/3/ - 3DCG

>>597180
tfw your shitty 3D app has no plugin for face weighting

Post best fwn plugins for 3ds max please. I've tried a few of them (FWN MAXScript and GetVertNormalsFromFaces MAXScript) but they don't work, or I don't know how to fucking execute a script properly. Also, I can't believe I can't find a tutorial for installing scripts in 3ds max which isn't like 5+ years old.

>>597180
>>Angry monkey noises<<

>>597185
I am curious, which one?
>>597192
Try this, it was recommended on polycount:
https://github.com/ByteHazard/wnormals
Not a Max user, so i can't help you further.

>>597201
Why not use something else but the shitty default texture painter?

>>597192
I use two scripts, the first is the one by Martijn Buijs on github, the other is the GetVertNormalsFromFaces script.
To install the first one, copy the .ms script file into your Program Files/Autodesk/Max/Scripts directory and in Max create a new script with the following code:

macroScript macro_ToDo
category:"WeightedNormals"
internalCategory:"WeightedNormals"
tooltip:"WeightedNormals"
(
local sFile = (GetDir #scripts) + "/weightednormals.ms"
if (doesFileExist sFile) then (fileIn sFile)
else (messageBox "Unable to locate the script.")
)

Then under Tools press Evaluate All and now you can add it somewhere to your interface or hotkey it via customize interface dialog. Running it makes a box appear with just one button which when pressed automatically adds a mesh and normals modifier to the stack of a selected object. Assuming your model had one smoothing group, it should average the normals correctly. Be aware that it only weighs the normals, so it still retains some level of "softness" to the shading of the object, and depending on the structure of your object or intent, it may not always provide a good result, which is why there's script two.
In order to install the second script, you just drag it into the viewport and it'll be available under customize. This script is a little more tricky because it ONLY works on a base editable poly object with no modifiers, in other words you should first make a copy of your mesh (so that you can always remove the chamfer modifier later), collapse it and then go to work selecting the faces you want hardened.
If there is a gap between two selected polygons, it will set the normals of each local selection, meaning that you can run this script on an entire object if you want to, so long as the only faces that are touching are the ones you want to have shared smoothing.
To use this script optimally, check out the ribbon tool in pic related, set it to normal only and also hotkey it, it'll make this process easier.

>>597215
With that out of the way, some tips:
- Remember to turn off backface culling! You may get flipped normals now and then, but won't be able to tell because when doing custom normals, they'll overwrite the result. The only way you'll be able to tell is if the backfaces are transparent. Flipping the normals into the right direction doesn't break your custom normal edits, luckily.
- Try and make all chamfers using the modifier, you never want to make them permanent as they can be a bitch to remove later should you decide to change something... and If you have to collapse the mesh, make a copy first.
- The best chamfers are obtained from ngons; any time you have a curved edge, remove any edges leading to it, as they may skew the chamfer in that direction. An exception is when the edges are directly parallel to each other, such as at the end of a tube.
- You may sometimes get hanging or overlapping edges, particularly after chamfering the object, dealing with them is as easy as selecting the entire element and pressing "Retriangulate". It's suggested to do this before weighing normals.
- When cutting holes into objects, preferably connect them to surrounding geometry along a clean edge and not from existing vertices, or as per the rule above, it may skew the shape of the hole.
- Intersecting two curved surfaces like cylinders is nearly impossible to do in a way that looks good (say a round hole through a round surface), if you can help it try and make curves meet up with flat faces. If you really have to do it though, you'll have to fall back on making a polygon shell over the edge and applying a normal map to it.
1/2

>>597218
2/2
- Cylinder edges may require a two-segment chamfer to avoid stair-stepping artifacts. The tighter the curve between two edges, the greater the need for this.
- Don't be frugal with edge count on curved surfaces, you're saving a lot of overhead with this method, so it would be dumb to ruin the effect with faceted sides. For small objects 16~24 edges are fine, but 32~48 are preferable for larger objects, and 64+ edges for very large structures.
- To clean up ProBooleans quickly, remember to check "no edge removal" and after converting back into a poly, under the Selection ribbon menu, in vertex section mode, look to the right side and use By Numeric = edges: 2 to isolate stray verts, and ctrl + backspace to remove them.
- Custom normals are rather fragile and will reset or break if you use various modifiers or tools such as detach, slice and symmetry, so save normal edit operations till the very end of your modeling phase. Even certain selection tools like shift-clicking in poly mode to select a loop of polygons will reset your normals.
- Attaching meshes together that have custom normals may screw them up, because the normal vectors are in local space, so remember to reset the XForm of all objects that you intend to merge.
- If you are building modular components and need to match a curved surface to a flat one, apply the Edit Normals modifier to both objects and manually tweak the normal vectors of the curved surface to be superimposed on those of the other, and visually inspect the blend.

I'm sure there's other stuff to keep in mind, but that's all I can think up for now.

>>597219
Next up are a couple of tricks you can do. The first is that you can extend out the edges of an object to produce an extra chamfer that will allow it to blend to an adjacent surface as though it were one part, and then remove the outer part as it's no longer needed.

>>597220
A slightly less obvious trick for optimizing surface details. First you model out some surface trim using the usual method, remove the bottom polygon(s), chamfer the object, and set normals from faces for all relevant sides.
Then you delete the bottom-most loop, and pull the chamfer loop down to where the bottom used to be. Visually it will look nearly identical, but with an entire loop of edges less geometry.

Oh yeah, before I forget, remember to set up your chamfer modifiers with From Smoothing: Unsmoothed Edges, and use smoothing groups on your base mesh to control where you want chamfers to occur. Once you're happy with the result, set Smooth threshold to 180º and weigh the normals.

>>597221
Why are you so good to us? Screenshoted the replies, thank you anon.

>>597222
Because why not?
Also, I made an error in >>597218 were I wrote to turn off backface culling when I really meant turn it on. I sometimes derp when it comes to this because I think of it as turning off the back faces, not turning on the culling.

Thinking about hairstyles

>>597228
Not sure if male of female...
What is IT?

>>597229
No degeneracy intend. I'm thinking of space suits.

>>597230
Then just do the helmet and one of those tight suits that also cover the whole head. Voila, you now don't have to make hair.

I wanna make a comfy mobile game

>>597119
no problemo, I'm interested in seeing the improvements you make

>>597240
I like it already

The man with the bunny's back, again... I was about to upload her model on Sketchfab, but this happened. Who cares, it comes with a dupe which is the more raw shit for the model.

Enjoy, you odd sadistic motherfuckers.
https://sketchfab.com/models/c437fa124a8045d0ac117f8dced3f85f

>>597240
I feel comfy already

>>597240
Needs more style. Looks neat and tidy for sure, but lacking something - if it stays looking like this it will end up being totally forgettable.

Think Transistor, Monument Valley, or Fez - they all have an easily identifiable 'thing' going on,I don't mean in terms of game mechanics but in terms of style and appearance.

Transistor; art nouveau / art deco styling
Monument Valley; acid-trip color pallette
Fez; fastidious pixel art, vibrant color pallette

If your look doesn't have a 'thing' - you got nothing.

>>597240
Why waste any effort on mobile shit?

>>597292
It makes more profit than desktop gaming, you know

>>597293
https://www.engadget.com/2015/09/18/konami-calls-it-quits-on-aaa-console-games-reports-say/

>>597293
Aye. See above. Sad times.

>>597294

https://www.engadget.com/2015/09/18/konami-calls-it-quits-on-aaa-console-games-reports-say/

Was actually intended for: >>597292

I've struggled to focus on one thing at a time recently. Trying to assemble a small team to make some baby steps towards making a game.

Last thing I made was a tileable brick texture, so I could learn the basics of Substance Designer, and experiment with my idea of having somewhat realistic reflections on surfaces combined with cartoony colors.

>>597296
>Trying to assemble a small team to make some baby steps towards making a game.

Dude... one thing at a time. From what you posted, you do not have anything like the experience needed to get a production team together. Don't take that the wrong way - you're on the right track, and it's fucking awesome that you have that level of ambition, seriously - but you have not eaten anywhere near enough shit to be a team-leader.

Keep yer head down, keep working, and make sure you eat big bowls of shit everyday.
(PS: If it wasn't clear 'eating shit' is a metaphor for 'pushing oneself to achieve things beyond one's means, only to be met with failure).

>>597296
>I've struggled to focus on one thing at a time recently

Trying to organize a creative team is not what you wanna be spending your time on right now then.

>>learn the basics of Substance Designer
THAT is your 'one thing' that you wanna focus on 'at [this] time'.

Here is your task - you have until February 1st to be good at Substance Desiger/Painter and B2M (a lot of people overlook B2M as a part of the Substance suite - don't; it's handy).
Fuck everything else. You like playing around in Substance? Cool - you're gonna be a texture artist. Focus on that, nothing else. Done.

Next.

>>597296
Think about what you are doing for a moment. Have you ever actually seen bendable bricks on the corners like those? Analyze your references constantly.

>>597324
He's just demoing a material - it's fine. People put brick materials and concrete tile materials on shaderballs all the time.

So there ya go - another lesson learned - show your materials off on a shaderball otherwise autists like this guy will think you're actually presenting a finished model.

