/3/ - 3DCG

Wew lad, high level stuff.

>>767850
High enough for the average /3/ pleb.

Now get in the wagon.

Whats pinching

Hope the CG Police doesn't shoot me in the back.
Not really sure what you mean by pinching though OP. Assuming you mean on the bevels for the ridges.

Pretty easy to do just slapping modifiers on top of a plane though. Only really need around 12 verts to make the basic shape.

>>767862
Assign a chrome shader with a fair amount of roughness and show reflections. If there is pinching it will show as "pinched" reflections... you'll know it when you see it.

>>767840
I'm just a n00b, but thanks this was a bit more of a challenge than I thought!

>>767863
>>767862
Had to remake it since I closed it without saving, but I'm not seeing anything like that.
Even put the roughness down real low to check. There was a weird artifact that shows up on the inside when the roughness was real low, but it turns out it was just reflecting itself.

But yeah, since it's all done procedurally I'm pretty sure the wireframe is a nightmare, but having a clean wireframe wasn't a requirement.

>>767854
Holy fuck you were right, give these retards a simple enough shape and they do post results!

>>767871
>you still haven't posted yours
>because you can't

Blender

>>767870
Wireframe isn't as bad as I thought, actually.
Clean reflections makes for clean wireframes, go figure.

>>767873
bretty

>>767872
it's literally a 30 sec model why bother?

>>767877
so do it. 30 seconds. go.

>>767877
>>767878
>he couldn't do it

>that kid that used booleans but got the sharpest corners in the class

384 Faces
Blender

>>767882
typical blendlet retardation

>>767882
>no bevel
>unnecessary loop

>>767873
only correct answer so far. also nice lighting.

Done, now make me a sandwich.

It look squishy

>>767918
Did you add bevels everywhere? Also, use face-weighted normals.

>>767875
Now apply subdivision surface modifier and show the pinch

The new bevel options are really forgiving regarding non-optimal topology.

>>767943
WHAT DID I SAY ABOUT FACE-WEIGHTED NORMALS, YOU FUCKING SUBDIV REBEL SCUM

>>767945
who the fuck is face weifht lmao

>>767945
Also i'm not >>767918

>>767958
>Having Maya installed
>Using Blender
Based

>>767840

>>767978
Based

Actually, how do you do this?

I don't even know how to make a hollow cylinder.

Is this a step in the right direction.
I am also >>767983 btw. Sorta new at this, but managed to get all quads, yes?

>>767978
Now make the >>>advaced Blender anvil

>>767840
Render faggs are disgusting.

If I hadn't gotten too cocky could I have gotten away with it? ;_;

>>768018
Based noise texture

>>768018
Make sure your UVs aren't warping so hard next time. Make more cuts if you need so.

>>768023
It's okay except for where it's obviously stretched r-right?

>>768027
I will. I just wanted to be part of a thread for a change instead of lurking around them.

Wew lad, I actually don't know how to avoid pinching on corners and if I should even use bevel modifier before sub-d modifier.

>>768035
N-GONS!

>>768035

kek

>>767873
Best in thread

Fusion360

>>768037
you technically can't avoid pinching, you need to use some tricks to work around it

Can someone post step by step for this?

>>768049
Not a trick, It's called sub-d modeling

>>768035
Woah, so this is the power of 3d modelling...

>>768057
Spotted the feminist.

>>767857
seconded

besides, everyone isn't ameritard

I give up on this area, hopefully somebody can help. Made a 16 sided cylinder, created a hole, moved a few faces upwards, added bevel modifier with angle constraint and then subd. Can't get rid of the pinch in this area no matter what I try. Fuck me...

