[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Settings Home
/3/ - 3DCG

4chan Pass users can bypass this verification. [Learn More] [Login]
  • Please read the Rules and FAQ before posting.

05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
06/20/16New 4chan Banner Contest with a chance to win a 4chan Pass! See the contest page for details.
[Hide] [Show All]

Meta on /qa/ only.
All meta discussion of boards is to be redirected to /qa/.

[Catalog] [Archive]

File: 1510314478593.jpg (438 KB, 1000x1414)
438 KB
438 KB JPG
Hello I'm asking your opinions on which renderer is better, before I spend hours studying it:
>Vray, Arnold, or Renderman

>Arnold is CPU-based
>V-Ray is CPU-based
>V-Ray RT uses CPU that can be used alongside the GPUs.
>Renderman rendering uses CPU+GPU

Should then renderman and V-Ray RT the best options?
I know some may say that my pc isn't that powerfull I have an i7 6700, a gtx 1070 and 16gb of ram is what I have and will have for some time.
Which rendering software would be the best for my machine?
2 replies and 1 image omitted. Click here to view.
GPU rendering is the future. CPU+GPU is a half-measure and CPU rendering might as well be fossilized dinosaur shit.
They all give good results and are industry proven.
Renderman doesn't use CPU+GPU yet, that comes in the next release. Currently CPU only.
Ignore Renderman and Vray for now, get Redshift and Arnold and learn these. Both are easier to learn.
Don't talk out of your ass. GPU renderer are not better, only faster, but the image quality of CPU renderers might be a little better.
>GPU rendering is the future
Well I can imagine that using multiple GPUs will probably be more powerfull than using dual CPUs
But if that the case GPUs price may skyrocket like ram prices did this year, something that may should keep in mind.
>Ignore Renderman and Vray for now, get Redshift and Arnold and learn these. Both are easier to learn.
Redshift looks very good, but should I really learn both?
I mean Redshift and Arnold.
Redshift. Gpu rendering is awesome
Arnold is piss easy to understand. Give it a couple of days of learning and you've got it.
Yes, learning more than one tool to get the job done is recommended.
Don't put all your eggs into one basket and don't be afraid, learning a renderer is mostly fun and easy.
Like i said Redshift is faster, but in some cases Arnold just looks better.
Learn Redshift first, if you 've got it, switch to Arnold.

Check out this channel for learning Arnold:

File: Bar No Frost.png (3.1 MB, 1920x1200)
3.1 MB
3.1 MB PNG
just wanted to know anyone's opinion on how the render looks.
31 replies and 4 images omitted. Click here to view.
I want realistic representations of what I'm getting.
This is why people were so pissed off when Nintendo started producing amiibos. The quality was nowhere near what we were shown.
The table is alright, the background looks very low def, as if you used a very small image of bokeh instead of actual lights, and the glass is missing a roughness map and dust. You have to give it some imperfections to make it look real. If you're trying to achieve a "comercial" look, then you don't need to, but lighting needs to be better. Give the table a rim light to separate it from the background, do something interesting with the glass reflections.

>oh look a new thread, maybe anons will actually try to help this newfaggot
>becomes software shitflinging within 20 posts
That's why I stopped coming to /3/, because it's more of a "Nuh UH my dad can beat up YOUR dad!!!" cesspool than anything useful for modeling
It does look like that. That is expensively made ice for an expensive establishment.
Imagine being so poor that you have never seen crystal clear ice in your entire life.

Is it possible to trace an image and model it in Blender? If so, i wanna learn how to do it.
13 replies and 2 images omitted. Click here to view.
Learn to draw. Understand basic shapes. Its easier to do it the long way then to try and find a short cut.
I don't understand the point of this? It *looks* like there's some height information but it'll look awful regardless of what you do with it.
Look up "camera matching".

Sadly, it can only be used for props and environments. For characters, you've got two choices: either modeling them in T-pose the proper way, or carving them out of a volume like a traditional sculptor (which means you won't be able to pose it afterwards). Either technique requires good knowledge of anatomy and sculpting, at which point you'll find you won't really need to trace an image.
Just redraw the character in t-pose and work from there.
This, and if you can't draw find an artist who can and commission them. Keep and eye out for somebody with solid anatomy, even when they are just sketching lose stuff. Everything else is secondary to that. Also talk to them an make sure they are aware of your needs; i.e. orthographic front/side view, with proportional consistence.