>>597298

Thanks. I am eating shit every day at my dayjob as an animator at an ad agency. The cap on creativity can get soul-draining, especially when I'm not even in the right field just yet.

>>597299

Done deal, anon.

>>597324
>>597328

That's right, it's just for preview. The reason why it's an arc is because I'm practicing getting straight UVs on bent objects in Maya. This one was beveled as well, so it was a bit more of a challenge.

>>597292
You can get much comfier on a mobile in bed

>>597328
Thanks for calling me an autist for trying to give some feedback.

>>597358
Get over it snowflake.

>>597358
>>597364
Now kiss!

>>597370

>>597254
Well, this is probably the last time you'll see my bunny... For now. I've got things to do, and if I want to make something in 3D again, it'll be a lot more of fanart or just a dune buggy, based off the one from Smuggler's Run (PS2), complete with spare wheel on the back...

>>597324
>ever actually seen bendable bricks

>>597364
>Thanks for calling me a snowflake for thanking you for calling me an autist

1 fps in viewport ffs

How is this topology for a finalized mesh? I feel like I've messed up with this extra belt around the neck.

>>597324
>What are cut bricks for archways?

>>597428

Yeah, if I modify the material to be single parallel bricks, I'm thinking I don't need to make things more complicated than simply putting straight bricks on a bent object with straight UVs.

Real time render test with the last textures set up.

>>597428
Not the same thing.

is this nose topology good? (i know the shape isn't great; i'm making a base mesh for sculpting)

>>597467
All quads...
Looks good to me, but i am not a topology turbo-autist.

>>597467
>THOSE 3 FUCKING LOOPS STOPPING ON GORILLAGONS

>>597476
care to explain what you're talking about? if you're talking about the fucked up eye topology, i just stopped working on that area to work on other areas

>>597477
>submit a picture
>Someone points out your 3 fucking loops stopping on gorillagons
>Have 3 fucking loops stopping on gorillagons visible on EACH fucking view on the pic
>You have the model itself containing at least 3 fucking loops stopping on gorillagons
>You presumably made the model yourself and placed those 3 fucking loops stopping on gorillagons your-goddamn-self
>Huh, what you're talking about?

I mean Jesus Christ, have you ever considered suicide?

Bonus points for
>I mean I just left that fucked up eye so I can fuck up rest of the model

>>597477
Yeah, my bad. I just looked at the eye area and there's actually more than 3 fucking loops stopping on gorillagons.

Trying to enclose as days pass. I'll add another thick layer to the roof behind the glasses, so you feel inside a big hangar when at ground level.

>>597488
are you retarded

>>597489
your proportions are really, really all over the place here. that doorway at the right size practically looks like it's sized for a dog. the structure doesn't really make any sense for the size it's supposed to be; the ceiling would probably collapse instantly. also those ridiculously fucking huge glass panes on the ceiling? there are many good reasons nobody makes glass panes the size of soccer fields. the light fixtures are also what, the length of multiple cars? i hate to be really harsh but absolutely nothing about this scene makes any sense

Need to work on the decals and the dirt, but overall I am nearly done.

>>597499
The two gray thingies in the bottom right corner don't look good, tweak them a bit.

Despite that, looks good overall

Well I aint one to be taken for them fancy texture type folk but I suppose I can do what I canses.

This is what I've got so far, any ideas on how to improve it?

>>597500
Yeah, you are right, i got lazy.
I also don't like that square thing with the screws/rivets and the 3 whatever right from it.

>>597392
I don't find her particularly appealing, and I'm a man who loves his rabbit waifus.
Maybe it's that she's not rabbity enough, maybe it's that she's not wearing any panties under her tunic, but I just don't find the character that appealing

Also her garb doesn't really tell me anything coherent about her. She's got a sword, but I think she's also wearing sneakers (?), she's wearing a belted tunic but the details on the tunic look a bit modern, I can't tell if she's supposed to be from a medival period or if she's a fish out of water, or if you've even given thought to such details

>>597502
Tip: slapping decals and random detail to shit, is not going to make that shit any better.

Whenever you are doing something, get reference of what others do: open pinterest or artstation, and search for what you are interested in. At first you should literally copy the best concept art piece you find; as time passes you will be able to combine a few of them, taking what you like most from each one. One day you might be able to do some awesome stuff from your imagination.

Trying to do something "yours" when you have not yet developed your artistic side will produce mediocre results like the ones you are getting.

Pic related, what you could try modelling for ez practice.

>>597504

anon's door looks better than the one in this pic.

>Let's have a random part of the frame stick out to make it hard to use it.
>Let's add panel with random triangular pointbecause fuck why not.
>Even though the frame is thick enough to fit at least one person in it, let's add a pipe on the outside. For the light.

>>597504
copying other peoples' sci-fi models isn't going to teach you a whole lot, either. it's far better to find examples of things similar to what you want to create by gathering references for sci-fi set designs AS WELL AS gathering references from real life things you can incorporate into your piece like examples of electrical boxes/access panels and stuff and what they look like after some abuse
an example would be that screwed in grey panel at the bottom right; going by the age of the door, maybe the electrical panel covering got warped from being knocked and it doesn't quite screw in properly and is at a slight angle. you've probably seen this happen all the time in your day to day life like how the electrical boxes on street lights sometimes get kicked in. you aren't really going to be thinking about or noticing these things if you're only looking at other peoples' models

>>597504

This is bad advice.

1. When your taste is not yet developed, how do you pick 'the best possible concept art piece'?
2. When you copy what others do, you will never develop your own sense of design/style.

I have seen it. A junior artist collects reference material for something, all happy and proud, and when I look at what they collect I cringe inwardly. How can you pick what's good when you don't have enough experience to determine what's good? Secondly, when you're using concept art from another artist, you take someone else's idea. Something that already has passed through a creative filter. If you use that as basis for your own work, you will always end up with something derivative.

Instead, look at real world things. Learn a little bit of engineering and how real things are designed and built. Form follows function. Once you understand this design language, you will be able to use it and play with it, but the end result will be based on reality, therefore feel more real, logical, and it will still be recognizable as what it is supposed to be.

>>597511
i think it depends. if your goal is to make something that sort of camp sci-fi, it's helpful to have some references of ideas other people have used in sci-fi set designs (as long as you have other real world references). you don't always have to be on the bleeding edge of originality. the difference is in also studying stuff from the real world so you know WHY camp sci-fi stuff looks the way it does and you're making purposeful decisions in your designs

>>597504
I am not copying other peoples concepts directly, i think that is lame, i want to come up with my own stuff, even if that takes longer.
That is the whole point of the exercise.

>>597504
I never copy whole stuff, i always break it down and make my own version.
This door for example
Mesh:
>45° shapes
>Bevels
>Indents
>Plates
>Bolts
>Vents
>Handle
>Button
etc..
Materials:
>Aluminum
>Painted Metal black, white and orange
>Dirt
>Slide Scratches
>Light/Emissive
etc..
Down to the smallest part, and then make your own door with all these parts.

>>597511
>A junior artist collects reference material for something and when I look at what they collect I cringe inwardly.

way to go anon!

>>597501
needs a working clock for one of the eyes.
it's not edgy enough yet.

>>597529
I decided to try and go for a more basic look first.

I'm unsure how it's called what I'm going to refer to but hopefully someone here will understand.

I want to have a character model follow my VR headset movements like they do in a game such as VRchar. I do not now how to word this so I can't even find anything on google.

Anyone willing to give me a hint?

>>597537
i don't now if this will work or not, because I have mever tried game dev tools.
but logically, you should parent character to vr camera, or it's target

some houdini constraints practice

>>597493

I will add pillars and beams when I'll be done with the walls and ceiling blocking.
The door on the right is the height of a human in the game, on the other side there's greffed pre-built offices. The left "door" is bigger since it's an entry for trucks.

The biggest fixture is a bit wider than 1 car in the game.

The top windows are huge you are right. But I'd like to make it work especially for light coming through it. I don't plan on making it realistic, just somewhat believable, following the initial game vibe (I'm making a workshop map).

>>597537
In whatever program you are using, you want to find the vector input of your VR motion tracking device and use that to control the head bone directly or control the playback of a head animation.

>>597543
>>597589
Thanks for the advice, I'll give it a try

>>597581
add little bit of friction to the surface of that obcect which explodes, because whendust falls on it, the dust slides hen it moves.

any tips on making stylized textures? i usually create a grayscale out of maps and then play around with gradients, but i think some parts of this are a little flat

>>597423
looks real nice anon

>>597604
I'm really liking this so far

>>597436
looks pretty damn nice

>>597511
You can't "develop" good taste, aesthetic is entirely arbitrary. Example: Everything in japan, whether it's color choices or materials or proportions or what they think is "cute", looks like absolute dogshit to me, but they fucking love it.

Most jobs, whether in film or vidya, involving modeling are going to have you make something based off someone else's concept art. Some fag in charge of hiring at pixar said at a Zbrush conference that the #1 thing you can do to improve your portfolio is to just take someone else's concept art and make it 3D.

And by the way, all of you are missing the bigger picture: anon's door looks like shit because the lighting is bad, the materials are bad, the color choices are boring, and the composition is shit.

