What's the best way to make anime-like waifus?
Low poly (non sculpt) or sculpt (zbrush)?
Choose your tool. It's a question of whether you like to stretch spheres or extrude cubes.
In either case, a lot of your success will be reliant on shader usage and compositing.
If you really need to ask, you should probably find a base mesh and decorate it. Waifucraft isn't appropriate for babby's first model.
Neither is inherently more or less efficient. It solely relies on your own strategies and heuristics when using either tool.
I really, really like bevels. All I want to see are bevels. Show me a bevel.
I also like chamfers. Actually, I like chamfers more. Show me a chamfer. All I want is a girl made of chamfers. But, I need to see the edges, or else I won't know that it's a real chamfer.
Smooth/Jaggy Pig Digusting. Give me pure chamfer waifus.
low poly is important for game development. don't you like the charming style of animal crossing? of course you do. it's also very low poly, especially older versions like the nintendo DS version. that was no more than 200 polys per model, maybe even way less than that.
Practice replicating visuals.
It doesn't matter what approach you take to it, if you're really interested in doing that(and if your idea of "anime waifu" is actual 2d and not the average western 3d-er's idea=pixar characters) the biggest limitations are going to be whether you're observant and intelligent enough to tell the 3d program how to show what you want it to, how you want it to, from the angles you want it to. And that requires math and programming more than anything else because that kind of NPR is based on 2d art is based on abstraction of reality, whereas realistic rendering is simply regurgitation of physical rules.
Or, if you're not that serious and you just want a little anime waifu character to mess with then yeah, go low poly. But in general, you should almost always try to stick to low poly as much as possible, moreso the better you are at programming and math(because if knowing programming/math enables you to not give much of a fuck about polycount due to the fact that almost everything is transferrable between a lot and high poly mesh since, speaking in terms of the object you're modelling as an abstraction, they're the same thing represented with a different level of accuracy).
Anyway if you've got none of the skills I mentioned and your goal is something like the OP image then don't touch high poly, I suppose. High poly is great and all but it's very difficult to manage compared to low poly, which is subdivisable in most applications anyway. The long and short of this post is that you should only work with high poly if you can afford it in terms of your technical knowledge.
These are classically floating geometry. You usually just put a separate quad where the eyes and mouth are, and animate textures for them.
This also applies for high resolution models, if you want to maintain a specific look. It's often a lot easier than lip synching and animating facial bones.
Or if you'e lazy, and don't really care about efficiency (and there's no real need to if you're just doing this for fun, or for something that's prerendered or is only going to run in realtime on a PC) you have each facial expression on seperate geometry and hide the unused ones inside the head.
Most game engines will just occlude that shit out, negating most performance concerns. Animated textures are more weight on texture memory than they're worth, sometimes, so there is definite value in just hiding expressions on toggled game objects.
I'll also post some inspirations
I'm a complete noob and I'm also trying to learn low poly for a contest.
Maybe if you sucked ass at modeling a face it would be.
Only because it's not cell-shaded. Really though he has no excuse it looks terrible.
This, a nice shader will always models look more sexually attractive than flat shading.
Did you even look at what I was quoting or were you just looking for an excuse to post a fucking tank?
Wasn't referring to all models just the ones in question. Those models that were posted have very basic textures that are little more than color gradients which you can easily replicate with vertex colors. If you're going to make a texture that basic then you may as well just use vertex colors and optimize as much texture data as possible.
Then just have a small 32x32 texture for the eyes. Some of these models have unnecessarily large texture maps for what are essentially just gradients in order to make them smooth and avoid pixel artifacts but with Vertex colors you can have perfectly smooth gradients because they're just color values.
The idea isn't to gradient every shape of the mesh, but to use gradients to draw attention to areas of interest. So generally the feet/legs of a character will be of less interest and the overall tone will become darker closer to the feet.
If you wanted to draw attention to the face for example, you could pull the gradient on the jacket tighter so it's only light at the top, and the bottom half is basically the same tone.
Btw those arms are about half the length they should be. hands general hang around half way between the hips and knees, and slightly longer arms always look less strange than too short arms.
Here's my attempt at char modelling. The hair is a huge struggle
this looks cute
maybe you should add hands and fingers to her because the arms look a bit short
Didn't really put much effort into the texture, just wanted something else than solid colors. Though I didn't know about the things you said so I'll try and keep them in mind for later.
And agreed that the arm look short. Guess it was just one of those things where I've looked at it for so long and just got used to it.
Tried some different faces, but ended up liking it better without. Might just be laziness kicking in, so I'll probably try a bit more.
I usually struggle with hair, but since this was low poly I found it a bit easier. I looked at different pictures until I found a hairstyle I liked and pretty much did what you said, started with a box and extruded while trying to accurately recreate the hair silhouette.
Agreed on the arms being too short. I'll try and add hands as well, might like it.