>>768065
>>768054

Me: >>767873

In Blender,
>modeled one "up/down" iteration flat out of a plane (with a couple subdivisions in anticipation of curving)

>add an array modifier to repeat 3 more times

>add a Curve modifier

>separately, create a nurbs circle

>back at the Curve modifier, set it to use the nurbs circle. scale the circle so your geometry gets as close to a closed loop as possible

>make sure everything looks good at this point

>apply all current modifiers. Highlight the vertices that don't quite join, and Merge them by distance

>Now, add Solidify and bevel modifiers to taste to add thickness and bevel

>set to Shade Smooth and use Auto Smooth in the Object Data > Normals section to control it better

>finally if you still need it, add Subdivision modifier but I didn't need to

>>768067
>finally if you still need it, add Subdivision modifier but I didn't need to
Ah, that's the problem here, I can do it without subdiv as well, and I can do it procedurally like you did, that's not the problem. But with a sub-d approach, and WITHOUT using fwn, I can't make this work perfectly, but I'm not sure what to do here. I tried with different bevels and with support loops, but nothing works great.

As I understood, this is supposed to be a sub-d challenge, you have to use a subidivision modifier.

That being said, are you sure you're still not getting some pinching as well if you're using bevels + fwn? It's too far away in your render so I can't see it.

I use fwn workflow more than I probably should, that's why I still kinda suck at sub-d and am slow as fuck.

>>767840
How's this?

>>768076
kek

Is that a reference to the sinking demonstrator boats? We heard the hilarious news in Europe.

>>768077
Yes it is! Turns out they did it to themselves too. Apparently they were all moving too fast and too close to each other, causing larger and larger wakes till they started capsizing and making things worse for rescue workers by panicking and zooming around making the wakes still even larger causing more damage.

My favorite quote from the article I read about it.

"The majority of vessels involved showed ZERO CONCERN for the well being of others, or the damage that they were causing with their wakes."

>>767937
It already has it dingus. It's like the first modifier.
I don't remember exactly, but it goes
>Subdiv
>Array
>Curve
>Solidify
>Weld
>Bevel

Having the subdiv at the bottom of the stack means you can give it as many faces as you want without any major fuckups because the source is just 4 faces.

>>768076
>>768079

go back to /pol/ deep state shill

>>768076
I like it. But why are those Democrats wearing MAGA hats?

>>768081

Did something similar

>Create a tileable pattern
>Array it 4 times
>Simple deform modifier, bend on a 360 degree angle on the Z axis
>Weld modifier
>Solidify modifier
>Bevel modifier
>Weighted normals modifier

>>767986

You've got uneven topology distribution, this will cause shading issues.

>>768077
>5 boats out of 7000 sank
Keep on gobbling that fake news yurofag, listen to your EU owners.

>>768101
Do it with subdiv modifier instead. Weighted normals can't be used for every situtation in production.

>>768106
>complains about "fake" news
>admits that what the news relate truly happened
Doesn't get maga'er than that.

>>768113
I can't hear you under all that censorship and absence of freedom. :)

>>768112
We get it; you want people to apply a subdiv to beveled corners without modelling specifically to mitigate pinching so that you can say "hah look you failed".
You might be retarded to enough to expect this to happen, but that doesn't mean everyone here is retarded enough to do it.

>>768114
Ah, yes, the land of cancelling. Nice freedom you got there, mate. Can't even say 那个 without getting fired. :)

>>768113
ORANGEMANBAD! Isn't news.
Boats sank, many of which supported Biden and some Trump. It's what tends to happen with boats and boomers.

>>768117
Better than going to jail for tweets and facebook posts. That chink professor will not be suspended long anyway even is he teaches nonsense.

>>768117
Are you a Euro talking about other countries not having freedoms?

>>768076
Good job anon. Oh no, not on the render, on fucking up the thread

>>768118
>>768119
>>768120
You hear that? It's the noise of maggers raging mad. And it's fucking beautiful.

>>768126
Alright bootlicker, you've had your fun and derailed the thread. Can you go back to plebbit now.

and another thread successfully derailed by stupid ass idendidy bolitigs faggotry

>>768120
Imagine thinking America is free. America is a shithole led by fucking retards. American patriotic propaganda is unreal.

>>768112
In my case >>768081 , I didn't need the modifier. In fact, it made it look worse imo. But..
>Weighted normals can't be used for every situtation in production.
Why not? If you export an object, then it'd export the way the normals are facing too. All weighted normals does is even out normal directions. I can't see a case where having it turned on when exporting would cause any problems. Especially when it's something that can be done manually by editing the normals and used to be done manually in some cases.