And finally, if you're in a position of authority and/or seniority and you see a fellow artfag doing something you don't like, as a leader you have a responsibility to educate them. You passive aggressive unlikable faggot.

>>597581
I don't know what real world material this is simulating, if any, but it would be something really bizarre like a box made of sheet rock with small rocks inside.
The individual things happening looking good (the dust, the fracturing, the bending, the scattering) it's just the combination together is strange

>>597489
>>597493
that other guy is right, your sense of scale is all out of whack
>>597501
lose the texture on the legs, or at the very least make it much more subtle, doesn't look right at all. rest looks fine to me

>>597613
Which door are you talking about?

>>597615
I don't understand really.

>>597533
give the wings a smidge of an actual wing profile. Not too perfect, but more to break up the symmetrical look it currently has. It's very obvious that it's 3d modelled.

>>597203

Like what exactly?

Texturing is hard

>>597613

1. Fair point. I should have used another word, or have been more specific with what I mean by that. So, can we replace 'taste' with 'design sense'? I.e. the ability to look at something, decode it (analyze the composition, shape, use of colour, texture, etc. etc.) and be able to make a reasonable judgement based on that.

On a side note, though taste is personal, beyond reasoning and impossible to come to a universal agreement upon, I do think that 'good taste' can be developed. Since 'good taste' is a set of choices and preferences set by a specific social and cultural group, the criteria/rules governing what is seen as 'good taste' can be decoded and learned and applied.


2. 'Most jobs' is a generalization. On what data do you base that statement? There are many game studios where you can both design and make your own models, for example in the mobile games sector.


3. You're missing the picture here. You're basically saying a military hummer looks like shit because it has no aerodynamic curves, no glossy glitter paint, no chrome rims.

Anon's picture is not bad because:
- anon was obviously going for an 'Aliens' military style door, ergo drab colours.
- all the various accessible elements look logically placed, look functional, with sensible readable markings.
- it actually looks like a door.

Is it a little boring and derivative? Sure, we've all seen something similar before. But if this would end up in some kind of Colonial Marines game it would be perfectly servicable.

4. You assume a lot here. How do you know what I did or didn't do? I wrote that I cringed inwardly at the beginning, but then I sat down with this person for many hours over several days, talking about what does and doesn't work, where and how to find the best reference material, how to build a solid reference library, and how to decide what to pick.

Learn to think before you call people names.

>>597613
"Most jobs, whether in film or vidya, involving modeling are going to have you make something based off someone else's concept art. Some fag in charge of hiring at pixar said at a Zbrush conference that the #1 thing you can do to improve your portfolio is to just take someone else's concept art and make it 3D"

Interesting idea. I've made some props via 3d modeling and printing that ended up pretty goddamn lucrative via this. Pic unrelated.

>>597629
It took a dude at a z brush summet to make people realize that? Well I suppose artists aren’t in the business for money anyway

Made another substance. Yeah, I obviously followed one of the most popular tutorials on Youtube, but it was a good intro to creating patterns, and I deviated a bit from it by coloring separate surfaces differently.

not the most original. learning rigging.

>>597628
>You're basically saying a military hummer looks like shit because it has no aerodynamic curves, no glossy glitter paint, no chrome rims.

That's not what I meant when I said "bad lighting, bad materials, boring colors, shit composition". You can have a technically excellent model but the final render will look like shit if you have poor lighting, simple materials (in the shader sense), boring colors (yes the source material has a set color, like olive green, so make your background shit a contrasting color or colors that makes the hummer stand out), and poor composition (I can't encapsulate a concise definition of composition inside parenthesis)

I should have said "big studios" rather than "most jobs", the smaller the studio the more generalist your skillset is apt to be, while a large studio you're more likely to be a specialist.

>>597625
>I don't understand really.
I don't think they are technically correct when they say your "proportions are wrong"; as long as you're modeling everything to sizes that exist in the real world.
I think the problem is one of perspective, as in the camera settings. Camera for the screenshot you posted with the figure on the ground by the door is setup / positioned / etc such that the figure doesn't look far away, he looks tiny.
As long as it looks right in-game or whatever. Level editor type shots always look a bit odd.

wrapped a bunch more engraving on this gun.

>>597278
I might make up for it with the characters

I still need a plot and shit to do as well. I always start making a game without any ideas :s

>>597743
Check out gold damascene, especially the way the old Arabs used to do it.
https://www.youtube.com/watch?v=4KM7ySNWuqU
There's a right way and a wrong way to go about doing it.

WIP Rei Ayanami.

I'm making her as a custom avatar for VR chat. I don't have much exp modeling with deformation in mind. Hopefully it doesn't turn into a disaster.

>>597747
Very interesting. I love this guy's channel, some pretty interesting gun lore in there.

I'm actually replicating an in game model for SLA 3D printing/molding/etc, though, so the final finish and desired look and feel is known.

>>597760
This is enough for tonight

>>597768
Nice hair.

Also, post wireframes so I can make autistic comments about your face topo.

>>597791
I've never made a 3d anime girl before so I really don't know if she will turn out freaky looking or not. Hard to tell without textures and lighting.

Am I doing it right?

Started cleaning up/remodeling/smoothing a destiny 2 game model for 3d printing.

First pass is just splitting out by manifold/section, setting up symmetry modifiers, quadding

first animation down
holy shit youtube compression wasnt kind
https://www.youtube.com/watch?v=1j_LF6F1f-w

>>597873
Reference more people walking. Those arms are totally wrong.

I think I'm finally getting better at boobs. Why are they so hard to make not awkward looking?

>>597880
because you touch yourself

What are some good zbrush tutorials I can find on cgpeers? Anything not lynda.

Do you guys even bevel when shooting for a low polycount char? Right now I have to shoot for under 10k polys for my character and beveling feels like such a waste.

>>597743
>Engravings give you no tactical advantage whatsoever

>>597886
See FWN technique >>597215
When two adjacent faces are in one smoothing group, the vertices aren't split, so a plain cube with hard edges and one that's beveled but with one SG both have the exact same performance footprint when rendered, this effectively lets you bevel any edge you want for "free". Even if you'll be baking normals, the resulting mesh will still be smoother overall and get a better bake. Editing the normals lets you maintain hard faces even when everything is smoothed together.

Are there some good FWN scripts available for Blender?

>>597904
YAVNE is mentioned often, so I'm guessing it's the defacto normal editor for Blender.
https://github.com/fedackb/yavne

>>597901
The explanation in the linked post goes over my head a bit, which is sad considering I've been doing this for years. You're saying to run that script instead of beveling hard edges?

>>597912
Basically you take your typical hard surface model, and place chamfers/bevels everywhere you would normally have hard edges, then you smooth the entire mesh with one smoothing group.

Once you have that done, you pick out all the faces you want to be hard again, and run the script on them. Some tool/scripts let you choose how hard you want the resulting surface to be, but generally you want every face except the chamfers to have influence over the shading. If you have ngons, then you definitely want full influence from the face direction.

Just to reiterate, any time you have a discreet smoothing group on a model (even if it's normal mapped), every vertex gets split by however many edges connect to it, so for instance a basic cube really has 24 verts and not 8 as the software tells you. A beveled cube also has 24 verts, but none are split if they're all smoothed together, so you get the bevels for free.

>>597880
Get some bearing on anatomy before tackling areas of the body. They flow down from the shoulder and sag outwards from the ribcage, and accumulate more at the side of the body than the center. You also need to take gravity into account. If you fill a bag half full with fat, it's going to settle in the bottom, and that's what breasts really are, half filled bags of fat.

>>597873

For swinging arms, use FK, not IK.

Also, head shouldn't swing from side to side. That just doesn't look right. Assuming you have rotation spaces in your rig, change it so that the head is not dependent on the upper body's rotation.

Finished modelling the turret, I've been following the star citizenn workflow, so it is midpoly with face weighted nromals. Now I'm going to start modelling the decals.

>>597967
Forgot pic

>>597967
>>597968
And this is the reference

>>597967
>>597968
>>597970

looking good, seems like animating it would be fun too

>>597976
Yup. I also plan on doing particle effects for the shots and all that stuff

>>597926
This is a skin tight body suit though, not a bare chest. And I know a body suit would normally squash them, but in this instance the designer is a perv with a boob window.

Make more sense?

How absolutely critical is it to avoid 3 / 5 sided faces when making a deforming (organic) mesh? Can I get away with it if there is no easy way to avoid it? I'm not even sure how I would make everything quads in some situations.

Rei progress. This thing is gonna be a bitch to downsize as I need about 1/3rd of the current polys and she's nowhere near done.

>>597981
Tris aren’t too bad, however you should never have 5-sided gons, they’re reasonable on hard surfaces, but for deformation it’s shit. Make a pole instead.

>>597503
Yeah, Vivi is an odd depiction of a furry rabbit, despite having a couple distinctive features, while looking more human. Her sword could cut open dimensional portals, and her belt keeps her garb sealed so it prevents wind from blowing in.

Yes, an odd fashion sense, but she doesn't always wear that, because that's when she goes out on some adventure.

