks like she's just floating in the scene, some shadows would go a long way in grounding her. Also feels like something's missing from the left side but whether that's a tree or some debris or some sunken structures out on the sea like electric towers or something, I don't really know.

>> No.975542
File: 144 KB, 449x450, 1708725690279701.jpg [View same] [iqdb] [saucenao] [google]
975542

>>975335
If you're going down this route, try to scale your UVs so that they use four sheets instead of three, pic related. With four you can assign a material to each sheet and paint each one separately (just like you could with three), but you can also do the sensible thing of scaling it all down to a single sheet and just upping the texture map resolution, say to 2k or 4k. Then all you have to do is paint the one map on the one material.
With three sheets you can just be sad. Do things in ones or fours for maximum UV happiness.

>> No.975546

>>975310
You could post without asking for feedback if you wanted to. Sometimes it's nice to come here and see what people are working on, sort of the point of it being called the "WIP" thread I would suppose.

>> No.975591

>>975546
This.
/wip/ is more of a public space to post shit. Works in progress, recently finished stuff, it all belongs here. Whether you ask for feedback or not. Which you'll get, whether you ask for it or not.

>> No.975602 [DELETED] 

>>975539
I'm glad you liked it, ty, I haven't painted in a long time, I maybe do it in the next models, I have a couple of Malifaux scans and some Corvus Belli miniatures that I want to sculpt next month if I have tiime.

>> No.975683
File: 108 KB, 662x941, chink sword.png [View same] [iqdb] [saucenao] [google]
975683

>> No.975684

>>975443
Definitely needs more workflow, keep going anon, just crank up the workflow
>>975362
Good workflow maybe more workflow would push it further
>>975683
Not enough, needs more workflow
>>975095
The likeness it's there, just needs more workflow

>> No.975685

>>974766
Needs More workflow anon

>> No.975695 [DELETED] 

>>975684
>>975685
So you just got up today. Sat on your desk, turned on your pc, went to boards.4chan.org/3/, made some idiotic replies prolonging the 'workflow'meme and said to yourself 'gosh I am so funny'?

>> No.975696

>>975684
is there a free blender addon to more efficiently add workflow into my workflow?

>> No.975700
File: 304 KB, 1030x731, 25.02.2024_21.56.27_REC.png [View same] [iqdb] [saucenao] [google]
975700

I thought I should actually give blender another try and came up with this. Now I have to figure out how textures work and then how to use it in unity

>> No.975712

>>975696
https://blendermarket.com/products/machin3tools
if you get your money for free like i do, it's free
otherwise it's £0.99

>>975700
you supposedly can just use the .blend file directly in unity

>> No.975714

>>975695
Why don't you make a thread crying about it? I'm just giving you constructive criticism

>> No.975716

>>975695
Cris can take constructive advice really well, I don't understand why you can't, are you really that fragile?

>> No.975722

>>975695
I think the workflow thing is funny, anon. Don't take it so seriously, geez.

>> No.975738

>>973530
>keep those normal map values low especially when you're using offline renderers.
Could you (or anyone else) please elaborate on this? Will higher nornal strengths ACK my render times?

>> No.975740

>>975722
I thought it was funny but I'm starting to wonder if the anon doing it is doing it because it's funny. It's more likely that he's literally insane and wants to kill /3/ because he lost an argument with cris two months ago.
For the record, I've never seen the "Maya has no topology" anon, the anon calling everyone else cris and the workflow anon arguing with each other. It's probably the same person. For a place like this it takes only one person to completely ruin it. I wish these types would use tripcodes still so everyone else could filter them.

>> No.975741

I’m a programmer looking for a 3D artist gf.

We kiss each other and we make amazing RPGs of the likes of Dwarf Fortress.

Graphics will be most likely a Kings field 4 style with modern lighting.

Reply to this post if you’re her.

>> No.975742

>>975741
I think the chair chud could be a decent substitute for a woman. He loves getting his gay asshole pounded by big black men like you.

>> No.975743

>>973023
can i add you on discord? Have a question regarding coding in blender/making a tool for pose drawing.

>> No.975755 [DELETED] 

>>975741
I can be your girlfriend yes.
>>975742
Yes. I could be his girlfriend yes. Thank you for the cupid service.

>> No.975787

>>975743
Uh, are you sure you replied to the right person? I'm no coder. I just learned a thing or two in geometry nodes. And I still have much more to learn. My little projection test is very inefficient, and when I apply it to a full body mesh, it takes like 30 seconds to process. Because it's creating about 40 million points, and then deleting a lot of them. Deleting points increases the speed when I'm only testing a few basic shapes. But I'm not sure deleting points increases speed when the point count is so high.

Also, after it's finished processing all the points on a full body mesh, it's failing to delete a lot of the points. A lot of the points are not touching the surface of the body. So it's losing accuracy somehow, and I'm not aware how. I can't even change the camera when so many points are on the screen, without it taking forever to update the viewport.