Reposting this from another thread
You guys might find this helpful:
>Browse sketchfab for lowpoly models
>Look for tutorials and inspiration in Pinterest. There's also tons of good shit from deviantart and tumblr.
You just have to know where to look. I'm learning a lot from just observing and training your eyes to see how topology works.
Made a shitty Miriam from the upcoming Bloodstained game
Now for some simplified lowpoly waifus
No anon, YOU'RE pretty good; in the niceness organ
Yeah I fucked up the moment I contemplated making a "generic animu" female mesh
Due to time constraints(but mostly laziness) I didn't fix it
I'm also missing two gold wristbands, the circlet earring thing and forgot to paint her left hand's SSS black since it's a glove
I may or may not fix it depending on if you guys think it's worth showing on ArtStation, because it's not up to par for to my personal standards, which some people may agree to and some saying that it's dumb to think that way
uh how do i do this
>people are gonna see my decimated no-topo garters
Does sketchfab even support multiple UVs on a single mesh and maya hair?
Tattoo is on a different UV set because I did my body UV seams without thinking and the tattoos would've been low res if I used the same UV set as the skin; also due to my stupidity
Also I can upload a higher res revolving render, but it's only 1280x720 because potato pc and deadlines
I have a still render at 1280x2000 tho
Because I posted that when I was out, basically before I had to present the model for class
But hey here, have some disappointment
I think you're confused about the workflow.
Look at the texture file on the left hand side. You simply cannot get such a clean result from painting on the mesh. Also there would be no benefit.
Can I please get a wireframe of this? I would like to know how you went about it.
Looked at reference photos of clothing I liked, hair I liked and body shapes I liked, and tried to just combine them. Also to help 'sell' it more I went into my lambert and turned the diffuse all the way down and Ambient color all the way up. That way you can only focus on your silhouette and not worry about the shading until you texture. Textures are really simple, just gradients, but it's not really that good looking so I'd suggest you spend more time on texturing than I did.
Do I need to know how to draw in order to get into 3d modelling?
It depends on what you're doing
But it's not technical skill from drawing that counts in modelling, but the understanding of form which comes with a proper education in drawing
Basically if you're making a torso, you need to know what the form is like, showing it through pencil or 3D is just the extension of your knowledge through a mean
You're gonna need good anatomical knowledge if you want to do humans, you need to know about basic joints and how plating/armour works if you're making a robot, you need to know about how many separate parts does a revolver have and how they separate, etc
I can't draw, at all, actually have a problem with fine hand motions that I was in treatment for
Doesn't matter at all
IF (big IF) I need to draw something I just do it digitally with stabilizers and endless undo functionality
>moreso the better you are at programming and math(because if knowing programming/math enables you to not give much of a fuck about polycount due to the fact that almost everything is transferrable between a lot and high poly mesh since, speaking in terms of the object you're modelling as an abstraction, they're the same thing represented with a different level of accuracy).
what the fuck are you talking about, programming and math have almost nothing to do with 3d modeling, from an artistic perspective, you don't need to know anything about programming to retopologize a sculpt, and you need to know very little(basically no) math.
>all that bullshit about knowing programming or math
Is Foxworth genuinely autistic, or is he a performance artist pretending to be autistic. I just can't tell sometimes if it's not just some bizarre Andy Kaufman-esque piece of performance art that none of us are getting.
>>You need knowledge in programming and math if you start making video games. 3d has almost nothing to do with that
It all depends on what you mean by ‘making games’. If you are talking about programming them, or writing shaders, then you are right. If you are talking about actually designing them, then that is not necessarily true. If you are talking about level/prop/asset design, it is even less true. If you are a one man show, then, yes, you need all these skills and more. Otherwise, you do your part in the pipeline, and those skills are not needed unless you are a shader artist or a programmer. Some basic maths help in PS if you know how the math works behind the various layer styles because it saves time knowing how that math will affect the pixels before you choose a given blending mode, thus saving you from having to test all the modes to get the result you want.
I have been diagnosed with autism by my psychologist, so yeah, I am genuinely autistic. Personally, people who try to be strange or interesting for "le irony xDD" or whatever make me mad, honestly. I just express the feelings I have through 1990's styled artistic means and I don't see why anyone would do what I do for any other reason. I am who I am and I have no reason to stop doing what I love. It's a hobby coupled with my feelings of love, and it makes me happy. Also, everyone should know by now that I fell in love with Yuki Nagato over Yuuko (which I announced on all sources a while ago) so that model is now invalid. I also feel stupid for putting so much work into everything I did for Yuuko and then realizing that Yuki was truly the one for me. I was blind all this time. Also, that's a blender screenshot and not from a video if it wasn't obvious enough, indicating that post as mine. In the picture is the new model I have made for my legitimate waifu, Yuki.
To understand why we must first recognize that fandom of anime characters and collectibles is a cultural phenomena that is widespread and therefore well known.