File: tight-loose-vagina[1].jpg (26 KB, 651x393)
26 KB
Wow im a little surprised, I cant seem to find any tutorials on pus pus modeling.

Anyone have any good tutorials on it? Also is it possible in Zbrush? I imagine if I set a blend shape to spread the legs open i could get in there and do the sculpting and then close them back when im done... I can try it in Maya too no problem but Id like to stay all in one program.
38 replies and 6 images omitted. Click here to view.
I got some idea after looking at https://3dexport.com/search.php?hidden=521640&search_keywords=vagina

but it looks like everyone just models that one perfect vagina from every porn movie
What’s with the exaggerated fundus enough hips to track a herd from out there yonder...
call a hooker, scan her vag with a 3d scanner
>3D Scanner
That's stupid. Use Photogrammetry.
I like it

Daily reminder to stay positive and ignore the haters. Stay positive all around. Stay based - lil b
Shit, wrong board, sorry, fucking Clover...
So the haters are right. I knew it was time to give up on /3/

File: ue4.png (31 KB, 254x207)
31 KB
It's time to start a new thread.

Some good new videos are now available on UE's YT channel. Check out this one about lighting, I needed something like this so much a while ago, but it is still extremely useful.


I also have one question, maybe somebody will know the answer - how can I make my deferred decals have smooth and faded edges?
303 replies and 32 images omitted. Click here to view.
haha loser.
I bet you're the funniest guy in the high school.
Need some advice. I have an object that's kind of a cross-beam similar to pic related; essentially a center junction surrounded by four struts that end in anchor points. I have a modular piece for the center, the strut, and the anchor, and they all neatly attach at 45º angles.
The thing is that the length of the struts are going to vary depending on where they are, and placing all of these manually is going to be a waste of time, so I know that I have to set up a construction script that can do it for me, but I'm not entirely sure how yet.
I know how to use splines to draw out stuff like tubes, and so far I figured out how to spawn the anchors some arbitrary vector locations, although as it is, it's an ugly set of Add Static Mesh Component -> Make Transform -> Get Anchor Location pasted four times, with the make transform scale used to flip the anchors around to face the center point. But I'm not sure how to actually repeat the strut between the two points, so that it maintains its length until it needs to add a new strut instance to make it longer if you drag the endpoint far enough.
Am I doing things backwards, or what is the simplest way to do this in general?
File: StrutFunction.png (304 KB, 1711x725)
304 KB
304 KB PNG
Ended up figuring it out myself, with the help of various resources and videos, although it's probably horribly inefficient, being my first BP and all that.
I packed it as a function that takes input from my end-point vector, gets the distance, turns it into an index, and uses that index to multiply by my grid snap amount.
Since I couldn't figure out how to repeat the function for each strut automatically, I simply exposed booleans that ask me which strut I want to create, and select the distance stepping in the axis combination that makes sense for the one I need.
The two extra inputs are the scale of the endpoints and struts, so that I can flip them in the axis I need so that they point in the right direction.
But at least it works, I drag in my center junction with the anchors at their minimum distance, and as I pull out the anchor widgets it creates sections of beam behind them that line up every 20 units.
should i be using the film tonemapper?

File: lovebrush.png (701 KB, 484x483)
701 KB
701 KB PNG
ZBrush 4R8 Love Thread.

Holy shit anons - I used to fucking *H*A*T*E this insane bullshit sculpting application.
It used to literally raise my bloodpressure. I would get a tension headache just thinking about it....Until I checked out the newest version...

Holy fuck. I get it now. I totally get why all those fucking weirdo masochists used to defend this shit. Only, I can't call them masochists anymore - because this shit now actually *works.*

Don't get me wrong - there's still a ton of legacy-workflow shit that is just totally annoying - (yes, shit still goes crazy if you do anything but hit 'edit' after placing your first tool) but on the whole the insane asylum that is Pixologic is gradually coming back to reality:

3D gizmo to move / rotate / scale. Yes - just like the one in Maya, Max or whatever. Yes - you just tap 'W' and you have the thing pop up to move your object around. Yes! It's true! I know right?!
You don't need to do the 'hold alt key, drag, then release' to zoom in and out! It's still there - but you can just use the Ctrl key now! Holy shit! Amazing! Just 'ctrl + drag' and you're zooming in and out! Incredible!
Live Boolean? It works. The icons in the sub-tool palette to use it are kinda indecipherable, but it works. It's good. I like it.
Scale master? Eh... it's still kinda dumb but it's a start.
Was the FBX import/export there before? If not - this is also a very welcome addition.
Image planes! Reference views!

