Thread #1025537
Anonymous
/SQTDDTOT/ Stupid Questions That Don't Deserve Their Own Thread 02/09/26(Mon)21:09:28 No.1025537
/SQTDDTOT/ Stupid Questions That Don't Deserve Their Own Thread 02/09/26(Mon)21:09:28 No.1025537
/SQTDDTOT/ Stupid Questions That Don't Deserve Their Own Thread Anonymous 02/09/26(Mon)21:09:28 No.1025537 [Reply]▶
File: Untitled.png (273.8 KB)
273.8 KB PNG
.
142 RepliesView Thread
>>
>>
>>
>>1025539
Used to be a regular thread. Questions like how do I get free models, how do I download jpg, how does rotate work, etc, all trivial "how do I" questions that clutter the board for zero general value go in there where they are easily answered and found by further questioners while keeping more interesting threads from falling off the board
>>
>>
>>
>>
>>
>>
File: Screenshot_2026-02-13_19-33-35.png (923 B)
923 B PNG
>>1025713
I can only bone select if I'm bone-select mode, which I guess makes sense. But I only need LMB for that. I want to select bones while in the other 2 modes. It's gay having to keep switching. Holding alt or ctrl doesn't work in those cases.
>>1025715
You have this problem too?
>>
>>
>>
File: eye-bn-728x336.jpg (27.7 KB)
27.7 KB JPG
I have a cartoon eye I want to morph with shapekeys, but I don't want to distort the pupil texture. How can I do that? I don't want a procedural pupil either. Just a texture.
>>
>>
>>
>>
>>
>>1025537
>>1025537
in UV unwrapping, how do you set the selected orange things to be in a precise unit or scale? i can't get the life of me to align it on a texture. say there is a 512x512 part in a 1280x1280 texture that i want to fit the orange box to, how do i line it up perfectly?
it's all scale and rotate and grab and distort everything and everywhere in blender. I know i am doing this wrong or using the app wrong, but can anyone enlighten?
>>
>>1025930
Assuming Blender
In the UV editor press N and you will see menus on the right. In the Image tab you can see the coordinates under UV Vertex. You can type in the coords into X/Y or move the vertex manually.
By default uv coords are displayed between 0 to 1 so if you want pixel coords, go to View tab and check Pixel Coordinates. Keep in mind the origin starts from the bottom-left, not top-left.
I wouldn't worry too much about "perfect" alignment anyway, nobody will be able to tell if you just eyeball it. There will always be some slight distortion unless all your faces are perfect squares and you are a god.
>>
>>
>>1025961
awesome, thanks! i'll try it out later. and yes, blender. trying get some pixel art seamlessly wrap in some kind of tile and coming from a 2D background, i guess some old habits still hold me back. im just too anxious
>>
>>
>>
File: téléchargement.jpg (9.3 KB)
9.3 KB JPG
Besides CGPersia, The Course Busters Telegram and online-courses.club, where can I find premium courses for free?
I'm specifically searching for this one: https://www.wingfox.com/c/9248
>>
>>
>>1025981
Let's say I have a one-off project as a VFX artist and I want to pick up any program to retopo a complex ZBrush sculpt with, for this one time, as fast as humanly possible.
I've experience in ZBrush & Maya, but have lost all the memories (like quad drawing) and can go in any direction.
>>
>>
Okay so this is actually exactly what I was looking for since I've got a simple question, does anyone have examples of what good seaming and UV mapping looks like if I want the most bang for my buck in terms of texture quality relative to texture image size
I'm better at learning by looking at examples over watching tutorials, though I wouldn't mind tutorials too if you guys have a good one you would suggest
All the links that relate to information in the UV section of the pinned thread for the board are dead, so that doesn't really help me
>>
>>
File: when life hits.jpg (82.3 KB)
82.3 KB JPG
>>1025537
has anybody successfully installed reallusion character creator 5 and iclone 8?
is there a portable version without doing all the hacky stuff?
>>
>>
File: fo4hazmattexture.jpg (558.4 KB)
558.4 KB JPG
>>1025537
How do I go about changing the colour of a texture without impacting stuff like dirt and stains on it?