>>767840

>Anon posts funny render
>MAGAtards start shitting up thread instead of just laughing
Sure makes you wonder who the real snowflakes are.

>>768055
subd modeling is what causes you to have to add supporting geometry, causing the pinching, and that's why i said you need to use tricks to avoid it

>>768152
dont u talk about my daddy like that

I'm retarded.
Please tell me to kill myself gently.

>>768115
>>768148
If you push the model through different software back and forth, it can get fucked if it's made with fwn. There was a case in my studio where we had to use subdiv only, NO weighted normals allowed.

It would also make this easy challenge harder, I'm not trying to fuck with you. Fwn is not a solution for everything and you can't always rely on it.

>>768199
How can that be? Normals are normals, calculated with face weighting or otherwise; only their values change. As long as programs don't modify them, interchange should be straightforward. What kind of issues were you having?

>>768206
I wish I could understand it better myself, but from what I've heard it's that different software can calculate them differently and they can get fucked at some point in the pipeline, also it's harder to make LODs with them. Again, this is just what I heard, and I didn't have issues with them myself. But when a client wants everything to be able to be subdivided and art director tells you to ignore FWN, are you gonna say to him he's retarded like >>768115 said to me, or are you gonna do it like you're told to?

I personally use FWN wherever I can, but there are times where I'm just not allowed to.

>>767882
>>767986
Typical blendlets making the software look bad.
Go back to your doughnut & anvil

How do you guys bevel the bottom corner without any pinching? What kind of a bevel is it? If I bevel only the horizontal edge, I'll get triangles which will cause pinching. If I bevel the entire loop, I'll get pinching on sides. I can bevel it nicely along the curve of a cylinder, but corners simply won't work and I'm going insane trying to figure this simple problem out.

>>768055

Sub-d modeling causes the pinching you retard.

>>768112

The sub-D modifier causes pinching. I used weighted normals here

>>767958

And it worked. Keep your terrible advice for yourself faggot.

>>768230
Same here, I think. The convex corner is easy enough, but I couldn't solve the other. Any ideas? I'm sure I'll facepalm when I see it, it must be a simple thing, but I'm a nooblet at modeling.

>>768076
Fucking based lol

>>768230
>>768247
This works I guess.

>>768365
Damn girl, show me your wireframe.

>>768366
Probably would look clearer if I decreased the subdivisions

>>768368
Actually this one's better

>>767840

>>768372

>>768374
The maxfag is so strong it's not even using any support loops for the creases.

OP here.

This was my method. I struggled with this for like 2 days.

>>767840
Jokes on you, pleb, I don't use subsurf.

>>768378
I thought I was the only C4Dfag on this board.

>>768378
The way people model this is fascinating. So complicated for nothing.

>>768378
>>768365 here
I used subdivision surface, array and curve. Doesn't really matter what the topology looks like when it gets bent into a perfect circle anyways.

>>768382
Then explain it. Nobody wants to help people who are stuck here. Barely anyone even shows wireframe.

>>768387
I fail to see the challenge, are people used to modifier doing all the work?

Maxfag here.

This was my method. I struggled with this for like 2 minutes.

>>768390
Seems like it's a challenge about working smart instead of working hard.

>>768391
Imagine actually making that image when you could have just typed
>smoothing groups :^)

>>768390
What is confusing here is that 1 half of the people are doing this only with a bevel modifier because they say it can't be done with subdiv because it creates pinching. Meanwhile other half are somehow doing it with subdiv without any problems, like a maxfag above and that c4d webm. And neither seems to be cutting topology manually, they just add modifiers. And for my retarded ass neither approach works.

>>768391
Does anyone know blender's equivalent of setting up separate smoothing groups and subidividing like this? So in max subdiv takes those smoothing groups into account? If I did this in blender, I'd just get 4 separate chicken nugget looking shapes if I directly applied subdiv like that without support loops. It seems so easy here.

>>768391
Well if you already know what you're doing it's going to be quicker, it only took me a few minutes to do >>768378 once I figured it out.