Comment too long. Click here to view the full text.
30 replies and 1 image omitted. Click here to view.
When are fuckwipes like you gonna discern the difference between not being able to use a thing, and pointing out that the thing could obviously be improved?
You want a weld that looks like a machine welded it, or do you wand a weld that looks like some beginnerfag welding student did it?

If you want the welding machine look, that's super easy--those things are fucking amazing. All you gotta do is get ONE bead correct and then clone them around.
At the part where he makes a knitting pattern: https://youtu.be/dvCompIskoY?t=191

Shitty welds require more attention to detail.
If you need references, just go to /diy/ and make a thread asking people to post their welds. It's happened before.
hey I like the transpose tool tho, maybe i just got used to it and I still like it more than the new gizmo.
can one of you fags explain to me how the fuck I close a project in zbrush? there doesn't seem to be an option

yes I've read the sticky

File: Atlas Rig Day 2 small.jpg (223 KB, 1281x810)
223 KB
223 KB JPG
Anyone have a good resource for rigging and how to set up IK properly?
Just need to be pointed in the right direction, please. Google turns up a bunch of stuff, but it's all different - not sure what's good-practice and what's bad-practice.
I should note that I would like to understand the general theory behind it, not software specific implementation instructions.
Think car controls, what do you need to do to get it moving, where the steering wheels starts, where the breaks are (gravities forces) etc
File: Sexy biped.jpg (96 KB, 816x624)
96 KB
I´ve been using Biped for 3dsmax for eons.
I rarely rig a character outside a biped, it has so many advantages, but it´s a bit weird to set up and get used to. Prime system to get motion capture data.
Damm Im struggling with this myself. Im thinking of taking a paid course
My own approach is going to be to pirate a bunch of tutorials off cgpersia or the like, and hope that I can come to the right conclusion via aggregation of many sources

why can't nobody crack it?
why don't all software use their anti piracy protection?
how can you branch with x-particles 2.1?
will x-particles 4 be cracked?

>>inb4, buy it you cheap fuck
1 reply omitted. Click here to view.
They’re using the system that Denuvo was derived from, so practically half the program is anti-piracy measures. Yes, Denuvo has been cracked, but games are a more mainstream product, and so there’s more demand for a solution, but who the fuck wants X-Particles? Like 10 people?

Honestly, fuck C4D and just go full Houdini, or at least Max with Thinking Particles and Phoenix FD. There are much better solutions that already exist and are cracked, no use in waiting for a ghost ship.
Its using Denuvo
>why don't all software use their anti piracy protection?

Because it puts unnecessary load on the CPU
How hard is it to learn houdini?
Pretty hard, but the interface is very nice, so it's not like it's hard because you'll be fighting with just trying to find out where everything is, but that said, it's the endgame for particle effects and has been for a while.

File: 1498359104515.jpg (9 KB, 225x225)
9 KB
>watching tutorial
>they keep hitting the enter key or lmb way too fucking hard
62 replies and 25 images omitted. Click here to view.
Does anyone has any tips on making tutorials? I made a thing recently and some people have been asking me for one
Plan out your talking points before hand, don't do it freestyle. Don't add music. Make sure your voice is loud enough. No mumbling, holding a pencil or something in your teeth and try to talk with that for a few minutes before recording to help enunciation. Legible UI. Screencast keys or your program's equiv to see what keys are pressed. Don't waste time on an intro just say what you're gonna do with an image of what the end result will be, then get to it. Put useful keywords in the description or tags so people can find it. Final product in the thumbnail.
>Plan your video beforehand.
>Practice so you know exactly what you're doing and what the results will be.
>Don't go fucking around looking for things in menus, know where it is and go straight to it.
>If there are multiple ways to access a feature/function you're using (menus, shortcuts, whatever), mention ALL of them and where to find them.
>Explain exactly why you're using a feature/function.
>Screencast your keys, use subtitles where necessary.
>If it's a lengthy tutorial with multiple parts, list time codes and chapter titles at the start of the video.
>Don't stammer or slur your words, don't say "eh" or "uh" at all.
>If you're no good at talking while recording, plan your dialogue and dub it over afterwards.
>Unless you absolutely require additional audio, the only sound anyone should hear is your voice.
>Fix your audio levels.
That depends. If the guy just fucked up at using a feature it's pretty bad. But if it's part of an iterative design process I don't mind as much, seeing the different tries and mistakes is a part of learning as even experienced designers tweak their models a lot.
why do you have that image of that mudshark ofspring on your computer
you cant seriously not see the canthal tilt, ugly nose and facial dismorphia
jesus fuck i hope your not a neil druckmann pajeet nigger.