>>
>>
I started only a couple weeks ago. Does anyone have advice on how to do irregular shapes on faces? I am trying to make an intricately shaped keyhole for example (not just a trapezoid cutout with a circle cutout on top), but am struggling. The best way I can think of at the moment is to draw the shape into the mesh somehow and use a mirror modifier so it is symmetrical. And even in my plan I am not sure how to do that correctly either. And I want to avoid shrinkwrap and boolean fuckery if possible.
>>
>>
>>
>>1025537
How the fuck do you rig and animate an eel? I've modeled a dumb simple eel in blender but have been at a loss for hours. There is next to zero tutorials online for how to do anything 3dcg for any sort of marine life, let alone eels. I could use snake rigging techniques but those don't exist online either. I ask Chatgpt to scour the internet for me and it is at a loss too. I just want something simple so I can make a png sequence and use it as a sprite sheet.
>>
>>1026247
Nevermind I made something satisfactory. Just extend one extra bone on the end of your chain and turn off deformation for it. Then apply the damped modifier to the one right behind it and make it target that end bone. Then go to the bone behind the modified bone and give it the same modifier, only targeting the modified bone in front of it. Repeat this for the rest of the chain all the way to the base. You can do this while modifying or go back through after, but you then need to make them go down in influence approaching the base. Highest influence on the bone on the end that targets the unmodified non deformative bone, lowest on the base. Then you can move the base and have it ripple in a wave behind it, I'm sure you could tweaks this to your liking and experiment with influence. Also, if well modifying you get an issue where bones start pointing towards each other and look weird, just undo it and retry to select your target. This is a simple yet versatile rig that can work for eels, snakes, tails, some tentacles, and anything similar. Only downside is that you can't export this but you can always bake animation and you'd likely want to do this procedurally anyway for a 3D game. I personally needed it for a 3D visual rendered as a 2D png sequence for a spritesheet so I didn't have to worry about any of this. Never give up anons.
>>
what is efficiency?
like, if i can add detail with more verts or more materials/textures/shaders, what is the rough exchange rate?
i know it's a dumb question because it varies by engine and implementation, so lets say its a game that has to actually run on normal hardware. For another example, i could use a transparent texture to make a leaf, but i could also use more geometry. which way gets me more leaves per scene?
>>
>>
>>1026364
Duplicate the box model and sculpt it full of details, then bake that down
>>1026363
Alpha blended transparency costs tons more than alpha testing/dithered transparency which costs more than no material transparency.
You'd think that means you should overwhelmingly favor tons of small polygons but on GPUs specifically triangles that are very small on-screen cost proportionally more resources so they should be avoided for real time rasterization. They're better for ray tracing though!
More different shaders cost more.
Triangle counts aren't that much of a big deal unless you've got costly modifiers or tons of models
Instance all you can as GPUs can accelerate that massively.
In Blender some modifiers, especially subdiv, perform much faster if they can be done entirely on GPU that means no other modifier, constraint or parenting that needs to read back from the post-subdiv mesh. The relevant modifier will display a warning that it can't run on GPU. That means if you use vertex parenting on a subdiv mesh it might be better to just duplicate the mesh without the subdiv, make it invisible, and parent to that. Or just set subdiv to 0 in the viewport when you need realtime feedback
>>
>>
File: ballarmarature.png (410.2 KB)
410.2 KB PNG
Im doing an animation course, and when i load this ball.blend they sent, it loads normally, but when i append the collection in a new file, it creates this WGT-ROOT, WGT-ROOT.001 and circle objects, why does this happen?
>>
>>1026388
Depends. There's usually two versions of those objects: The ones actually used on the rig itself, and the ones that are just there to serve as reference for the rig's controls.
Click and see if you can get in pose mode from there. If there's no option for it, then it's the object. Hide or delete that as it doesn't affect the final rig.
As for why it happens, I don't know, I personally haven't found a workaround for it, so I assume it's just a "feature" of Blender's that they don't plan on fixing anytime soon.