Basically, my method failing, because it's creating and deleting too many points. I have to figure out a way to make it only create the necessary points. Or at least a number near to the necessary amount.

Anyway, you didn't ask me about all that. If you actually replied to the right anon, then you give me your discord name, and I'll consider adding you.(I'm trying to retain what anonymity I can) Just know that I'm not a coder.

>> No.975794

>>975787
I think that he is talking about something else anon, he needs help to pose his little friend in a more standing position if you know what I mean

>> No.975797

One question for everyone here, has someone else spend so much time 3d modelling his waifu that accidentally created a Tulpa?

>> No.975822

>>975797
No.
Tulpas aren't real. You just have an overactive imagination.
No, I don't care if you think "it's totally real, guize. Please believe me". Stop being an dumb tumblr idiot and talking yourself into an autistic belief that will harm your mental health more than it already is, and ruin tour relationships with everyone you know, after you spout some bullshit.

>> No.975826
File: 270 KB, 885x758, retop.png [View same] [iqdb] [saucenao] [google]
975826

Here's the start of a retopo of a basic cute girl head sculpt I'm working on.
Nothing noteworthy but I like the way the nose turned out

>> No.975848

>>975822
>mental illness isnt real
This is your assertion.

>> No.975850

>>975848
>>mental illness isnt real
This but unironically

>> No.975853
File: 493 KB, 500x281, 1698593683720114.gif [View same] [iqdb] [saucenao] [google]
975853

>>975826
lookin real cute there
(it is indeed a nice nose however)

>> No.975867
File: 199 KB, 1844x392, fuller.png [View same] [iqdb] [saucenao] [google]
975867

now that is a sexy ass fuller and I'm not afraid to toot my own horn over it. god I love a good fuller. aspiring bladesmiths, take note.

>> No.975872

>>975826
wish I was good enough to do that not even gonna lie to you family

>> No.975930
File: 126 KB, 960x540, HelicopterHeliCopter.jpg [View same] [iqdb] [saucenao] [google]
975930

>>975542
I went ahead and crammed the uvs onto one map. Everything turned out pretty well.

Used a Phong shader to look somewhat like the old renders of the game. Saved me the headache of painting it for a flat shader.

>> No.975932
File: 761 KB, 650x867, DoDonPachi_PG2.jpg [View same] [iqdb] [saucenao] [google]
975932

>>975930
I made this render because I couldn't find an official render of the helicopter like the red and blue ship renders in picrel from the devs.

It isn't to jarring and fits well with the official renders in my opinion. could use some tweaking with the settings, but I'm not familiar with them.

>> No.975935

>>975930
>>975932
Looks really good.
Great job.

>> No.975961

finished my evangelion animation :)

https://imgur.com/a/7zpaJQ9

>> No.975963

>>975961
Neato.
I think the waves not being a clean loop, and the overdone camera shake kills it though.
Probably could have been enough to make the waves just play in forward and reverse to loop them. Camera shake (while I use it as well depending on the scene), I'd just get rid of entirely in this case. There's enough movement in the scene where it's not doing it any favors.
Overall though, it' ain't bad. Nice work.

>> No.975978
File: 79 KB, 640x480, 0001-0100.webm [View same] [iqdb] [saucenao] [google]
975978

>> No.975986
File: 198 KB, 3840x1640, spaes.jpg [View same] [iqdb] [saucenao] [google]
975986

making a cool space short

>>975961
this looks amazing, but yea the frequency of the camera shake needs to be reduced

>> No.975988

>>975685
true. i am a noob in 3d modeling.

>> No.975997

>>975930
KEE-NO
Good shit anon
Keeping all UVs onto just one map makes things easier if you wanna make alterations to it in the future

>> No.976023

>>975978
That t-shirt is too elastic maybe it's more heavy than it should?

>> No.976024
File: 694 KB, 1441x1936, subaru_01.png [View same] [iqdb] [saucenao] [google]
976024

>>974655

>> No.976028

>>976024
Nice, are you one of those Tulpa creators?

>> No.976032

>>976024
Good luck with your tulpa waifu my friend

>> No.976041
File: 26 KB, 640x480, undress_01.jpg [View same] [iqdb] [saucenao] [google]
976041

>>976023
The spring forces are low. High spring forces just require more CPU time. For the most part I wanted to show how I'm able to conform the dress to the body in such a way that the shape underneath is not completely lost.
You know, maybe to make something sexy but not porn.

>> No.976050
File: 30 KB, 640x480, undress_02.jpg [View same] [iqdb] [saucenao] [google]
976050

Because if you want the dress to stay put you need to cheat little bit by attaching it to static goals. Because over time it will drift no matter what. But then maybe I want the character to experience wardrobe failure.

>> No.976139

Hi all, I'm currently working on my first serious game project; I've dilly dallied a lot in unreal, but decided to buckle down and actually make something.