>>597980
>>597988

Still don't buy the breasts, they look like protrusions on an otherwise male chest. The armpits also sit much higher. The hip flare is low, lift it up for a much more sexual feeling.

Look at reference of real features and not just anime girls, because anime girls are just exaggerations of those real features to pleasure the dick.

>>598002
>Look at reference of real features and not just anime girls, because anime girls are just exaggerations of those real features to pleasure the dick.
But this is an anime girl. I'm not going for realistic. I can fix that part up though. Good call on the hips, I'll try to move them up more.

>>598014
But anime girls are based on real girls, and knowing how real girls look will make you better able to stylize those features into anime. There's a lot small details to anatomy that make all the difference, some of them are exaggerated in the anime style, some are watered down, but understanding the real function and form of some of these features will help the end result in both aesthetics and topology.

Like for instance how the breast flows from the shoulder. Making it flow down as such means changing the topology flow as well, which will result in more realistic deformations and better shading.

>>598015
How's this looking? Just a quick 2 min change. Ignore the slight bulge.

>>598017
>>598015
Meant to post this one first. I think the ass needs adjusting in the other one.

>>598017
>>598018

not the boob autist, but i like to think that the top of the breast flows more into the shoulder muscles like in >>598015

>>598023
Again, it's boob armor, her actual boobs are well below that (flows into the shoulder / armpit area).

I would post more to try to show what I mean but at this point I'm spamming the thread with my renders so I don't want to be annoying.

>>598026
But imagine how uncomfortable it would be to wear that, clamping down the side of the breast in a very unatural way. And it also doesn't look quite as breasts without that detail.

>>598026
Ah ok, maybe sell it more with the texture then

>>598018
Unless you planning to get women wet or give men erections, get rid of the boob socks.

I continued to work on the textures for this pool scene. Any suggest so far?
I am still far new to unreal and feel kinda lost on what I should do to improve the look of this.

>>598070
more polygons for the pool

>>598044
>>598042
You can kinda see here what im going for here. Theres like a latex type fabric underneath the boob plating. Sure its incomfortable and impractical but im trying to stay faithful to the original design (with some modification based on personal preference).

Forest floor substance.

>>598070
Your buildings need awnings. Everything is too much one color. Think of how nature would affect these buildings. Would dirt be building up or staining the bottom trim? would there be weathering on the edges of the buildings? The roof?

On another note, you should add more color. There's only green and white for the palette. Get some flowers in there.

>>598097
Your colors are too saturated. PBR workflow requires you to use extreme values sparingly, *and* greens in nature just don't have that level of saturation unless you're going for a stylized texture (which you're not).

Big rocks are off to a decent start, but have way too much tiny noisy variation (cut down on the really high intensity noises and slope blurs). Would be nice to have some large, hgiher frequency details going through them. Also, I'd personally make em more like polished pebbles (perhaps with some weathering), these all look like they just broke off some volcano.

Tiny white pebbles are too bright and noisy. Keep it subtle, it looks like someone spilled popcorn over the forest floor.

Once you clean off all the noisy shit, you could try to add some more layering and variation to your grass. Thicker blades, as well as blades with sweeping curves (this ain't a haystack of needles).

I'm not sure what the yellow and lime green things are.

Depth of field + lens flare = why?

>>597968
Personally, I think I would've given those handles insets + panel cuts rather than joining them to the main body - it currently ends up forming a big awkward piece that looks like it would be a pain to manufacture.

Also, your edge width is inconsistent, some of the stuff having much thinner bevels, the two cameras for example.

Good job otherwise, seems like a fun exercise for midpoly+FWNs.

>>598111
Thanks for the critique

I plan on doing a "weld" normalmap for the "decal/trim" texture. That area you marked will have that so it won't be as a single piece, it will look like two different pieces welded together.

I will also try to match the bevel size, already fixed a few but forgot about small detail like the cameras.

Is it happy or sad? no one knows!

>>598110

Did a small update based on your feedback.

> unless you're going for a stylized texture (which you're not).

I am, though. I just don't subscribe to the idea that stylized has to mean simplistic, so I see how far I can push it.

> I'm not sure what the yellow and lime green things are.

Those are leaves. I thought it would look more interesting if I put in more color variations, but it doesn't really help when you can't really make out what they're supposed to be.

> Depth of field + lens flare = why?

Depth? Dunno, same reason anyone else would put it in. I couldn't import PBR into Maya 2018, so I had to render in Substance Painter, which doesn't really have that much to offer for rendering.

>>598118

Those leaves look like beans. You should create a texture with a few photoscanned leaves, import them inside substance to combine them with your grass and rocks.

>>598118
Everything just looks pasted on. Nothing overlays anything. First you have your jelly grass texture, then pasted on top of that are your slimy lima bean leaves, then your copy pasted rocks. It just looks bad, not stylized.
It looks like a skin that I could use on a spore creature.

>>598114
Holy shit that's cute as fuck boi

>>598114
Do you have a blog or something? I want to see all these things you've been doing

I'm making this low poly character in Blender and I'm almost done modeling except for the hair.

htts://orig00.deviantart.net/71c3/f/2015/029/6/0/eqg_adagio_dazzle_flash_puppet_by_yoshipapaercrafter-d8fxihm.png

How do I model the hair? Making a bunch of spheres skyrockets the poly count and will be hard to paint the yellow streaks on. I tried intersecting them in another program because Boolean refuses to work, but the end result required too much fixing.

>>598121
>>598123

Last update to this material before moving on to a new one.

I increased the roughness overall (Painter render looks glossier than Designer viewport), reduced the height difference in the leaves, amplified the fiber texture I made for them, and chipped off some of them so they would look less like beans, and added some dirt on top so that I would have something that overlaps. Turns out that the dirt made the material less saturated as well. Of course the things don't overlap, they're meant to be separate objects.

Though this is just practice, and I'll look further into the rocks and photoscanned leaves for another time. For a game, I would probably leave out the large rocks, just model them and place them on top of the material, and have leaves on separate planes that move around with the wind, and put other larger plants on top as well.

>>598167
do the hair in 1 piece. it is possible

>>598172

I want to but I don't know how. I tried combining vertices but I wasn't sure about how it was turning out.

>>598174
keep welding in the front, make sure you delete anything hidden. use proportional editing to quickly move large amount of verts

>>598171
No, I have to be honest - delete it, go through Substance Designer tutorials (there's plenty of grass ground breakdowns on YouTube) and after that start again. This barely resembles anything. IMO, ofc.

>>598171
absolutelydisgusting.jpg

>>598181
>>598178

Not good enough, can't do anything with this feedback like this.

>>598118
Don't try to reinvent the wheel when you can't even make a cylinder yet. Do you have a clear vision of what you're hoping to achieve by "seeing how far you can push it"? Or do you hope it'll just come together, somehow, if you tweak it enough?

The cornerstone of stylized art is the exaggeration of main, striking features, while toning down minor details and noise. You're trying to introduce dense, intense noise all throughout your substance (just look at the texture on those rocks), which makes it far from stylized.

There was a pretty good article about stylized art on 80.lvl recently, you should give it a look: https://80.lv/articles/realistic-vs-stylized-technique-overview/

>>598183

Thank you, this is criticism I can work with. I think the problem for me so far, is that I start with tutorials meant for realistic materials, then I end up doing a more stylized version of that in the middle of it. So, I think I'll improve once I begin making substances without starting off at a tutorial.

>>598182

The rocks look out of place, the leaves look like beans and it's way too saturated.

>>598182
Sorry, it's just beyond saving. As I've said, you need to learn more before doing your own stuff. What are your references even? I don't know how to critique this, it doesn't look like a grass and rocks at all to me. Go through breakdowns and compare that work with your own.

I give up, just call me an asshole if you want, but I don't know how to reply.

>>598171
can we see it applied to an uneven plane? Or is it intended to be applied to cubes?

>>598156
>>598162

thanks :))

tumblr and instagram
https://www.instagram.com/ribberart/?hl=en
https://www.tumblr.com/blog/ribberart

I have 13 followers on tumblr and 261 on instagram for some reason

>>598209

No, it's supposed to be on an uneven plane. I haven't modeled what I'm supposed to use my substances on yet, but I have a certain idea of what kinds of material I need.

>>598176

It looks like crap and has a bunch of tris. Isn't there some way to box model it with modifiers?

>>598279
i think you should start with a box, subdiv it once and just go from there
slowly extrude it until you get a good basic shape

>>598282
oh also some useful modifiers you can use are skin,mirror and lattice.
especially skin, i find it annoying to use but it should be useful for blocking hair

>>597180
Re-doing the model. Going for a cuter pint size ninja. After trying to get corrective blend shapes to work properly it be faster to just redo her low poly. I figured less geometry less problems.

I learned the hard way that getting corrective blend shapes to fix arm deformation automatically is a pain. I have not quite figured out how to fix it. Also I can not figure out for the love of god why the left arm rig is off and why the wrists keep rotating.

This is just a block out. Will have a proper model later. working every day till Jan1.

>>598308
your dedication to making your own rat waifu is admirable, keep it up.

>>598311
Thank you. Learning how to do new things and do them faster is making it rewarding.