>>768382
It certainly felt that way, but nothing else I tried worked. Bevel deformer gave one issue and edge loops gave another. I could have just used subdivision creasing/weighting but that felt like cheating.

Maybe the bevel modifier in C4D is just shit. Maybe I should just learn Max.

>>768403
In blender it's called edge creasing.

https://docs.blender.org/manual/en/latest/modeling/modifiers/generate/subdivision_surface.html#modifiers-generate-subsurf-creases

Solution for Blender users, with subdiv, no bevels, no edge creasing.

>>768408
Small update: to avoid one sharp corner where both sides connect due to the bend modifier, place SimpleDeform after Subidivion Simple, then apply both of the modifiers and weld vertices again. THEN add a second subidivision modifier and it will fix that.

>>768408
this

So toolbag 3 seems pretty cool

Took a few attempts but it wasn't too hard.
My two best looking approaches.

lmao look at all these retards doing unnecessarily complicated processes, just draw a couple of edges bro

>>768787
Is this bait?

>>767886
>uses the rounded edge shader to cheat

>>768408
that's how I'd do it in 3dsmax

>>768788
What blendchad there did actually works. Sloppily executed though, seems like he just arrayed and curved the pieces without welding them together

>>768787
>retards doing unnecessarily complicated processes

Do you think those edge loops are necessary? You realise they don't do shit, right? Pinching is caused by inconsistent polygon density, not specific topography. If you start with a higher density mesh the problem solves itself.

>>768807
Not him, but the edge loops definitely have a use.
I used basically the same method here >>768369 The "inconsistent polygon density" has no impact on pinching.

>>768767
Decent looking ashtray, bro

Also, thanks OP for the thread, I'm a retard so I've actually learned a bit from this. Would happily lurk in another thread where everybody uses different methods to make a simple shape.

I unironically made mine without, curves, arrays or bevels.

>>768831
Post it here then

>>767840
and it still fucking pinches REEEEEEEEEEEEE

>>768846
try this

There is a type of a workflow which is sometimes used, and it's a kind of a mid-poly approach. Basically, you're supposed to make a mid-poly model which should look like a high-poly (with nice bevels, curves and shading), but it should still have a much smaller polycount as it shouldn't use a subdiv modifier. However, that mid-poly model should be able to be subdivided nicely as well, i.e. it should be optimized for a subdiv.

I wonder how would it be done with this piece. Only support loops won't work because the model won't have any curvatures without a subdiv. So it should have a bevel modifier, but then the problem is how could it be properly subdivided because a bevel modifier will cause pinching.

>>768840
it's that one >>767905

Adding a new loop to tighten edges is one of the most satisfying thing.

>>768801
That's because of the array and the mirror, I didn't want to do the full mesh, but that can be easily solved anyway.

>>768807
>Pinching is caused by inconsistent polygon density

wrong, pinching are also caused by n-gons and triangles in strange positions, do you realize how retarded you are anon? With the extra edge what I do is avoid both and that the mesh is more subdivided where I really need it.

Tried my hand at it. Can anyone explain how/why what I did is wrong?

>>768864
Wel, how did you do it? Bevel shader on a simple low poly?

>>768220
>ask for help on a 3D board
>get meme words

>>768909

>>768408
>>768410
>>768915
Thank you!

>>768921
Nice, glad it helped!

This thread is a testament to /3/'s yearning to participate, learn and share projects and challenges. Too bad there's always gatekeepers to hold people down and belittle them, or trolls to derail.

>>768961
Humanity is like that, most people are regressive retards that love to hold you back or tell you why you shouldn't try something.
Or well meaning but simple people who think what is impossible for them is also impossible for you.

If you're in the habit of listening to naysayers you're never gonna accomplish anything.
If you are someone who succeeds you need to get into the habit of traversing pathways they don't even realize are there.

Think of /3/ like navigating a sidewalk full of dogshit, just take enough care to place your steps where the turd is not.
When you see a really vile turd that is fresh and glistening, don't pick it up. Don't interact with it. Don't smear it onto yourself.
Just leave it alone to dry out in the sunshine. Left alone will succumb to the elements and turn into a harmless powder which the rain then washes away.