Since it's my first commercial project I've kept it pretty simple. Basically there is a zombie outbreak (very original i know lol) and you have one day to prepare for the oncoming horde, where the actions you took affect what happens when night falls. The actual gameplay is similar to a point and click game where you look around the environment and solve some basic puzzles, but in 3D.

Point and click games are actually why I decided to make this game in the first place, as I always wanted to freely explore some of my favorite point and click games. (mainly from Scriptwelder with his "Don't Escape" series, which this game is inspired by)

On the more technical side, I originally was making it in UE5 but decided to downgrade back to UE4 as I wasn't going to use any of the new Lumen & Nanite features, plus UE4 seems more stable anyways.

Here's a very early look at the game so far; file is very large so I uploaded it to yt. (https://www.youtube.com/watch?v=eOFEciaBHhk))

The compression will likely ruin the effect, but I've gone for a pixelated 3D art style. This has actually come in handy since I can render the game at 20% resolution, and then upscale with nearest neighbor filtering. As a result, I've gotten this game to run at 120+ fps on an old Intel Integrated GPU. I suspect once the game is more complete it'll be lower, but I'm really hoping I can keep performance high on low end machines.

The name of the game right now is Infected Dawn. I'm relatively happy with it, but I might change it later down the line, who knows.

Obviously things are still largely a work in progress, but I'd love to hear your thoughts on what I have so far.

>> No.976147

>>976139
It's very generic. Very. Generic.

>> No.976150

>>976147
Good point, I think I have some ideas on how I might fix that.

>> No.976155

>>976150
I mean... If you're going to make a game about a zombie something you may as well ask ChatGPT to write a better script for your game for you.

>> No.976166

>>976139
I don't mean to dampen your spirits, and I don't really expect the world from regular people or beginners, but like, where's your hook?
What sets your game apart (or will), from anything else? Why should anyone play it, or even buy it?

It's pretty obvious you're in the very early stages, so obviously it's a bit much to expect things to be compelling, but it's something that you should definitely think about. The whole "if you build it, they will come" mentality isn't really a good one to have, since just because you "made a thing" doesn't mean you're entitled for it to be successful (even very marginally). Especially these days.
Leaning on visuals and an aesthetic is only going to do so much if the gameplay is bog standard fare. Though I have to agree that things are pretty generic, and looks like a million other Youtube-bait games I've seen on Steam.

That being said, your game has yet to take shape, so you still have plenty of room to really flex your creative muscles before things become more or less set in stone. Hopefully you keep working on it, and I hope it shapes up nicely.

>> No.976172

>>976155
yea, im not super great at writing stories lol

>>976166
>I don't mean to dampen your spirits, and I don't really expect the world from regular people or beginners, but like, where's your hook?
What sets your game apart (or will), from anything else? Why should anyone play it, or even buy it?

Tbh it is actually kinda refreshing for people to ask the hard questions, since I genuinely do want people to enjoy my game.

I suppose I didn't explain the intended hook super well in my original post, but one of the things that I really enjoyed about Scriptwelder's games was the presentation and environmental storytelling. But I also thought it was neat how there wasn't just one way to complete the game, as if you didn't find a specific item you could still complete it. (and this also added to the replay-ability, since I've replayed the games many times trying to find all the different endings and ways to progress)

I also really liked how it encouraged exploration and thinking out of the box, which is something that I would like to do as well, along with everything else mentioned above.

>It's pretty obvious you're in the very early stages, so obviously it's a bit much to expect things to be compelling, but it's something that you should definitely think about. The whole "if you build it, they will come" mentality isn't really a good one to have, since just because you "made a thing" doesn't mean you're entitled for it to be successful (even very marginally). Especially these days.
Leaning on visuals and an aesthetic is only going to do so much if the gameplay is bog standard fare. Though I have to agree that things are pretty generic, and looks like a million other Youtube-bait games I've seen on Steam.

Yea that is something I haven't quite figured out at this stage, but hopefully once I've fleshed out more of the gameplay I can more properly assess how engaging it is and what specific things I'll need to add/change to make people actually want to play it.

>> No.976174
File: 511 KB, 638x691, image_2024-03-01_175537411.png [View same] [iqdb] [saucenao] [google]
976174

Working on a low poly thing for fun and general experience to get better, any tips for how to improve?

>> No.976191

>>976155
>ask ChatGPT
You are so lazy, next time you are going to ask chatgpt if you need to clean your ass when you finish going for a shit.
The answer is yes, you should. Just in case

>> No.976212
File: 144 KB, 640x480, 0001-0150.webm [View same] [iqdb] [saucenao] [google]
976212

There are still some issues but it's coming along. It also does joint bending corrections and... other stuff I don't want to talk about.