>>597180

>>598315
haha he has a little willy

Looks nice though, some kind of Tzeenchian birdman?

i need a little bit of help
im trying to make the selection into a type of cuff, an eliptical tube of sorts to serve as a trim.
now i try extruding but when attempting to scale it, thats where things go wrong as it will throw off the alignment.
i need to scale it parallel to the faces i extruded them from.

This isn't so much of a WIP post or request for feedback....
Working on an alien planet Unreal environment with all sorts of creepy gross looking rocks and plants. Part of the visual language is that things are full of holes and little networks of tunnels.
This is a small crop of what is gonna be a large sort 'hero' rock piece. I'm just kinda getting worried about how it's gonna bake and how the hell I'm gonna retop it.
I guess the reality is I'm looking at a fairly expensive lopoly mesh, which I don't mind because as I said, it's gonna be large, be at eye-level, and viewable right close up, but I don't know what other kinds of problems it's gonna cause down the line.
Thoughts, anybody?

>>598353
Another angle....

>>598353
>>598354

Last one...

>>598353
>>598354
>>598355

And just to you time if you were wondering - yes, it is completely taken from concept artwork for Mass Effect Andromeda, which I own but have not played - so I actually have no idea if this kinda stuff actually made it into the game.
Regardless of whether or not the game is any good (I've heard it sucks) the Dark Horse art book for it is really, really impressive.

>>598321
)) is there any other way to sell bird-mage-monster without Tzeentch? ))

so yeah, it'll be Tzeentch cultist-bird-man with chicken legs and second fish head, probbably instead of it's dick.

>>598353
>>598354
>>598355

If you want it to have holes all the way through you should stop worrying about low poly right away and just get it around mid poly, it's not like UE4 cares about polycount anyway. The mannequin they supply in the default projects is at 40k tris or something like that, and is all smooth surfaces with little detail that couldn't be baked. If they don't care, it probably doesn't matter much.

>>598361
>If you want it to have holes all the way through you should stop worrying about low poly right away and just get it around mid poly, it's not like UE4 cares about polycount anyway.

That's kinda what I figured - just needed reassurance. Structural detail like this (and yeah I want holes that can be seen right through, I don't want to cheap out on this particular prop) has to be done with polygons - there's no other way round it.

>>598308
dude... your model was very good, to begin with...
but, if you are doying this from scratch, you'll have two nezuwifus now ))

>>598279
Why don't you try with metaballs and then decimate?

Wip.
https://imgur.com/a/KM5q1
And two turnarounds.
https://streamable.com/aluew
https://streamable.com/niyya

>>598390
cool work, bro.
and always ignore haters.

>>598390
uuuuuu!! sexy!!
that's my jazz, really..


I was going to go BDSM/Fetishy look for this character,but then got carried away
https://www.instagram.com/p/Ba6tL1aB7Ot/?taken-by=gam3_b0y


>>598392
agreed 100%, fuck the haters!

>>598393
>https://www.instagram.com/p/Ba6tL1aB7Ot/?taken-by=gam3_b0y

I need l3wds of this

something's wrong with my geometry here but i can't figure out what it is. what's going on with these normals on the inside corner of the inner shelves? i'm not able to insert edge loops properly. been trying merging all the points together to see if that was the problem but it doesn't seem to be

>>598395
Looks like those edges are unconnected, or you have doubles there. Are you using array modifier? If you do, you need to check the "merge" checkbox in it.

>>598396
oops i found weird faces inside the negative space of the shelves. basically there would be a face spanning the entire width of the bookshelf inside the negative space. wasn't easy to find. thanks though

>>598394
https://www.instagram.com/p/BVcHYpFBH7X/?taken-by=gam3_b0y

here is without pants, in very early stage of design, but body is done.

and here's topless )) (image)

>>598398

I want to cum on these tits.

Been ages since I did 3D stuff, getting back into the swing of things.

>>598421
That doesnt look anything like a sexy 3d waifu

>>598398
Looks nice. Keep it up.

>>598429
I've finished that model for some time already tho.

PS: second day of bird )
I scul[ted the body, ditched placeholder basemesh

>>598434
I'm really enjoying Creature sculpting tho!
trashing anatomy and also trying to be anatomically correct is us awesome !

I think I found my thing...
(btw that mouth on stomachsucks either I'll make it really large or delete.

>>598435
forgot image...

>>598225
can we see it just dropped onto an uneven plane then? Not what its specifically for but just any wiggly plane.

The depth of the material looks strange, but depth always looks strange on cubes.

blocking out a hog mechanic, i really like trying to convert a 2D concept (not mine) into 3D

some minor polymer material tweaks are left and i'll call it finished.

>non-degenerate character sculpts and photorealistic models
Please keep going and stay that way, /3/.

>>598451

Fuuuuck that looks good.

>>598451
My penis goes hard for hardsurface. Looking good.

>>598451
>jew gun

>>598451
I'm not a fan of the noise you're using for your plastic/polymer, and would strongly, *strongly* recommend changing whatever grunge is creating height in your rubber(?) stock pad (the shoulder one), cause it looks like you put a random SP grunge into your height layer and called it a material. Do you have references for that area? I can't think of any credible material that gets worn that way.

It's solid work overall, but if I had to give more general critique, I'd say that a lot of your materials feel too rough (roughness-wise) and noisy/cloudy. You've got a subtle digital camo as well as plastic bump on almost everything, but little variation in unique wear for most surfaces. What would really help sell the believability of the texture is some unique dirt, oil leaks, scratches, basically whatever kind of unique and interesting damage you can source from real guns made out of the same materials.

What is some modern weapon tutorial I should go through? Chamferzone's AK tutorial is 17 fucking hours long which is crazy, and I heard he is not using the newest techniques (something about bringing the model in Zbrush instead of using a floating geometry for baking, or whatever that means). Isn't there something more similar to Tor Frick's method of fast modeling, but for 3ds max or blender?

>>598497
>the newest techniques
You're going to be surprised, because a vast majority of weapon artists still use sub-d. Not all of them, mind you, and it's slowly being replaced by various other workflows, but it's still the only way to get certain complex shapes in hard-surface (especially the flowing, organic ones).

There are several hardsurface methods currently being used, they each have their pros and cons and none of them are really more "up-to-date" or taking over another (except sub-d which is losing ground). Some of them are:

Sub-d modeling (with control loops)

Crease + opensubdiv

Quad chamfer (Chamferzone)

Rounded edge (Tor Frick)

Zbrush booleans + dynamesh (The Division, various)

I can offer a breakdown of each if you want.

>>598451
Is this an original model or did you model it from some concept art? Honestly, I get depressed when I see how much time it takes to model any kind of a weapon. And that doesn't even take all the design process into account.

>>598501
What method of those mentioned would be the fastest (even slightly)? You don't even need to explain it, just some good video on YT would be enough (and probably better than just a written explanation for me).

>>597968
>>597970

"Done", but too bad I can't actually finish it.

The second UV workflow works wonders in keeping detail sharp, no matter how much you zoom is. It is still more limited than the traditional substance painter workflow, but hell, it looks cool as fuark.

This is after setting up some quick ugly materials just to show it off, and iterally just two planes as light sources. So no surprise it looks like total shit.

It will look 1000x times better once I can do proper substance designer materials and set up fancy effects like parallax in unreal. Probably my next project will be a sci fi environment so I can get this turret in an actual setting with actual lightning. But that won't come in at least a few months more...

>>598475
>>598490
>>598491

thank you. Have one more shot

>>598492
valid points are made, rubber butt is just a drag n drop painter tubber, as it wont be seen in-game.
Roughness and digital pattern is a direction from concept artist. Also it isnt supposed to be worn out.
>>598510
after a concept art with vew revisions, i cant post it here as its not my work, however it follows concept pretty well - at first it looked like an assault rifle, i've managed to discuss change of the magazine/receiver to make it bit smaller. Its for a fan project im working on when i have spare time, there is no public announcement yet, but its starting to look pretty good so i decided to put my time in it.

>>598514
But just look at how sharp detail is. Keep in mind there are only two 512px textures used.(one for normals, other for Ao opacity emissive and masking)

I am finished now with the first iteration of the texturing process.

As you said, I believe the scene needs more character as it looks to clean, therefore I will at some decals and stuff. I might also add some more stuff to get color variation into this, maybe some red flowers.

If you have further suggestions just let me know. Thanks!

>>598519
You still need more vertices along the pool. It's too jaggy and jarring to look at in comparison to the rest of the render.

I will do this as well, thanks

>>598515
>rubber butt is just a drag n drop painter tubber, as it wont be seen in-game
With the amount of work you seem to have put in the rest of the gun, I find it strange you'd leave that part in that state, if only for the sake of portfolio shots. Hell, having literally no height information there would look better than what you have going now.
>Roughness and digital pattern is a direction from concept artist
And it is your job as a 3D/texture artist to interpret them in a believable and satisfying way. There's plenty of interesting uses of digital camo finishes on firearms you could reference, and I doubt your concept artist told you "use X roughness value and put this specific noise on it", so that excuse is moot.
>Also it isnt supposed to be worn out.
Maybe a misunderstanding as I'm not referencing extreme, fallout-style damage. But even brand new objects have slight, subtle imperfections, which we often need to push and exaggerate in videogame graphics to compensate for realtime lighting being imperfect.