>>
>>1026388
Since this is deffo the p2design ball rig, make sure to watch chapter 02_15 to know the difference between linking and appending, as it'll explain exactly why you can see copies of all the used widgets.
Basically because you appended, it copied every mesh widget it'd need to *edit* that rig in the scene. Thoses extra meshes are the meshes referenced by the custom bone widgets, hidden in the original file.
Instead, you should link, not append, the BALL collection from that ball file, and it'll link it to the scene in a configuration where all those auxiliary meshes are in a hidden collection. Then right click the collection in the outliner and choose library override->make->selected and content. Then you'll be free to animate in pose mode.
>>
>>
>>
>>
>>
>>
>>1026002
Basically efficient mapping can be done with SmartUV unwrap and then tweak so that important areas get more real estate than bland/unimportant areas. If most of your model is symmetrical, mirror that part so the texturing is more efficient. Just make it one square texture if you can for portability.
>>1026032
Make an outline of the hole with vertices, extrude out, merge vertices so it's a hole in a plane and extrude down to make the hole deeper.
>>1026363
Less polygons is usually the key unless the textures are huge or harder to manipulate. If you're making shit for games you'll learn how to make it in such a way that it's easy to decimate the poly count for levels of detail while maintaining the basic look.
>>
File: ezgif.com-optimize.gif (2.8 MB)
2.8 MB GIF
>>1025537
How do you get trains to follow curves properly? Especially ones with bogeys or tenders, they seem to all react differently. Yes they're all following on the same axis, all the origins are where they should be pivoting.
Weird place to ask something like this, but idk where else to go. And apologies for the shit gif quality, had to compress it so my demonstration would fit.
>>
>>1025537
Whats the proper workflow for animating game animations? How do you animate boring loops like characters leaning over a desk, standing and waiting for a response, etc, and have it not look overly static?
Should I be doing longer loops w/ less variation, or many short loops with lots of variation (3-5 variants per "action")?
>>
How in the FUCK do I even advance in environmental 3D modelling/design?
I understand the blockout/greyboxing phase but once I reach the baking and texturing I'm pretty fucking lost.
Everywhere I look, I can only see blockout tutorials and tutorials about adding props, details or instances.
The part in the middle seems to be extremely gatekept for some fucking reason.
>>
File: tv.png (1.4 MB)
1.4 MB PNG
so i modeled this simple tv object.
Now my next step is to export the png of it so that i can import it to a game i made this for.
The texture has two materials atm. The body of the tv is grey. The second material is grey, but i lowered the alpha channel to make it see through.
So i want to export the the texture of the object as a png so i can import it to the game.
i asked chat gpt for help and it had me bake the texture and unwrap the uvs using smart unwrap. But the issue is that when i export it from the image editor, its just a black square. How do i bake a material with a lowered alpha channel in addition to the material for the rest of my tv object? Goal is to have a png of the texture that has the lowered opacity of the tv screen and the color of the tv. thanks.
>>
>>1025537
Is there a polyquilt version for Blender 3.6, or something equivalent to it? I'm looking to speed up my modeling process a bit and I'm tired of doing everything by hand. My two most consistent time sinks are rigging and retoppoing (texturing too but at least that's at least semi-interesting). I'm planning to use rigify to address the first issue but the retopology market seems to be a bit more contested.
>>
>>
>>
fa/tg/uy here.
What is a reasonable price range for commissioning 3D sculptors to make STLs for 28mm miniatures for printing?
I decided after getting a 3D printer and Wash and Curing Station to try commissioning a 3D printing sculptor for two 28mm miniature STLs.
He said he was gonna charge seven hundred plus US dollars for those two models.
Now I know that 3D sculpting and with the added caveat of STL printing in mind means a big thing, but I wasn't expecting this.
I think I am in the wrong for complaining about the price but at the same time I'm wondering if this is actually the reasonable price.
Wish I was a Saudi Arabian oil baron's kid so I didn't have to do normie shit like being financially responsible.
>>
>>1026803
I'm sure anons here will crucify me for saying this but the truth is that if you wanna make your own custom minis and don't plan to sell them on a marketplace, just learn a solid 3D gen workflow through ComfyUI.