>> No.976218

>>976174
To start id just axe the notion of "low poly". Model it as best you can without restricting yourself.
Your basic shape is an accurate representation of your design, which is good. The design is just lacking.
I assume you made it. That's good and very helpful to know how to make your own image planes. If you want to improve youll want to push your comfort zone. Draw a more elaborate design and model it.

>> No.976277

>>976174
Don't listen to this guy >>976218, just complete the model you're doing fellow furfag fren, model it and then texture it to the best of your abilities, that's it.
3D modeling is iterative: you'll get better with each new attempt you'll make at a particular subject, don't worry about "improving", you're already improving by making this very model, and you'll get even better by making more down the line.

Your model looks good so far, in the drawings you used for reference the arms were freakishly long, like SCP-96 tier, but you corrected that a bit in your 3D model; the fact that they're unusually long isn't as evident thanks to the fact that your character's torso is also quite long so the two even themselves out in the viewer's eyes. Remember that a person's arms go, length-wise, all the way down to the crotch-level, the tips of one's fingers extend slighly below it. But seeing the stylized nature of your character, anatomical liberties are acceptable

>> No.976316
File: 292 KB, 739x658, dudeweedlmao.png [View same] [iqdb] [saucenao] [google]
976316

ye nigga that's the good kush

>> No.976390
File: 1.74 MB, 1000x1080, Sanja.webm [View same] [iqdb] [saucenao] [google]
976390

I love servals.

>> No.976404

>>976390
There she is in 3d.
Not bad anon, captured your sketches really well.
Planning to do more with her?

>> No.976414
File: 1.21 MB, 960x540, 0768-2077.webm [View same] [iqdb] [saucenao] [google]
976414

>>976404
Thank you!
>Planning to do more with her?
Mayyyyyyyyyyyyyyyybe...

>> No.976441
File: 518 KB, 557x817, monroe.png [View same] [iqdb] [saucenao] [google]
976441

>>976316
Givin' mah nigga some clothes. Texturing is at about 65% progress. Next is weight painting. God I fucking hate weight painting.

>> No.976443

>>976441
This isnt a recycle of that krystal model that was floating around is it?

>> No.976461

>>976443
Nah I made this from scratch, didn't even know there was a Krystal model going around

>> No.976507
File: 79 KB, 1620x902, T-Pose_1.jpg [View same] [iqdb] [saucenao] [google]
976507

>> No.976511

>>976507
oh, hi Fertility Goddess Sculpture from 28300 BC

>> No.976527

>>976441
Weight painting is done, and I have to thank this tutorial for it: https://www.youtube.com/watch?v=IPQ2QVVIxLU
>tl;dw
If you already have a basemesh weight painted, you can just Transfer Weights from your basemesh to your clothes mesh. It worked like a charm, all I had to do afterwards was just manually weight paint the mouth and neck areas, that's it

>> No.976528
File: 1.01 MB, 1510x837, RUN THROUGH THE JUNGLE.png [View same] [iqdb] [saucenao] [google]
976528

>>976527
Oh whoops forgot pic

>> No.976530

>>976507 >>976528
Why am I not surprised that you're using Blender?

>> No.976534
File: 60 KB, 1096x890, T-Pose_2.jpg [View same] [iqdb] [saucenao] [google]
976534

>>976530
Why would I need Zbrush or Maya just to sculpt a pair of breasts?

>> No.976541

>>976530
I dunno brah, only you can know that

>> No.976552

>>976534
So you can make something that has some sort of resemblance to a quality product.

>> No.976557

>>976507
Semen

>>976530
>>976552
Blender is the only good software for making erotica

>> No.976564

>>976557
I don't have issues with erotica. It's the anthropomorphic animals part that that bothers me a little bit.

>> No.976576
File: 37 KB, 791x805, R.jpg [View same] [iqdb] [saucenao] [google]
976576

>>976534
I dunno mayne

>> No.976588

>>976530
Why am I not surprised that you're a faggot?

>> No.976623
File: 611 KB, 1920x1080, 2024-03-05 (3).png [View same] [iqdb] [saucenao] [google]
976623

Trying to model a Subaru...

>> No.976628

>>976623
Better start thinking about edge flow right now, anon. Good edge flow is the difference between shitting out a car by the end, or a vaguely car-shaped object.
Keep it up though, cars are pretty tough. Really fucking good exercise in edge flow though.

>> No.976630
File: 1.04 MB, 1920x1080, Hollywood Hill.jpg [View same] [iqdb] [saucenao] [google]
976630

>>975125
Alright, finished up the scene and I'm doing the final renders now. Had to take a bit of time to finalize the compositing stuff and figure out how I was gonna put things together for the final video, since each layer (foreground/mid/other shit like spotlights/airplanes) all had different timings and loop points to save on render times.
Like the foreground runs for 500 frames since I have movement on the plants that's essentially invisible at this point, but the city runs for 1000 to give the cars a bit more time to travel. The airplanes are like 4000 frames, but I can manually layer them into the scene whenever I want, so I rendered out their full travel. They didn't take too long to render since they're like 6 pixels.