If the gun is indeed intended to be new, I'd actually tone down whatever kind of damage is on the grip - it looks like your guy tried to eat it.

>>598529
well you got me with the grip, im just lazy.
While most of your points make sense, and i've already considered them, I dont think i'll be changing the way that pattern is applied - i specifically didnt wanted it to look like duracoat or paint, but more like embedded into the structure of the gun, as hell i know, it micht be 3d-printed gun or something like that. I've played a bit with roughness details and bring it to the level of microsurface details it have now, and as for now im rather satisfied with it. Its possible that at the end, when all the weapons are developed, they'll ger a last pass of worn layer, but afaik it'll be masked in engine for short periods of time. Thanks for input tho!

>>598510
The fastest of these methods is probably rounded edge - the downside being it somewhat limits you in the kind of shapes you can create. The rounded edge shader is great at quickly adding bevels to straight, regular shapes, but begins to struggle (produce some artifacts) the more complex curves are involved. You can actually mix it with sub-d modeling (using creases, for example) to get the smooth shapes you wouldn't be able to get with a mid-poly alone. Another downside is the shader doesn't produce 100% immaculate results when baked, even under ideal conditions. Some people mind it, some don't - I personally don't find it's a problem at all in game art seeing as your textures will cover most of the (small) artifacts and mipping/compression is still a bigger (unavoidable) issue for prop quality.

Another issue you might run into is that the workflow only has documented support in Modo as of now. I believe it's possible to bake it in Max and Maya, and that it's currently being integrated into Blender, but I have no clue if the shader is in a "useable" state in any of those programs.

Here's a couple videos about rounded edge:
https://www.youtube.com/watch?v=EHkbCQQykAg
(An overview by Warren Marshall)
https://www.youtube.com/watch?v=N5UPaoQm7kI&t=695s
(Combining the shader with sub-d modeling)

https://www.youtube.com/user/Askguden/videos
(Tor Frick's channel, which provides a great overview of a rounded edge-based modeling workflow in use.)

so when trying to extrude along a curve all i got was this bullshit creating 2 flat strips side by side where i want a damn tube like negative space.
i followed this tutorial practically to a T but i cant figure out whats going wrong
https://www.youtube.com/watch?v=79-nBN2QGjw

>>598530
looks fine. share concept art?

I've been creating these engravings by booling away from post-nurbs surfaces, but maybe there's a better workflow, this is seriously time consuming. Any suggestions? Needs to be actual geometry for a 3d print.

>>598545
cant share concept sadly.

Engravungs are from heightmap. If you need to make them on geo, experiment with topology tool-diamond shape, and beveling it. Should work.

>>597427
Looks good to me.

>>597700
Very nice, Anon. And you don't need to be defensive about following a popular tutorial, it wouldn't be popular if people didn't use it.

>>598070
Add normal maps for stone to the pool edges and the other bricks. Look up Michael's house in GTA V as a reference for what the house walls should look like. If you decide to pirate the PC version of GTAV to get a better look, get the "nosteam" version off piratebay.

>>598440
Looks good so far, but the nose doesn't look the same as the drawn one.

working on a car

>>598171
I don't even want to imagine what your graphs look like. There's no shame in being a beginner but Designer is all about being familiar with the toolset. Your not gonna get better on here without having a full grasp of the tools which you clearly don't have.

tried modeling a bottle of milk but... i dont know if its the bottle or the milk itself, it just isnt white enough. maybe i have bad lighting.

>>597760
Shoulders too wide I think

>>598590
Bad lighting (looks like you're using the default World setting), you probably should use the Principled Shader with a realistic IOR, also geometry (particularly if it has poor edge flow and quad spacing or edge distribution, or if it's lowpoly with sudiv turned off like in a viewport preview) will significantly influence how your glass shader (custom or default) behaves. Also your camera settings are extremely important, you're viewing it with a 35mm lens extremely close up, the distortion isn't going to fit your mental conception of milk bottle. Consider a setting between 50mm or even 120mm if your camera is going to be so close.

You should also consider blocking out a scene, even a very simple one from one of your ref images of like a countertop + drinking glass + milk bottle. If you don't have other things for the glass bottle to interact with, you're going to get a very unnatural look to your glass. You can counter this a bit with an HDRI but you should still combine that with some simple scene.

>>598308
This is still the apex of mouse wife (not my work, from a guy on vimeo).

got the engravings done for mah space gun

>>598421
Progress

>>598531
What do you think about Fusion 360? I don't know who of you guys mentioned it, but now I'm going through some tutorial and it looks awesome. Relatively simple to use but I probably shouldn't have gotten into an advanced tutorial right away, because now I'm struggling a bit.

>>598598
Tried it but it's not for me. It's certainly not bad but for me it's too CAD-like also it's online/cloud only. That's what I fear will happen to MAX/MAYA in the forseeable future.

>>598593
True, but its not my mouse wife.

Progress,,

I'd be happy to hear any suggestions from anatomy fags.


PS should I remake the head?

>>598676
I really like the head! But that lump in the center of his neck may be too much of a protrusion.
His belly button is too low. Raise it just a little higher. He's also missing the hip area that rest just below the belly fat and just above the mons/groin.

TIP: Belly buttons lie exactly on the horizon line at the top of the hips.

>>598679
>>598679
agree on a belly button, but that part will be covered with big belt thing (armor)
it'll be mostly covered in armors,but I want to leave exposed

>>598598
Neat, clean, and (in-theory) nondestructive. Everyone I've seen use it so far (industry professionals in hard-surface) has been fairly slow with it, so compared to other workflows I don't think it's always the fastest (though it is definitely viable if you wanna check it out).

Of course it has its pros and cons, I've seen concept designers bash together some hardsurface stuff super quick in moi 3d, it might just be better suited for certain shapes and aesthetics than others (cause you can avoid whatever operations take the longest time to execute in CAD programs).

https://streamable.com/jvs1y
https://streamable.com/bjsga

>>598701
And.

Both are in progress obviously.

Please explain the difference between substance designer and zbrush + photoshop for creating textures. I see that some artists used the latter in horizon zero dawn. Also, PLEASE link me a place where I can learn how to sculpt plants from scratch in zbrush, polypaint them and bake to texture, since I can't find any resource.

>>598750
>Please explain the difference between substance designer and zbrush + photoshop for creating textures.
Substance Designer is a fully parametric/procedural system, meaning that absolutely everything can be altered at any time, from the creation of the initial shape on down. And since you can load substances directly into your 3D package, depending on how you set the substance up, you can expose parameters that let you change its properties right from your material editor, without having to output any materials.
To give an example of how this is useful, you can swap around shape nodes at the start of your graph to change the appearance of your texture, but keep everything else like color and details in place. So you can say generate a bunch of various tiles but have them look generally the same.

ZBrush + PS sounds more like the workflow that's utilized when making scanned surfaces, because with Substance around there's no real point to doing so unless you're working off the best data set possible (an actual sample of a surface). Essentially you take a whole bunch of photos of a surface from every angle, reconstruct it with PhotoScan or RealityCapture, import the resulting mesh into ZBrush, and use RGBZ grabber to make textures off of it, including normals with the appropriate shader. This is a pretty heavy workload, but the results are extremely impressive.

>look at artstation profiles of pro artists
>Look at their UE4 renders
>Beautiful interiors
>Natural exteriors so shitty looking that I would think they were done by some indian kid in a spare time if I haven't already seen their interiors
What did they mean by this?

>>598764
it's much easier to create a small closed environment with nothing organic in it than a large open one that possibly has organic stuff in it. duh

>>598770
That doesn't make much sense since both types of scenes have equal amount of models and textures. But I'll agree it's harder to make organic stuff.

>>598776
eithe you are retarded, or you are renderig rooms and think you are good.
anyone will be able to model and then render modern interior.
it's mostly basic geomrtric shapes, everything is symmetrical. every texture is a tile.
there is no real texturing challenges, especially now in interior rendering. if toy'll try you'll be able to do it in couple of months, I think it's the easiest thing in 3d.. (if not say what is easier)

but modeling and texturing foliage is really hard, you have all organic shapes, there are o straight lines. nothing is symmetrical, textures must be accurate in color and resemble objects they are shading, some "google image wood textre for furniture" won't work on real trees. nothing is polished, you won't be able to shade 10 different objects with one procedural matrial slughtly tweaked. you must UV everything.. etc.. etc...

>>598776
>>598764

nice WIP

>>598785
I actually don't know how easy it is to render interiors because I have more experience with exteriors. But I get what you're saying, I know exteriors are harder, but was still surprised by the differences in quality from pro artists.

wip

>>598797
looking g00d, buddeh

>>598797
Are using references? Could you show them if so, looks familiar

>>598398
>https://www.instagram.com/p/BVcHYpFBH7X/?taken-by=gam3_b0y
where is the porn nigga

I added a small hungry dog.

animation will be done in 4 days after I realize that I forgot to hide one of the layers during rendering :(

Whatever happened to that guy making the ninja mouse?

Did he add a vagina to it after all? Would love to see her getting fucked.