>>767882
>384 faces for a stupid crown
anon, you can make so many fucking things with 384 faces and you make something like that

>>767882
you're lucky we are anonymous on 4chan

>>768372

Based Maxchad dabbing on blendlets.

>>768403
yeah, it is fucking easy
its why max is the best hard surface modeller

>>769369
It can be equally easy to do in other software, as long as you know what you're doing. Max is great, but no need to flex with it everytime, especially when talking about simple things such as this.

wtf i'm actually having trouble with this, i feel stupid

>>769373
its understandable if you are using blender, don't feel bad

>>769373
People posted tutorials how to do it in multiple programs, check it out if you haven't already.

>>768961
>learn and share projects and challenges. Too bad there's always gatekeepers
Some beginner faggot coming in and telling everyone to make a lamp or a spray bottle because they assume everyone else is beginners and also need the practice is the "challenges" that pop up here mostly.
No one gives a shit about whether or not someone can make a basic ass model.

The reason this is successful is because it's simple, but uses "advanced" modelling techniques to achieve a clean result. It's not a subjective challenge "oh wow, that's a nice looking lamp", it's an objective one "look at that pinching, you did it wrong". There's one result that is acceptable, but many ways to get there. Which is why this gets multiple replies.

The only other threads that work for this type of subject are shitpost threads like the duk thread or wallpaper paste. Which is arguably better since it brings out the "creative" side of /3/.

>>769368
Based blendlet doesn't know his free software can do the same thing.

Solid master race reporting in. I feel bad for anyone that spent more than 5 minutes with this shit.

lol mine definitely came out more ashtray looking

>>771082
that's ok it looks pretty good (for a blendlet)

>>768902
Ok but really I'm a noob, what did I do wrong? Because I refuse to believe I got it right

>>771085
thank you, I've been putting off learning 3d modelling for years now and decided this month to actually stick with it a bit longer

>>771089
1) /3/ is really slow, in fact if you get a speedy reply it's probably a shitpost
2) shade -> smooth

let's face it, this is the only correct mesh

>it takes 3d anon weeks to autistically engineer one little part

>>771190
Not versatile nor optimized

>>771077

>>768048
You cheated and used nurbs

>>768408
Why does it do this, what have I done wrong, it all looks just like the picture

>>771516
I had to make an "empty" and it was fucking up because it needed to be rotated, that's why. Now I took forever getting the angle and colours perfect

>>767877
>haha, you guys are wasting your time, he says
>while posting on 4chan

>>771516
Be sure to apply position, rotation, scale and center the origin because it uses that to determine how to manipulate the object. If you mess those up you can use empty object instead, but it can be a bit confusing, yeah.

>>771546
>draw a stick figure
>you're an idiot if you don't

why can't blendplebs do anything without using rounded edge shaders aka cheating?

>>767840
ur fucking joking right?

>>771516
literally my mums company logo

>>767958
>using firefox

>>771610
Rounded what? Are you having a stroke?

>>771501
OP didn't state you can't

>>771675
True, but he said "without pinching", which as far as I know doesn't happen with NURBS. It's typically an issue with subdiv meshes, so I'd assume that's the target.

>>771653
>he doesn't know

>>771646
>Using chrome spyware

>>771610
>rounded edge shader
Who used the bevel shader?

>>767875
You got tris in a poly model motherfucker. You are under CG arrest.

>>768149
4D users cant into poly modeling. Always showing off babbys 1st modifier and mograph garbage.

>>771854
On the inner bevels? Not a big deal. I could have turned up the bevel by 1 and get rid of it no problem. Though it doesn't matter. No deformation, and too small to make a difference on the reflections. It's a non issue and still satisfies the "non pinching" requirements.

Still, I hope I don't get softbodied in CG prison.

>>771793
the blendlets that didn't show wireframe

How much would you sell those for it you got those 3d printed? You know Halloween is coming up

>>767840
Got it done in SketchUp.