Now that that's all out of the way, I can finish up rendering the rest of the animation. Nothing major, just a few different camera angles. Close ups of the car with extreme bokeh, a wide shot showing more of the area, an aerial view showing the radio tower, maybe one from inside the car, then probably another angle I haven't quite figured out yet.
All the shots should end up being a few mins or so of video that I can cut between.

But yeah, enough of the shitty blog post, here's a vid of the first shot at full res.
https://files.catbox.moe/rtd5zw.mp4
There's a visual glitch at the end because my dumb ass accidentally bumped the main layer back 2 frames.

>> No.976631

>>976628
Thanks, Anon. I'm getting my feet wet into Hard Surface modelling. My focus is going to be solely on cars purely because I am passionate about vehicles.

I'm learning about a lot of things as I go. I am willing to pursue a "professional" workflow with time as I go, learning from my mistakes etc..

>> No.976632
File: 327 KB, 1920x1080, 2024-03-05 (4).png [View same] [iqdb] [saucenao] [google]
976632

>>976623
If anyone knows of any good tutorials out there on mastering good topology and edge flow or would like to explain it...it would be greatly appreciated.

>> No.976633
File: 349 KB, 1920x1080, 2024-03-05 (5).png [View same] [iqdb] [saucenao] [google]
976633

>>976623

>> No.976634

>>976630
I assume you know already that the exhaust visibly glitches at 13.5 and 27 seconds

>> No.976635

>>976632
Thomas Colin 3D is a topology chad. Works in the industry, but I think he has a >discord where he helps people out. Seems like a good person.
If you're not down with >discord, or just in case I'm wrong about him having one, he has videos on topology so he might be worth checking out.

>> No.976636

>>976635
Thanks mate

>> No.976638
File: 715 KB, 800x337, Exhaust.webm [View same] [iqdb] [saucenao] [google]
976638

>>976634
>I assume you know already
Fuck, I do now.
Good catch, I've been staring at this shit so long I've become blind to it. The exhaust is supposed to be a seamless loop, but I think I made an error somewhere.
Actually, I know exactly what it is. Foreground anim isn't 500 frames, it's 400. Out of habit, I start anims on frame 100. Exhaust anim is 500 frames instead of 400. Blunder of the fuckin century.
Should be an easy fix, just render out the section with the exhaust fixed and plop it over like a patch.

Nice save anon, thanks.

>> No.976651
File: 224 KB, 1543x906, T-Pose_3.jpg [View same] [iqdb] [saucenao] [google]
976651

>> No.976658

Does anyone here use a render farm? I'm just curious how they are and if they're worth it.

>> No.976665
File: 339 KB, 1080x1920, Screenshot_20240305-101842.png [View same] [iqdb] [saucenao] [google]
976665

>>976651
Your butt topo is going to cause you headaches down the line.

>> No.976666

>>976658
render farms inspect your renders and dont allow adult content, so people here dont use them.

>> No.976676
File: 99 KB, 1000x1000, monkey.webm [View same] [iqdb] [saucenao] [google]
976676

Just gonna drop these here

>> No.976677
File: 387 KB, 1000x1000, monkey dance.webm [View same] [iqdb] [saucenao] [google]
976677

>> No.976678
File: 191 KB, 1000x1000, monkey bars.webm [View same] [iqdb] [saucenao] [google]
976678

>> No.976679

>>976666
Id like to hit up some render farms with a few thousand work requests for detailed renders of goatse. Let them inpect that

>> No.976680

>>976679
you would get banned and lose a lot of money

>> No.976681

>>976666
Is that true? Fucking state of 2020s internet.
I'm not even doing adult content, but things like that annoy me.

>> No.976696

>>976658
I doubt it counts, but I've used SheepIt on occasion when rendering an animation is too much to do on my own in a given timeframe. Wasn't bad, some frames come out shit due to whatever hardware someone is using (90% of the time is the 20xx series that shits the bed), but I can deal with it if it multiplies my render power by several times. Just keep a notepad handy, mark the frames and re render them locally or through the farm again.
Only shit that sucks is that there's a file size limit, so you have to try and cram in all the shit you can and stay under.
But it's worked handy in a pinch. I can only imagine a paid one would be better since you don't have to wait in line.
Files are apparently stored locally when rendering, but I've not been able to actually access them when I tried rendering for other people, and like the other anon said, adult shit is a no go (at least on that service). I'm sure there's a few around that do allow it if you shop around though.

>> No.976702
File: 1.40 MB, 1920x1080, viewport render.png [View same] [iqdb] [saucenao] [google]
976702

what is it gonna be?

>> No.976704

>>976680
Theres no way in hell they wont give you a refund if theu refuse their business plus just use different emails. Set up a bot to do it. Make their little censorship gremlin peer into the void 100 times a day.