>>
>>
File: body pose progress.png (542.2 KB)
542.2 KB PNG
I made my own base body model, and now I’m trying to create clothes that match a nun character reference. I’m having trouble making the clothing fit properly especially because the outfit is fairly form-fitting, and the clothes I make keep looking too baggy, bulky or the character's body "tears" through the clothes. Since this is my first time making a model, would it have been better to start by drawing a turnaround/reference sheet in something like Krita first? I’m wondering if that would make it easier to try to build the clothing as one connected mesh instead of trying to add everything modular afterward.
>>
>>1026803
>He said he was gonna charge seven hundred plus US dollars for those two models
You'll have better luck asking on the printing thread on /tg/, but that seems a bit high for 2 figs, unless the artist has some clout
>>
>>
File: photo_2025-12-16_17-21-30.jpg (229.3 KB)
229.3 KB JPG
Sup /sqtddtot/
I'm looking to get into 3D for FDM printing, which software would be best to make stuff like this?
I know that the knitted effect can be easily done with Blender, but what about the base model?
I don't really care about polygon count, just need a good way to make a clean model and separate it into pieces (like separate colors, limbs, etc).
I'm a 3D /beg/ who did some Blender tutorials and am taking a look at Zbrush right now, but I don't know which would be best for that need
>>
>>
>>
>>
>>
File: Stuck.png (126.5 KB)
126.5 KB PNG
Following Issue:
- Mirrored (X-Axis) Character
- Armature, correctly named (.L and .R for their respective sides)
- All transforms applied (Object + Armature)
- Automatic weights applied
- Weight-Painting X mirror is off
- Mirror modifier is beneath the armature modifier
Yet when I move my left leg, my right leg also moves. This is a bug I have encountered a few times already but I can't remember the fix right now. I know that this is supposed to work, I have multiple other characters where it does.
Does anyone have any idea why this is happening?
>>
I want to learn how to make human characters but when I search up tutorials on Udemy the authors all sculpt a character first... Is this the new way of doing this? I remember back in the day you had four perspectives and two or three reference images t-posing, and you'd go from there... Is there a specific name for this kind of modeling? I'm intrigued by sculpting, but it feels like a whole nother, in depth step I'd have to undertake to get to my goal instead of using using 2 different reference images.
Any help or suggestions?
>>
>>
>>
>>
I have a question about whatever the version of "tracing" is but for 3D art. Let's say I like a guy's personal model of a preexisting character so I took screenshots of his character and box-mapped it in blender would that be "3D-tracing" or would that be just references?
>>
File: kitty katswell.png (639.7 KB)
639.7 KB PNG
>>1026893
Is it this character?
>>
File: claire refrence model.jpg (893.6 KB)
893.6 KB JPG
>>1026894
No, it's Claire, here's my references I've gathered so far. I think this should be enough wanting to take main influence from the middle's example. You reminded me of the danny phantom fortnite collab and looking at the 3D models people made for her why does Butch Hartman's characters translate so poorly into 3D?
>>
>>
>>
Is UI lag on Win11 for 3dsmax (2022) a common problem? Every time I do anything whether it's select an object, right click for the menu or add a modifier there is a 1-2 pause that's getting really annoying to deal with.
>>
File: rightclick_2.mp4 (2.7 MB)
2.7 MB MP4
>>1026938
Video to show what I mean.
>>
>>1026715
>>1026920
Here, this guy makes a lot of tutorials recreating Nintendo models.
https://www.youtube.com/@TheObservatoryShow
https://www.youtube.com/playlist?list=PL5F48bE3OZbvnG-zWvuqmMeIkCq5_1H _p
>>
>>
>>
File: .jpg (35.1 KB)
35.1 KB JPG
how did they do the upper eyelid here?
https://www.youtube.com/watch?v=LrO4myCqNsE
it's completely flat in the model. I assume just baked down. are there any limitations to this? it's extremely rare to see no eyelid.