I have no experience modelling at all, but I wanted to make my waifu to use in animation. followed some dude's Blender tutorial

The hair is messed up though, there are seams where the underside meets the top part, and don't know how to connect it properly to the scalp

A simple shitty Snek. Probably a little off and the scales are just a rough displacement for some quick easy details but only need him for a small detail in a project.

>>598828
how you do the nice scale texture?

>>598865
Not that anon, but i did it before in an old job.
Displacement texture map does the trick in Mental Ray, but it takes forever to render.

Well its done. The colors unfortunately look like dog shit on instagram but what am I gonna do about that :(

Well its done. The colors unfortunately look like dog shit on instagram but what am I gonna do about that :(

https://www.instagram.com/p/BdYZ8CNDcT3/?hl=en

>>598874
I tried to do some scales with normal mapping before, but seams would always give me bad results

>>598876

Here's a solid/wire if interested

>>598880
I have too many questions every time you post, honestly I can never compute how you can nail those textures every time and how do you even create them. Not to mention that ice-cream sculpt, it looks so realistic. Everything is just too perfect. I really appreciate that bear in the woods scene you shared, it looked so easy and simple, but it isn't. And animation is just next level, do you have a link to some yt video which shows that function in blender? You explained it a few times and I still don't understand shit nor where to find it. And the lack of followers is really strange, something doesn't add up. I mean, come on, just go to reddit and post this in a few subreddits, preferably popular ones and you'll instantly get at least a couple hundred followers, if not thousands.

>>598876

Curious, how long did that take to render out?
Good job, looks pretty damn good.

>>598882

If Im making organic textures, I almost always use the cycles noise textures. They look really natural and I dont have to work on them too long. You can combine different noise textures with different vertex color maps on the model for influence control to get some cool blending of generated textures and materials.
I just add nodes and mess with colors until everything looks ok basically :s

The ice cream is pretty much all done with displacement modifiers and a blender internal noise texture. The screenshot shows teh extent of sculpting that I had to do. After you have that base layer done, just layer displacement modifiers over the top of it with vertex groups to control displacement

I'll have to give reddit a try. Thanks for the help :)

>>598883
I cant remember but with both my 1060s it was 1 minute 45 seconds per frame. So that would mean.. 5.8 hours to render? It kept crashing every once in a while so that added a few hours. I think blender crashes a lot with booleens on really dense meshes. The subsurface scattering adds a ton of time per frame too.
Thanks!

>>598885
are you rendering with a CPU? that's a bit too long for anim without reflective materials

>>598886
two Nvidia 1060 cards. Theres a lot of subsurface scattering on the ice cream and a little on the turtle skin that ate up all that time

>>598885
so you use subd surface and lowered it in the viewport?
the icecream looks detailed so i thought you use dynatopo.
or sculpt first, then baked a displacement map and added it to cycles?

>>598440
finished baking, working on textures for the hog. deviated from the concept because lazy

>>598890
well its all 3. I used dyntopo to make it look like the last screenshot. then I used displacement modifiers on the mesh (this looks like the first screenshot). Then in cycles, I used a noise texture to add finer details for the final render.

i have a wip for something related to skyrim and need advice for it but... its... not safe for work...

>>598899
hey man you gotta roll those dice.
Mostly if it's just a naked person in the t-pose or w/e you're fine

>>598906
ok well you see im trying to have the flaps of these pasties conform to the shape of the breast but when i try things like shrink wrap the meshes are mangled and chewed up.

>>598908
you've done a fair amount of work on them already so if oyu dont want to start over it would be worth it to just carfully move all the verts by hand

If you're up for starting over, or just a tip for the next set of pasties (or any skin conforming garment) duplicate the mesh and then carve out the shape you want. Then extrude it and delete the excess geometry.

>>598912
hmm ok thats fair
im really hesitant to start over considering these things are actually the mouth parts for a... creature... to be used in an animation so editing the verts seems like the most viable
(in hindsight i should have kept the rig i used to put them in that position)

>>597240
what's your lighting setup on this look like? i have a hard time with interiors like this.

>>598877
As i said, DISPLACEMENT map, not normals or bump maps - it´s a entire new type of map slot. This is in 3ds max, i mean.
And yes, seams are a nightmare and you have to battle against the bitmap until it looks right.
...last time i did this was in 2012. Gosh, i´m getting old.

>>598921
looks pretty decent but the shadows are pure black lol

>>598908
should be able to click the magnet and snap to face.
>>598921
yeah I know displacement maps. figured they'd have the same seam issue tho. this was for a humanoid reptile character, and places like the shoulders and face just looked bad with scales cut in half. I've seen someone make procedural scales in substance designer I think. haven't touched the character in forever though, ugly beginner stuff.

>>598923
so snap the face of the flaps to the face of the breast?

>>598924
its snaps vertices to faces as you move em. still vertex by vertex, but it goes where you need instantly at least.

>>598924
be careful of that though because if you grab groups of vertices they'll all snap onto the surface.

>>598927
ok so one vert at a time. would it help to do proportional editing? because the whole thing is solid all around, the flaps have two sides of faces to them and i want to make sure the verts behind the ones ill move will move accordingly

>>598928
should be fine to proportional edit. try it and see

>>598932
well i dont think im doing it right, i cant seem to get the verts to snap to the body as they are two separate objects

'Sup, the guy who made the cute bunny (and this username is my nickname). I'm giving you a teaser on the next 3D piece I'll be makin' in January... Oh, and she's from a porn comic.

>>598934
will she get a lot of mayo?

>>598933
yeah it works with separate objects. do you have it selected to snap to faces? The default is a different setting.

Rigging the outfit and hair is finished. Now I just have to make the different body type morphs and rig the staff.

>>598936
Yes, a frigging jar of it... Even though this is going up on YouTube as one of those time-lapse videos, rimjobs are out of the question.

>>598942
This is aesthetically really bad.

>>598934
Sorry if I missed it, but is there a place to follow your art? I personally quite enjoyed the bunny you made, love her rogue-esque aesthetic.

>>598942
This is all sorts of ugly. Not even the loli tit can save it.

>>598942
Good 3D takes time. Keep working. Don't feel bad to try over again, long as you learn something new and get faster. Also look for character already done in 3D and reference it.

Kinda looks like your rushing it to just get it done. Maybe look up snake witch or snake themed villains. Stick to few simple shapes.Maybe make the hair chunkier for cuteness?

>>598963
That character is already done in 3D. I'm just making clothes for it.

>>598815
3d takes time.
and if I'm gonna jump into porn, I will create other characters for it

>>598964
Tunic, short skirt, headband that looks like it has eyes. Herald staff with evil eye at the top. Hair that looks like fangs. Hair that looks like tail with beads at the end. Regret not making scarf that looks like tongue.

>>598978

The snek was an afterthought. I was originally going to make a broom but I'd already made a broom for another witch outfit I made over a year ago. So I thought, "maybe a staff with a giant snek head would be fun".

Collar for Sansar, almost done with the high poly.

>>598308
So I re-used that other kensu model as a base mesh scene I forgot how to make a butt.... Didn't have much time to work on it scene Dec27 till today. Doing the later today if nothing comes up.

>>598953
Usually, my art is of mixed quality. Some are fine, some looks like I could do better, and some looks like shit.

Hope you like DeviantArt: https://outmaced121.deviantart.com/

>>598685

Blocking out armor

PS:
Happy new year, !!

>>599047
I think both the armor and its edge trim look much too thin atm.

>>599047
i love him, i concur with making the armor fatter tho

although, he is a birb, and birbs are lightweight. so do wat u want

Is this ass looking ok? I keep thinking something is off about it but I'm not too great at anatomical correctness (especially idealized figures).

>>599053

>>599052
>>599048

) thanks for kind words,
This armor is mainly shape blockings, I want to make it more of a organic+metal thing, with eyes, teeth, fangs feathers n' shit and "evil" ornamentations.
I also want to somehow tie in Fish in this character. Not only birb

As lightwait goes, he's a mage archetype, I don't want him to be tanky, but we'll see.. I usually design as I go. I know it's bad practice, but I'm too lazy to create concept beforehand.

>>598515
The text and logos looks terrible... Why is all of that part of the normalmap?

>>598827
Why didn't you just use a ripped model

>>598891
calling this finished, want to move on to the next project

>>599054
it looks weird
like it's too high and not tilted enough.
Ass cheeks are too far apart (INTERGLUTEAL CLEFT).
I can't pin point SACRUM (you need triangular shape at the top)
You also don't have a nice GLUTEAL FOLD (horizontal crease between buttocks and HAMSTRINGS)
>>599053
That small hint of ILLIAC CREST is off, and I can't even see it from the back view.

the model has many problems, but you only asked for the ass. I am not going to scream at you to study anatomy, because you want to jump into making anime models as soon as you can, but I hope you are going to search for references of everything I wrote in CAPS, and try implementing them in your model.

It's easier sculpting a muscular model, and then smoothing everything if you need less muscular body type. That way you at least get the muscle group and bone structures right.

For models like that contour is very important, it's good to stand back far away from the model and imagining outlining all the shapes you see.

>>599115
and that's what people should do more often. The more projects you finish the better.