>> No.976705

>>976702
2 hot air balloons racing around the world

>> No.976706

>>976696
I did look into SheepIt, but I didn't like the idea of people using my machine to render their shit. Reminded me too much of Hola! and how that used to work.
I've seen people mention them in the past and they all seemed to be fairly priced and snappy when it came to rendering. When I looked it up I needed to have a Pixar budget to get anything done. Although I think they actually make money from 3D which I'm yet to do. I think that's my next step which is why I was interested in render farms. I don't like the idea of not being able to use my computer when something is rendering.
Appreciate the help. Maybe SheepIt won't be that bad. At least in the early stages until I can afford to pay for it.

>> No.976707
File: 43 KB, 1280x720, 2001.jpg [View same] [iqdb] [saucenao] [google]
976707

>>976705
nah, remake of 2001: a space odyssey iconic opening scene

>> No.976708

>>976706
sheep it is terrible. There is nothing worse than having someone else use up all your compute. Also, your room gets loud af and hot af

>> No.976709

>>976708
Fuck that then I won't even bother with it.

>> No.976710

>>976676
>>976677
>>976678
Is that... a vagina?

>> No.976721

>>976708
>>976706
>I didn't like the idea of people using my machine to render their shit.
>There is nothing worse than having someone else use up all your compute
You actually don't need to do that anymore. They used to have a retarded point system where you needed to render out other people's shit to have a positive balance, otherwise you couldn't render. Now you just get placed at the bottom of the queue. That's it. You don't get bumped around for having a low balance compared to someone else.
I think I've rendered out like 10k frames at 4k without rendering anyone's. Saved me weeks of rendering for fucking nothing.

Obviously if you abuse the system outright, they'll take note of it and take action, but if you're using it sparingly for final renders every few weeks/months, it's just gonna look like regular use rather than abuse.

>> No.976741
File: 760 KB, 1882x923, T-Pose_4 .png [View same] [iqdb] [saucenao] [google]
976741

>>976665
Thanks.

>> No.976757
File: 244 KB, 1000x1000, bear flip.webm [View same] [iqdb] [saucenao] [google]
976757

>>976710
degenerate coomer
it's just where I fucked up the UV's

>> No.976769

>>976757
lol silly bear

>> No.976774

>>976757
>>976769
Such a joker.

>> No.976808

>>976710
>Kore wa...manko ka?

>> No.976812
File: 586 KB, 1910x1962, comparison.png [View same] [iqdb] [saucenao] [google]
976812

Continuation from this:
>>976139

Been experimenting with the visuals quite a bit to try make it more interesting by adjusting things like the fog and color temperature. I swapped out the tree and grass models, although the fence and car will likely have to be swapped to match the more retro style as they kinda stick out. As shown in the pic, I've also been messing around with different lighting scenarios that the game will switch to depending on how much time is left in the day to serve as a visual indicator that you're almost out of time.

I've also been working on the player movement a bit, adding running and jumping to make exploring the different levels easier and more enjoyable. That is, if there was actually something to explore. Still have a lot of work left to do in giving the player a reason to actually walk around and discover stuff.

https://www.youtube.com/watch?v=UJKqOSTZVdo

Unfortunately even uploading in 1080p isn't enough to preserve the film grain and pixel effect, which I also overhauled to be much cleaner.

One thing I'm still struggling with is the choice of music. I've tried quite a few different music tracks (all of them being royalty free) and I'm not sure if they totally fit (along with the fact that royalty free music is used in many places and could stick out by how common it is). I've been looking into Epidemic sound, as a lot of people seem to use that platform to find music for their stuff. If I like it I may buy a few tracks to use if they're worth it.

>> No.976815

>>976812
I quite like the visuals and things. It really reminds me of Dayz so it could be that because it's one of my favourite games still.
You could take some kind of inspiration from it too. For instance it has no music, it barely even has background noise. All it has is footsteps and wind (sometimes thunder and rain), but that adds to the atmosphere because if you hear a bear/deer/wolf that's nearby you can hear it better. If you hear a sudden gunshot you go into survival mode etc. So no music in it is a benefit. It's also makes sense in the context of the game because it's set after an outbreak of some unspecified virus/disease/military experiment so you wouldn't have any machinery/transport/sounds of human activity (aside for other "survivors" ie players) etc.
There's no actual endgoal to the game neither, it doesn't even have a story to it except what I mentioned about the outbreak you just larp and make up your own. But it forces you to explore because you need food/clothes/weapons/arguably a base etc and all of the good stuff is further inland (you always start near the coast). If you want to build a base you need a saw/axe/hammer/nails/wire. If you want a car you need to find one first and then find a battery/radiator/spark plug/fuel. Etc. You can't just find everything in one place so you do need to explore because you have no other choice.
Obviously it's an online multiplayer game and my assumption is that you're making a sp game so the two aren't exactly the same, but you can take something from it maybe.