>>771938

>>771939
that looks like shit

>>771939
anon what is this

>>772019
Everything in SketchUp looks like shit, that's why only hobbyist like myself use it.
>>772024
>anon what is this
That's a visible face. Nothing is actually round in SketchUp, everything is literally just lines and polygons.

The autism in this thread is incredible

>>771922
I think there's one in total

are pol like hackers that are racist

>>771645
Is your mom a blendlet ?

>>772612
Yes

>>767840

>>768378
I like your workflow. more viable than the "Make a billion vertices, then there's no pinching" approach by others in this thread

How in the fuck did I manage to do this

>>773194
Show wires, show modifier stack (if there is one), presumably there's some weird viewport shading going on as a result of inward facing normals and/or culling fuckery

>>773194
Reversed normals and backface culling.

I'm kind of new to this, so don't bully me too hard. I'm autistic and slow so this took about 3 hours. I could certainly do it faster the next time, and better by implementing a normal map and using more faces on the cylinder and fewer on the bevels.
Also maybe I just missed something but fuck Blender for not having edge/face selection saving. It always takes some weird round-about way to select the specific things I want without just selecting everything manually. (which is fine for this specific thing, but might be a problem for something bigger)

>>773905
>fuck Blender for not having edge/face selection saving
I believe it has selection groups, no?

>>768035
clicks, nice

>>768076
top kek
3D shitposting = best posting

>>773816
>>773810

Backface culling was off, I made sure.

idk, it's something. Will show wire if anyone actually cares.

>>773905
I've used vertex groups and also even assigned specific materials to specific faces to select them later

>>771568
Hello idiot

>>773114
shit

>>773905
this is not good

Took few mins in moi3d, and I cant even model. It there more to it or what??

>>767840
This topology doesn't give me any pinching but it still doesnt seem right to me. Using turbosmooth. any tips?

>>773905
>using more faces on the cylinder and fewer on the bevels.
Starting with cylinders with more sides solves 99% of problems with using cylinders. Also let subdivision help you round your edges, not excessive bevels.

I can't solve it

Prop modelers are truly the lowest of the low. How can you peasants waste so much thought on a mundane everyday object? Even modeling rock formations is more interesting and rewarding.

>>775235
Have you followed a course on modeling? Looks like you lack fundamental knowledge, for example about holding edges, smooth shading, etc. Don't worry, it can be solved.

Nobody cares about your topology autism on a static 3rd rate prop

>>775253
>>775390
It's not about this particular model, it's about modeling principles. Damn retards.

>>775235
Second one cording to
>>768408

How can i weld it?

>>775423
Nevermind

>>767918
I'm this guy and I gave it another go. Still really new to all this but look! I gave it a shiny new texture! woah!

Huh, that was more of a challenge than I thought

thanks op

>>768035
as expected of 4chan grown models

>>767840
Did I do it right

>>775974
no of course not

>>775974
does it fucking look like you did it right

The answer to modelling within curved surfaces is usually to have more geo than you think.

i cheated like a fuccboi
before you ask: yeah its subsurfed 3 times, no it wouldn't pinch even if i moved the top edges

Close enough

>>771930
>How much would you sell those for it you got those 3d printed?
I would use a retard reaction image but the one you posted already goes perfectly with your text.

>>768067
This is perfect. Made the thing in a minute or so. Thanks for teaching me somethiing.

tbf I just looked at OP's image and did mine before looking at all the shit you're all nitpicking at. my basemesh is clean though.

Next challenge when?!

>>767882
simplicity at its finest

>>777440
Unfortunately, it's wrong. It won't subdivide as desired.

It's okay for a low poly model though.

>>767840
Modeled it in both mesh and NURBS hope you're happy anon. did it in about 8 mins in meshes and 2 minutes in NURBS.

make it something harder next time.

I copied and then tried to do a volumetric sim and fucked it up so it was unusable but here's the homework, teach.

>>777112
>all these triangles

>>767882
I'm new to 3D modeling, what's wrong with this?

>>778248
It doubles the faces being calculated with the useless supporting loop in the middle while adding no detail, and it has hard, unbeveled edges which exist nowhere in real life.

least polys itt