Alternatively working on multiple projects at the same time.

>>599115
Makes me think of some of the Donkey Kong 64 models, if they had a bigger budget.

>>599115
Oink, oink. Looks good hombre. Good job. Best of luck on your next one.

Getting there with my model of Elder Centipede, render on keyshot without texture for now.

>>599131
gee whiz how many triangles is that?

>>599131
Looks great, too bad King killed it.
How are you going to texture it? I don't think there is an official colored version of a cover or some promo art by Murata.

>>599132
I was thinking just now if maybe all those body parts could be instanced.

>>599136
>King killed it
saitama pls go

>>599141
>saitama
Who is that Saitama? He's not in the magazine.

I've been doing this for 16 hours /3/, but it impossible to do what Grassetti is doing. He just makes some random movements and it looks really good, I do the same exact thing and it looks crap.

Its impossible

>>599153
>16 hours
Did you start from a cube?

It's fine to start from a base mesh anon.

And if you don't know anatomy, at least have some reference mesh you can try copy.

You are not using a mouse, right?

>>599154
>Grassetti

I started with polysphere

That the thing someone told me not to start with a basemesh because that is cheating and you will never improve if you do use a basemesh.

I'm using a wacom tablet.

>>599155
As much cheating as using models for sketching.

If you feel like you want to learn everything from scratch, then practice making base meshes, and after your base meshes look good, re-use that one good base mesh you created yourself for practicing your next step in sculpting.

It saves you time
Even making a well proportioned model out of spheres, tubes and cubes is going to help you visualize the figure.

>>599132
500.000 or 2.000.000 depending on the amount of details of the body parts.

>>599136
I already finished texturing it. Gave him some Tyranids palette look.

>>599155
>and you will never improve if you do use a basemesh.

That's silly.

If you really want to try making it from "scrach", then do it with Zsphere instead.

>>599155
>That the thing someone told me not to start with a basemesh because that is cheating and you will never improve if you do use a basemesh.

that is absolutely stupid

sculpting from a sphere is fine - yes, it's a bit more work in terms of needing to plan your proportions/ drag stuff out etc and it's nice for shaking your dick around, but it's entirely acceptable for you to work from a basemesh first of all

How you make that basemesh is your decision; like >>599162 said, you could create a lot of primitives like spheres/cylidners etc. and orientate them together to make your base proportions:
https://youtu.be/tZnUgt659oI

with Blender, you can use the skin modifier - with a good reference, it's immensely easy to rough out the base proportions and then sculpt the detail and muscle into it with dynotopo, or with a bit of work nudging around the applied skin modifier mesh, you can have everything more or less plotted out at a low subdivision, then slowly build things up with the multires modifier:
https://www.youtube.com/watch?v=jAI4tnqv77A&t=30s (dynotopo example)

Then there's the more detailed base mesh creation, where you capture the basic forms in a first pass through poly/box modelling, then sculpting from that, with your proportions and basic forms already built in:
Part1: Box modelling https://www.youtube.com/watch?v=spi4lGxnMZg
Part2: Sculpting from box modelled mesh
https://www.youtube.com/watch?v=9oFHCMa5tlQ


Lastly, you need to get a good grip of anatomy, a decent knowledge of surface muscles, their origins and insertions, and bony landmarks visible through the skin. Hard for me to tell if you presently lack that, or if you are just having trouble actually sculpting it in - which doesn't surprise me if you're new and just starting from a sphere

>>599190
Tyranids work well, I can imagine being it part of the hive.
It looks great

>>599190
>>599208
> I was an old man that liked centipedes and collecting tyranids figures so much that I became a monster!

>>599214
Very possible back story for the Elder Centipede.
but I'm afraid to imagine what Black Sperm's past would be like.

>>598514
What do you mean by "second UV workflow"??

>>598827
hair doesn't need to connect to the scalp

every single limb could be a separate mesh if you wanted it to be, lad

>>599116
She is wearing a suit, so a lot of that detail is not as apparent (such as the gluteal fold / lilliac crest).

With this in mind, you said the model had a lot of problems. What else is evident? I don't want the suit to just be an extension of her skin.

>>599116
Forgot to add here >>599226 but I was working on it more yesterday as I noticed a few things you had said.

>>599191
What is the difference in cube or sphere? You moron..
you just learn proportions when you start from scratch. When you start from basemesh you'll just learn how to give proportions a shape. But when you don't understand proportions you won't be able to reshape these Base meshes in exact proportions. It sucks when you are an artist and you can not use your tools to their fullest.

>>599153
Actually anon this is not so bad. You are just givving it musculature too soon. Try to give it correct form first. And watch as many tutorials as you can.
Art is hard you won't get any results in mere hours..

>>599107
I do, but the Japanese hate rips. Plus the rips have limited facial animation
>>599225
By connecting to the scalp I mean closing this hole, the hair is a separate mesh from the head

Am I tacticool yet?

>>599227
Your limbs look like tubes, they don't convey basic shapes. Try forming triangular shapes, and then smoothing them.
And if the crack is that visible, then the fold would also be there
look at the white alternative https://www.zerochan.net/256645#full

Define back of the knee. You could do it by placing a flat cube at the back, and blending it into hamstring to form the tendon. In drawings it's often hinted by just 2 vertical lines.

There are no ankles, there are no Achilles tendons.

Lower back looks better.

Switch between different materials and lights as you work.

>>599234
>I mean closing this hole, the hair is a separate mesh from the head

so how about a solidify modifier in blender (I'm guessing you are working in blender)
check out https://www.youtube.com/watch?v=_jpZYlWVm8o he made hair mainy out of planes, divided them into strands, moved them around to form some folds, and thickened them with solidify mod

>>599256
Thank you, I'll play around with the Solidifying modifier

I increased my resolution
divided the model into basic forms, and started shaping them separately.
I'm going to work on the upper half, mainly urethra this weekend.

>>599269
and here is with the glorious puffiness!

I still need to make the hole in the body deeper to accommodate the genitalia, because it still clips through.

also work on the folds closer to the lips, because you can see stretching on the body

>>599269
>>599270
We had this discussion countless times; post your cunt in /ic/.

>>599270
>>599269
nice work

looking forwards to seeing the rest of this creature

is it some kind of crab or worm thing?

>>599131
>>599190
A first Vray textured render.

Should I keep the bottom legs fat or thin? Had to re-draw the side view.

>>599256
Hmm ok, I'll play around with it. To be honest I thought I made her pretty shapely already. I do 95% hard surface stuff both professionally and as hobby, so I'm kinda illiterate when it comes to humans.

I could exaggerate the shape further, but the style from the anime (super "spindly" for lack of a better word) is not exactly what I'm going for. Trying to make her more womanly and less kid-like. I think the legs are fine, but the arms do need more definition.

>>599273
I don't think /ic/ is 3d

>>599283
You keep me grounded

>>599341
your image planes should be unimpeachable dogma, but they should also be perfect. Redraw them if you want to make a change. The more little changes you make without adding them to the reference the bigger hole you'll find yourself in when at some point things dont line up.

>>599341
actuallynice drawing skill you have there. I'd strongly suggest to try out Zbrush!!
your skill will transfer and you'll feel like fish in the water right away.

just literally try to sculpt the head of that character. you'll be amazed how fast you'll be able to get results

>>599155
Sphere or not, don't make an entire human as a beginner. You'll get just pissed and frustrated because it will look shit even on your 20th try and 200 hours in. Start simpler, first make a finger, than a nose, then a foot etc do several parts step by step and learn the anatomy of each one, you will improve and get familiar with anatomy much faster. Just like when you model a house you need to model each individual part. Divide everything into smaller parts instead of directly going for the whole thing.

WIP stylized Townhall with handpainted textures

>>599297
And the model now available on Sketchfab.
https://sketchfab.com/models/34412d5384574fd4aa1a7691ea94f26e

>>599467
interesting how there's a sudden dropoff in topological density as you move away from the head. i guess that explains why the triangle count isn't as high as you'd expect.

>>599471
Yeah, I felt keeping some details on the first ones was enough, otherwise I had 2.000.000 polys, while decimating most of them lowered it to 500k.

>>599414
>>599371
Thanks for the feed back. I will dive Zbrush another try. I live on the East coast so i may not have power till late Sunday. I'm surprise I still have it now. Cooking potatoes and making ice.

I think this looks better.

Just looking at anatomy for sculptors and trying to match the reference.

>>599194

Maybe, it just lack of anatomy understanding and not modelling. I feel like it would be backwards if I go back to poly modelling.

Finished up my Prinny!

>>599534
looks like it's made of plastic. i would tone down the specularity a bit if i were you

gotta figure out the shader now

>>599536
where are the lewds

>>597489
IS THAT A FUCKING MANLET?

>>599729
They are learning

>>598980
For me please can you show me how to bake this onto a low poly? Would you just make a ring about the same size then bake it on with a ball?

>>599536
You got her rigged?

>>597863
Rough update, smoothed in horns and cap. Boole tools in for 2/5 engravings on horns.

Still need a good metallic shader. Be nice to have something active in the viewport to compare with game screenshots & references.

Also, this helmet has built-in ventilation spots in the top, should help with comfort.