If you don't want it to be zombies you could try something else. For example the Forest has cannibals, that might even be a better game to take inspiration from because it's sp too, albeit with a storyline. Dayz, despite the name, has "infected" rather than zombies.
Everything takes inspiration from something else.
Sorry I couldn't be more help, but keep at it.

>> No.976822
File: 285 KB, 1885x910, T-Pose_5.jpg [View same] [iqdb] [saucenao] [google]
976822

Finally got some clean topology. But I have no idea where to go from here.

>> No.976825

>>976822
gross looking model desu

>> No.976829

>>976822
Next is to rig it

>> No.976833

>>976822
I could be wrong, but from the looks of it, you still have the mirror modifier on. I suspect that's the case because it's difficult to affect the center line of mirrored geometry, causing a seam to form down the center. Now that you topology is clean, apply the mirror modifier, making the thing 1 whole piece. Then, go to sculpting tools, *make sure symmetry is enabled*, and fix the center of your model. You'll find that as a whole piece, the center will smooth out just fine.
The particular area I think needs fixing, is her crotch. You should define a mound. I don't know what you plan to use this model for. but considering her giant tits and hips, I suspect something lewd. But that doesn't mean you need to give her a big bulging mound. She doesn't have any genitals anyway, so just give her a little V shape there. Subtle. Also, when a butt is high and tight, and you're looking straight from the back, you can see the mound peaking from behind the cheeks.

I think it should be possible without changing any topology here. Just some careful vertex pushing. Don't go crazy and break your model apart.

>> No.976842
File: 499 KB, 1080x1080, Monster.webm [View same] [iqdb] [saucenao] [google]
976842

Just wanted to try making a quick animation because I haven't done one before.
Would you drink it /wip/ bros?

>> No.976849

>>976842
Juice that makes you trans

>> No.976866
File: 322 KB, 1000x1000, burger.webm [View same] [iqdb] [saucenao] [google]
976866

>>976842
never use the linear interpolation mode - rookie mistake
bounce might work better in your instance
add some textures to the ground as well
do a 3 point lighting set up and at least get a studio hdri from polyhaven
Stick some bones in the can and get it to do a wave or something

>> No.976869

>>976866
>never use the linear interpolation mode
Fucking retarded.
There's a time and a place for it. Anon's Monster though is just the wrong place for it.

>> No.976901

>>976842
lel nice one, would drink

>> No.976903
File: 128 KB, 1880x907, T-Pose_6.jpg [View same] [iqdb] [saucenao] [google]
976903

Now, if only I knew how to do hair, clothes, uv, rig, animation, etc...

>> No.976916

>>976822
why does her head look so cone shaped? Why is her chest not obeying gravity, even a little bit?

>> No.976918
File: 35 KB, 864x875, T-Pose_6_1.jpg [View same] [iqdb] [saucenao] [google]
976918

>>976916
How would you do it?
Here's the model: https://files.catbox.moe/3fafrx.obj

>> No.976919

>>976918
>How would you do it?
-make the underlying skull look like an actual human
-have the chest obey the laws of physics
I'm not saving and opening any file from this god forsaken website

>> No.976932
File: 457 KB, 738x864, Flab.png [View same] [iqdb] [saucenao] [google]
976932

>>976919
If you like em flabby then you only had to ask.

>> No.976936

>>976866
Why not use linear interpolation?
>bounce might work better in your instance
I'll give that a try.
I haven't really done anything to the ground it was just for the can to stand on really. Like I say it was just to see how difficult basic animation was. I will be texturing it and whatnot though.
>get a studio hdri
That was my mistake. I thought I had picked a studio because the thumbnails and names are similar. Already changed it and it looks better. Glad you pointed it out though because I wondered why it looked so off and I couldn't get it right.
Wouldn't it look trippy if I made it do a wave? I might be imagining it wrong. Do you have a video of what you mean that you can show me?

>>>976869
Why is it the wrong place for it :(

>>976901
Based I'll save you one and put it on ice.

>> No.976952

>>976936
>Why is it the wrong place for it :(
Just the way that it comes to an instant stop while going up and twisting. You'd want to ease it out ever so slightly.
It's not a huge deal, honestly.

>> No.976958
File: 184 KB, 1000x1000, monster energy advert 2024.webm [View same] [iqdb] [saucenao] [google]
976958

>>976936
Here you are lad
I couldn't find a monster energy model for free online without logging into some stupid account so I made one myself.
When I said wave I meant wave like the pixar lamp.
Your linear interpolation looks shit because it looks like a robot is moving and is very jarring on the eyes. A bezier interpolation is a lot softer and more natural because it has non linear acceleration. You may as well go for a constant interpolation mode it's that bad.
Zoom a bit too, maybe change the aspect ration to a 1:2 vertical rotation video. The background adds zero value to the render in your case.