>>
Hey guys, sorry that this is more of a weird, silly question for /gd/ than /3/, but that board seems almost completely dead and fossilized, so I also wanted to ask you. I'd like to get into some kind of very, very casual "generative" (not AI) art. Kind of like what game devs do to generate and modify abstract art, beautiful landscapes, mechanical things, architecture and stuff like that based off of certain mathematical functions, but simpler. For example, right now I'm messing around with Excel conditional formatting, making neat, abstract patterns to kill time at work (I know that this is incredibly autistic and retarded, but that's the only thing I have at my job and it's easier to hide when I'm slacking a bit). Is there something more powerful than excel but still casual and simple to use compared to something like photoshop/blender and other "serious" programs?
>>
File: eyelid.jpg (16.6 KB)
16.6 KB JPG
>>1027036
But there clearly is an eyelid? What do you mean it's flat? Are you okay anon?
>>
>>
>>
>>
>>
>>
File: Capture.jpg (42.7 KB)
42.7 KB JPG
>>1027049
uhh no? this is an eyelid. it protrudes.
>>
>>
I asked this in another thread but i'll ask it here
I haven’t kept up with the progress that AI has had for 3D asset making. How capable is it? Is it still worth learning if I want to make money off it?
I don’t want to sound like one of those people who coped when AI art was in its infancy and we thought the pictures of Master Chief standing in line at gas stations couldn’t possibly make art as good as a well trained artist, but even with my pessimistic view on how AI is going to ruin the internet, I just can’t see the technology being able to replace a 3D artist. It would probably make the process easier in some ways but you probably are going to need someone to use it until we have AI that can literally think like a human with the full knowledge and intuition of a 3D artist, which if it existed, it’s probably way too advanced and expensive for most people to afford
I’ve seen those AI programs that can take an image and generate a 3D mesh, but even when you take account that the topology usually sucks and needs manual editing, you still have to UV unwrap it carefully so it doesn’t show distortion, texture it realistically or fine tune it to fit an aesthetic, bake features onto it for extra details, then put in a skeleton and weight paint it in a way that prevents unwanted deformation and achieves specific effects like jiggling or rigid movements in certain directions
I don’t think an AI could look at a chubby character and rig it in a way that an autistic but wealthy furry would want, or see a robot/mecha and precisely place the bones and restrict their relative movement that the design implies.
Can someone cut the doomer mentality and tell me what AI is actually good for with 3D asset generation? I have no problem adopting it as a tool to make my workflow more efficient if it won't significantly cut into my quality
>>
>>
>>
>>
File: 0001-0240.mp4 (2.4 MB)
2.4 MB MP4
>>1026986
Super cool, thank you!
>>
File: 1774170144643732.png (53.5 KB)
53.5 KB PNG
>>1027103
This is so good omg.
>>
File: turnaround_bsdf.mp4 (2.5 MB)
2.5 MB MP4
>>1027107
Thank you! here's another turnaround
And to stay on topic: is there perhaps a way to transfer the weight painting from a regular armature to a rigify armature? this model's rig is as primitive as it can be (it doesn't even have IK or anything of the sort) and it's kind of a pain to move around. I'd like to have a more sophisticated rig to have better control when posing it and animate it eventually
>>
File: flcl baseball.jpg (304.7 KB)
304.7 KB JPG
If you're a newbie to 3d Modeling and look up references in a artbook but discover that there is no proper "turn around" reference for the character you want to try to make without a side profile what are you to do? If you genuinely can't draw for shit and don't want to pay someone to make a turn around. Would the best course to try to make your own turn around no matter how shitty? try to find other outfits or references that are relatively proportionate to the character? I ask in this case because if you look at the artbook for FLCL Haruko's proportions are rather inconsistent when it comes to her different outfits. Though this more "casual" wear Haruko seems like it would work?
>>
>>
>>
File: BedFuckery.jpg (159.2 KB)
159.2 KB JPG
I have been trying to learn, and I am trying to make the back of a bed like in the image on the right, but I am not sure how to do it properly. None of the tutorials on youtube even bother with this kind of detail. How can I make it look better? Or at the very least closer to the reference image without that damn point at the top.