Secondly we must recognize that autistic individuals don't have any interpersonal skill sets or much of a need to really interact with another person the way you would your partner.
But what they still have is their sexuality and a need to live out their own romantic emotions in some fashion.
The illusionary one-way companionship of a imaginary friend is typical of people with this psychology, the 'anime waifu' is just the template that is projected on in in this day and age.
I speculate that '2D' girls is also highly appealing since it dehumanizes the subject of attraction making the surrogate lover seem more valid/real as no IRL counterpart exists.
If you fantasize you have a relationship with a imaginary real human woman this illusion may get shattered more easily when you become attracted to someone you actually meet.
No need to spoonfeed I already found it I forgot Google was a thing
I'm just going to read all of the issues out and determine who's best girl so I can model her
See you when I'm done anon
Art is about manipulating perceptions and playing with what the viewer expects and fills in with his own mind.
Jaggy and harsh CAN be made cute by a skilled artist. And not only that, It might need even more work and talent to do something cute with more polis.
This happens because just a line and two dots can be used to represent a perfect smile ":)" but if I ask you to use 1000 lines and 1000 dots, if have parkinson the smile will be uglier and even more jagged, even if you were allowed to do more strokes.
Blender perhaps with its horrendous UI, but any decent program shouldn't require mathematical knowledge more complex then basic algebra for calculating annoying shapes and architecture (if you're one of those weirdos that models environments in the maker itself rather then assembles it in whatever program you want it rendered through).
Keep in mind that the lighting engine in whatever you're rendering the low poly through is very important in regards to how the model will actually look as a complete render: good lighting can largely mask the fact that a low poly model is well, low poly.
Blender certainly doesn't require you to understand advanced math to model. I don't know where you're getting that from (although, I suspect you're sitting on top of it).
You can certainly _use_ advanced math in Blender. You can write shaders or do python scripts that use it. But modeling? Not at all.
It's probably just one of those "le blender sucks" anons that sits around this board. Now granted, blender does have a DF-tier UI, but hey, it's free and unlike various other programs won't beat down your door for using it to make commercial stuff.
I want to make video games but i just started out in blender, is this good low poly or nah?
i'm surprised i can still find my stuff here.
best way is lowpoly with baked normal maps from a sculpt so you may use that for anything you want.
i still prefer lowpoly with level of detail reaching between a ps2 and ps1 polycount.
you can do the classic unwrap detaching certain parts of your character or you can just go lazy and ask your texturing software to unwrap automatically, but that means your uv islands will rotate and you'll be able to spot pixels going different directons. (notice the pixels on the leg going 45°, that can look bad on square shaped objects with your texture going a different direction than your shape)
for that character i just asked 3dcoat to mirror unwrap the body so i could paint a mirrored texture for it.
you can paint it a few ways, but i like texturing my models in a texturing software that lets you paint directly on your mesh, gives you a better idea on how you want to do your lighting on you texture, like when you paint a fig. i use 3dcoat to do that, its relatively easy to learn and is layed out a little like photoshop. you can also use mari or zbrush.
3d coat also lets you paint normal maps, so i added a little bit of volume on the hair, so now it looks kinda like those models you can have on 3ds, still pixelated but with some nice render. the nintendo ds had no complex litghing or shading
as for the size of the texture, you can choose wich system you wanna work on, nintendo 64 only had 64x64 textures, i found that out by ripping n64 models and EVERYTHING was made of only 64x64 textures, its crazy.
basicly the more powerful the system is the bigger. but most of time i just do 512x512, most 3ds games do and also phones games use it. 1024 is a max.
you can also resize some parts of the uv to have more pixels, ex. for the face i the made the uv 2x larger
I use Substance Painter in the past, not heard of 3D Coat. I thought that programs like that would be overkill for lowpoly models, but I suppose there's no real reason it wouldn't work out, It's certainly an easier workflow than trying to texture pixel by pixel in a disjointed program like MS paint.
I've ripped and re-textured N64 models in the past (pic related) but they're easy cause someone already did the UVs and textures about 20 years ago and all you need to do is rebuild them. Trying to make my own shows I've still got a lot to learn.
yeah.. making uvs like that require you to anticipate where the texture is going to be and its fine for flat shaded textures like that but when you want to do complex shading it just becomes too hard to guess the shape of your model from you UV and you just start twisting your mind.
they're not reall overkill (they allow me to make pretty stuff), only if you want to make the models like they were made at the time.
this character in the pic has a real uv like those n64 uvs, but i still painted it on the mesh like the other, i think is just way more fun.
Not him, but you should really learn how to reverse google images, it's right there
Reverse google image, first result should be Danbooru, artist name is stated in the tags
Just letting you know, it could save you a shit ton of time
Oh fuck isn't this from cortex command ? That is cute as fuck.
Do you have more ?