>> No.976979

>>976932
sigma female

>> No.976984
File: 1.51 MB, 2622x568, pauldron geo.png [View same] [iqdb] [saucenao] [google]
976984

I'm so happy bros, it took me a month+ to figure this out but I finally found a way to make a subdividable pauldron with minimal shading artifacts and is suitable to be sculpted on...

Basically I had to make the pauldron without the gap, then boolean cut the gap and then shift around some verts. Was having trouble maintaining a nice sharp angle using crap like bevels/creases, it made the model look abhorrent.

With one of those nifty quad remeshers it could probably be refiddled into something less cumbersome now.

>> No.977009

>>976932
Based GILF enjoyer

>> No.977022

>>976952
Ah yeah I get it. I took the "inspiration" from the Pepsi Perfect scene and it does pretty much come to a straight stop in that. You'll have to blame them for tricking me.
https://www.youtube.com/watch?v=11BwLs3pHF4

>>976958
Thanks I get what you mean with the wave now. I'm not sure why I would do it though?
>it looks like a robot is moving
But isn't that the way it's supposed to be? Like the platform it's on (that the can is on, not all of the floor) is connected to something mechanical and something mechanical would move like a robot?
>Zoom a bit too, maybe change the aspect ration to a 1:2 vertical rotation video
I did mean to animate the camera so it would zoom in as the can rises, but I totally forgot. I think you mean have the camera zoomed in overall without animating it though.
>maybe change the aspect ration to a 1:2 vertical rotation video
I'll do that too, but since you said
>The background adds zero value to the render in your case
You've given me the idea to try something else with the background. If it doesn't work I'll change it a 1:2 ratio and keep it as that.

Thanks lads.

>> No.977024

>>976842
I wonder if it would look better if it kept rotating after it comes to a stop. Or if the can wobbled a bit afterwards.

>> No.977061
File: 5 KB, 114x73, heh.png [View same] [iqdb] [saucenao] [google]
977061

heh
hahaha
ehehehee

>> No.977066

>>977022
>"inspiration" from the Pepsi Perfect scene and it does pretty much come to a straight stop in that
Fair enough. I think it's the speed of it that's handling that though. And if you watch it very closely it does have a teeny tiny bit of slowdown near the very end. It's so subtle I might be imagining it, but I'm pretty sure it's there. Even has a tiny bit of bounce up and down, if you slow it down to like 0.25x speed.
But yeah, carry on. Like I said, it's not a huge deal. It's not going to break the humor in your anim if you don't have it.

>> No.977067
File: 72 KB, 569x357, curve.png [View same] [iqdb] [saucenao] [google]
977067

>>976984
never mind. the method I used doesn't let me round the edges of this part. This stuff is too hard.

I don't suppose anyone has any insight into how I could model the above shape in a way that can be subdivided and without crapshit shading?

>> No.977128

>>977067
Couldn't you just bevel that corner?
Might be easier for someone to help if you post your topography too.

>> No.977147
File: 481 KB, 1000x1000, THE monster energy official advert 2024.webm [View same] [iqdb] [saucenao] [google]
977147

>>977022
When I say robot I mean a more extreme version of these:
https://www.youtube.com/watch?v=cR-YlZ9NdIA
Even these top of the line robots have noticeable acceleration and overshoot which your animation lacks
A little goes a long way with acceleration values and overshoots. You can exaggerate it as little or much as you like but it must be there otherwise your animation will look jarring and offensive
Look at a slow motion video of machinery and you'll see that everything moves and vibrates (which is just oscillating movement), even solid metal.
https://www.youtube.com/watch?v=kel_OwSYos4
A drinks dispenser will certainly be no better.
If you want to go very fast and very rigid, adding motion blur will help you

>> No.977183

>>977128
Sure. Here is my topology: >>977110
I bevelled the corner of the redlined area in the original image, but I'm still stuck with a round inner corner. As you can see in the image I linked, bevelling or creasing that area leads to undesirable artifacts/fucked up shading.

>> No.977196
File: 485 KB, 1296x662, Screenshot_20240310_212713.png [View same] [iqdb] [saucenao] [google]
977196

>>977067
If you converge enough edges there, it'll probably look good enough.

>> No.977275
File: 669 KB, 1376x1397, newpauldron.png [View same] [iqdb] [saucenao] [google]
977275

>>977196
Thanks anon, this seemed to be the solution I was after, seems to have minimal shading problems when everything is spaced out well enough.

>> No.977285
File: 45 KB, 748x971, T-Pose_7.jpg [View same] [iqdb] [saucenao] [google]
977285

>> No.977314

NEW THREAD
NEW THREAD
NEW THREAD
NEW THREAD
NEW THREAD
>>977313

>> No.978859
File: 218 KB, 1024x768, bizcat (1).png [View same] [iqdb] [saucenao] [google]
978859

How close did I get this to looking like a 90s CG render?