>>
File: projectioncage.png (522.8 KB)
522.8 KB PNG
Anyone know what causes these baking defects in Substance 3D Painter when using a cage? I've already checked that the cage is clearing no problem against the high poly mesh with no casting errors and it renders okay in 3dsmax using the same cage but in Substance these defects appear.
>>
>>
>>
>>
>>
>>1027136
Be aware that, for 2D turnaround sheets, they don't always perfectly translate to 3D forms. In particular, anime faces have slightly different physical designs "face-forward" vs "side-on." Some advanced modelers will use automated shape keys tied to the camera's position, adjusting the skull shape on the fly.
If you're a newbie, don't sweat that sort of thing yet. Get the basics done. But be aware of the issue so you don't drive yourself crazy trying to get the face just right.
>>
>>
File: [Erai-raws] Akuyaku Reijou Tensei Ojisan - 08 [1080p HIDIVE WEB-DL AVC AAC][37859944].mkv_snapshot_12.39_[2025.02.27_19.53.47].png (3.3 MB)
3.3 MB PNG
I am kind of stumped on how to make some of the details. A column is simple enough, but I have no idea how to make the carving details on the wood or the repeated details on the walls.
>>
>>
>>
>>
>>
>>1025537
Not completely a 3D question but I'm trying to replicate slit scan effects in blender like the star gate in 2001 space odyssey but I've seen two methods 1) replicates the effect as they did in camera and 2) with two planes and a camera moving down a "hall" but what I want to know is if the moving down a long hall really the same result?
>>
>>
Blender. Made an action in the dopesheet. Idle and Run. Run is 16 frames while idle is 30. When i switch between them, blender keeps the old timeline range, like wtf. how do i, when i switch to a new animation, get fucking blender to set the timeline range to the last frame of the animation?
>>
>>1027455
An action doesn't inherently have a range, you have to set it in its properties (because for a cycle like a walk there might be a need to have keys outside the playback range just to getthe right curves), but beyond that there's also currently no command to set the scene range to the action range. You can use this though: https://extensions.blender.org/add-ons/action-to-scene-range/
An alternative for working on multi-action characters is to make one scene per action you're working on, that way you can easily change range and action in one click and also have a test/review scene where you're mixing action strips
>>
>>1025537
trying to rig a disembodied dick for game animation.
I was planning on trying to use modifiers to follow or at least point toward bone(s) that are part of the base female model, for use in making sure things line up properly, allowing for unique dicks/dildos to be swapped out.
Some questions:
Should the root be at the base, or in front of the tip?
any good videos on the subject? I had a few videos bookmarked, but Youtube took them down.(They rigged a banana and showed it *slightly* deforming a default toroid)
>>
>>1027461
Some links there: >>1024300
For the dick, put the root at the base, but basically the idea from that vid you lost (animenyan's) is to use a spline IK chain, then the curve itself has 3 control points which you'll hook to controller bones, root, middle and tip. This makes in so the middle is freely controllable but the dick will always pass through it, so you can link that to orifices or hands.
>>
>>
>>
>>
>>
File: gun.png (344.5 KB)
344.5 KB PNG
I want these to be separate objects but what would be the smoothest way of having the wooden handle fill in and "connect" to the metal or vice versa, as in are there any special modifiers or tools that could be used that i dont know about that arent just trial and error manually moving the vertices until it looks "right" but actually isnt cuz its been bent at an invalid angle or other errors like that
>>
>>
>>
File: 3d.png (2.2 MB)
2.2 MB PNG
How do you tell good topology from bad topology?
Pic was something I worked on a while back. When I posted the bottom anons told be it was bad.
I redid it copying the topology in a Capcom PS2 game (except for the breasts, you can see the splice there). It's way fewer verts but idk.
Ignoring the body forms themselves, which would you say has the better more usable topology?
>>
File: Screenshot_2026-04-12_12-18-45.png (170.3 KB)
170.3 KB PNG
>>1027516
Top is better but they are both shit
Good topology prioritizes even spacing and quads and has good "edge flow" i.e. follows the contours. Good flow is also good for deformation. If you have random faces breaking up your loops, it's not going to bend well.