File: Capture.jpg (149 KB, 1090x545)
149 KB
149 KB JPG
Substancefags BTFO

>completely new UI

40 replies and 10 images omitted. Click here to view.
desu i cant see anything from this video that would make me want to swap from substance anon

its not like substance is expensive either
this! i' ve only stumbled on one literally one Mari skin painting tutorial....
mari supporters.. please show us where we can learn it for single users.
I'm open to softwares. I like Quixel, because you can litrerally work in open project/layers with photoshop and quixel doesn't notice and updates the project., this workflow gives you huge flexibility.
I don't care about 16k images tho, I need reason to use Mari
>I don't care about 16k images tho, I need reason to use Mari
animated textures

I wouldn't recommend to swap to Mari, i would suggest to learn it and use it when needed in addition.
Nothing beats Substance or Quixel when it comes to speed and the texture driven smart mask workflow isn't really there in Mari.
Mari is for actual painting and unless you need the power Mari gives, it makes no sense to use it since you are way faster with alternatives.

The original Foundry channel on YT and Vimeo has some beginner tutorials which are good enough to get you started.
>I only use Mari to texture organic stuff. Faces, monsters, eyes, etc. The only reason I use it is because I think it's workflow and tools are better suited for organics but I'm sure you can get same exact result in SP.

File: Capture.jpg (57 KB, 1176x632)
57 KB
i have this bug in substance painter and i dont know how to fix
it and i know that theirs board for this /qst/ but they wont even replay to this question.

the problem that i have is that when i paint on one specific part of the model
it also involve other parts of the model

>how can i fix this
16 replies omitted. Click here to view.
I'm fairly sure we're being trolled here dude. Let's leave the little 'uns to their fun. No one can be this dumb for real.
>export the entire mesh with the same material out of whichever modeling package you use

You don't have to do this. Using multiple materials will give you multiple texture sets. Let's say you're texturing a truck and it's so large that you can't fit all the shells in one texture at a reasonable resolution, and it also has some glass that should get a shader you've made in your engine. You can assign multiple materials before exporting and do uv layouts for all of them seperately and it can still be one model but in painter you will have 3 texture sets to use/paint on
Or use UDIMs... but dude; OP can't even scale a UV set correctly... you're wasting your time...
Your uvs overlap, this is so obvious its painful
when uvs get out of coordinates they get wrapped back around on each other

Indeed, in summary - OP is the bug

File: cb9.jpg (209 KB, 1100x900)
209 KB
209 KB JPG
All you will ever need.
72 replies and 15 images omitted. Click here to view.

Autodesk woke as fuck
But the guy wearing a blender shirt lol

He also has his arms crossed, which implies he's insecure.
You do realize the guy next to him also wears a Blender shirt?

Kek at these blendlets trying to exist

File: mray.jpg (73 KB, 1239x1361)
73 KB

>Nvidia attributes the reason for retiring the seminal renderer as a change in focus. Under the question “Why is Mental Ray being discontinued” they will that “NVidia will focus on bringing GPU accelerated ray tracing technology to every rendering product out there”, listing OptiX, MDL, IndeX, and IRay for CAD as technologies that will benefit from further investment.
15 replies and 3 images omitted. Click here to view.
grandpa software.
Looks like another retarded Blendlet can't comprehend, as usual. SI|3D was a godlike piece of software. First thing I learned, together with Alias. It's next incarnation -XSI- shat all over Autodicks puny efforts. Autodick knew that and killed it.

Ergo: fuck Autodick. wank software by wankers for wankers
Everything about this is extremely aesthetic.
File: torvalds_bird.jpg (27 KB, 442x293)
27 KB

You're joking right? We're talking about Nvidia here...
a new characters for rule 34

File: 3D printed toys.jpg (41 KB, 810x456)
41 KB
Here is a link to the video: https://youtu.be/GGp4gqnFaLI
"Like and subscribe"
"Like and subscribe"
"Here's an arrow to the subscribe button!"
While I am interested in this topic, this video of yours is insufferable.
>shit video annotations
>shit camera recording
>1 minute wasted on Makerbot settings
disliked, reported and shat on.

Delete Post: [File Only] Style:
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.