You have so many random edges that don't add any detail and ruins loops. Your collarbone/neck is littered with tris that can be simplified to quads. The face below that is massive and should be divided to match the density of the surrounding faces. You have a pointless loop between your boobs. So on and so on. There's too much shit here to correct, better to start over. Idk how you accomplished such a mess. It's like you took the knife tool and went to town with it.
To better imagine good topology, you should separate your mesh into areas of interest (joints, loops, change in direction). Then fill in lines while prioritizing quads, spacing, and contour.
>>
File: Outfits2.jpg (89 KB)
89 KB JPG
>>1027520
the topology is basically copied from a capcom PS2 game
I understand that game engines turn everything to tris, and alt+J doesn't quad everything properly, but I noticed a lot of older models have seemingly random topology outside if key areas like the joints and face... idk if it even matters for this level of detail, or if I'm misunderstanding something fundamental about topology
>>
>>
File: Screenshot_2026-04-13_10-53-02.png (668.9 KB)
668.9 KB PNG
>>1027539
>I understand that game engines turn everything to tris, and alt+J doesn't quad everything properly
Yes. So you are aware the model is likely messed up. So why the FUCK are you studying from this model?
The devs worked in quads, the engine converts it to tris, then the tools used to rip these models (like ninjaripper) aren't 100% accurate, and then you convert to quads again. There is a high chance this model is fucked. You shouldn't be relying on ps2 rips to learn the basics of topology.
>I noticed a lot of older models have seemingly random topology
How can you say that if tri->quads fucked everything up? You haven't noticed shit and you clearly don't have the basics down to comprehend what you are even looking at.
I found the model of your character. The only place I could find it (https://www.deviantart.com/goodgamer280/art/Fiona-Cowgirl-986300474)
I used tris->quad and yeah I can say this model is kinda fucked. Unless you know how to clean it, which you DON'T. It leaves a lot of random tris and isn't symmetrical. Even then, it looks like you took some liberties and made it a little worse because you don't understand which edges should be erased. I hid some clothes, but as you can expect from these optimizations they leave giant gaps. This is a bad model to study from.
My advice, go watch a real modeling tutorial. Or continue down this retard path, I don't care anymore
>>
File: fake-anime-mouth.png (334.4 KB)
334.4 KB PNG
>>1027547
Here I made a guide
It's based on this (https://en.wikipedia.org/wiki/Hollow-Face_illusion). It's the same principle some anime models/costumes use so the eyes are always looking at you without any animation. They do this by placing the pupils farther inside the eye cavity. In this case it's the mouth.
This only really works well without shading though. In my example I inverted the whole sphere, but when you are actually modeling a human, you'll want to carefully select just enough faces around the mouth.
>>
File: recording-face.mp4 (386.7 KB)
386.7 KB MP4
>>1027553
This is what the end result looks like
>>
>>1027553
Thanks for the guide anon
>This only really works well without shading though.
I'm not to worried about that I was primarily planning on doing shading in compositing, I'm not planning anything for a game primarily animation and renders. But I am curious would shading be broken if you gave it a SDF but inverted the mask so it reacts opposite of the direction of the sun?
>>
>>1027539
>I understand that game engines turn everything to tris
No, game engines need to be fed tris in the first place so it's the tool chain that converts them into tris. It is not a mysterious destructive process artists have no control over, it can be done with intentionality.
The lower poly you go the less relevant the quad approach is. The quad approach is great because it leads to great subdiv, but if subdiv is not part of the workflow you'll have to learn to work with triangles to get the deformation you want. And the fundamental of topology is about that, getting thinsg to deform right
>>
>>
>>
>>1027580
I'm not saying it's a bad model, or that you can't use tris. The problem is anon was using it as a study piece for topology and failed to clean it up properly because his understanding was insufficient. There are way better base-models and tutorials out there he could use instead of fixing a ripped ps2 model. Also he said to ignore the body forms which makes his topology look worse than the source because he's not using the tris to add any depth.