[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

Thread archived.
You cannot reply anymore.

last one hit the bump limit >>595914
File: fumefx preview.png (471 KB, 966x586)
471 KB
471 KB PNG
I watched some vids on FumeFX and I wonder now, is there anyway to separate maya's own fluid simulations from viewport like fumeFX does?
File: Untitled-1 copy.jpg (794 KB, 1922x1018)
794 KB
794 KB JPG
What the heck is going on with this windshield?
I verified there are no duplicate vertices and the faces are all connected. Also smoothed faces.
Texture is just using the glass shader.
File: T.rex.png (97 KB, 1668x619)
97 KB
How would you unwrap a mesh like pic related?
I accidentally added a sub surf modifier too soon, and I'm not sure where to put the seams now so they don't show.
File: 1516319022266.png (91 KB, 1668x619)
91 KB
i did a rough sketch but you get the idea

organic unwrap is simple, simply draw a border with a loop then make a cut across it by selecting it horizontally

use hdri and see what results you get
File: Untitled-2 copy.jpg (486 KB, 1920x1080)
486 KB
486 KB JPG
I think its something about the model itself. It's a poser asset.
File: simple house.png (20 KB, 781x449)
20 KB
Newbie here. How do I make custom textures properly? I've made this house, but I want to texture it in a very specific art style that I'm unable to do with texture painting within blender.

I know how to unwrap UVs, I just don't know the procedure to properly texture those UVs outside blender.
camera distance/range wrong
maybe some bad glas shader
File: Head.png (391 KB, 809x1044)
391 KB
391 KB PNG
How do I get rid of all these excessive faces?
Theres like 15 of them crammed on there.
you can alt select a loop press delete and select dissolve faces/edges/verts accordingly
I'm looking closely and I see very narrow double bands that might contribute to the problem. Removing doubles won't help because it looks like they are slightly separated. You are going to have to dissolve those edges.
It's either the doubles or its that you have faces joining the two sides of the glass windshield (perpendicular to the windshield faces)
File: wtf.png (221 KB, 700x500)
221 KB
221 KB PNG
I created a low poly mesh, ported it into zbrush, sculpted and painted it, created maps for normal, diffuse and displacement then put them back into blender and put them all on my model.

And this dude still looks like this. What have I done wrong? I made sure the maps were faced the correct way and have them all slotted into the right nodes and the details still aren't coming out.

1) You need to select "Non-Color Data" instead of "Color" in the Image Texture

2) Normal Maps need LIGHT to do their trick. You have flat World lighting going on. The same is true for any sort of metal material or glossy material; if you have shitty lighting, you will get shitty results.

Google "3 point lighting setup" as a starter
I forgot to mention that for a Displacement map, you don't just connect it to 'Displacement', you can get away with that if you have to by using a ColorRamp (I think, or maybe it's a Multiply node) to amp up the effect, but you should really be doing a whole node setup with Bump + Mapping to control it, or use the Displacement Modifier (if you want true displacement of the geometry)
File: wtf2.png (528 KB, 1653x877)
528 KB
528 KB PNG
Well, I'm getting the normal map to actually do something now. But the displacement still doesn't do anything even after moving the value to an extreme amount.

I guess it's time to figure out how to use nodes now. More googling tomorrow.
How do I fix a lumpy Zbrush model when smooth is not enough?
trim brush
In Marmoset Toolbag, the global illumination combined with camera animation is giving a crap clipping effect, is there any way to fix it?
add a normal map node
you shouldn't use a displacement map tho
What art style?
File: Screenshot_14.jpg (91 KB, 1560x715)
91 KB
How should I attach the thumb to this exceedingly bad hand?
make the rest of the hand, and a make a thumb that ends in an 8 vertex edge loop. Then you can delete a group of 4 faces wherever you want and attach it there.
I mean at what angle etc.
Full disclosure: I've got no hand refplanes
File: Screenshot_15.jpg (17 KB, 649x435)
17 KB
Is this a hand? Is this what "hand" looks like?
File: Screenshot_23.jpg (42 KB, 338x535)
42 KB
How many holes in spacetime will I open up if I do this?
oh. anatomically speaking, a thumb is usually at rest a little on the underside of the hand. though it doesnt have much trouble coming up to be beside the rest of the fingers too.
anyone got experience with lip rigging in blender. bendy bones looks promising, but Im having trouble figuring out parenting and bone constraints. instead of a control bone stretching out and pulling on the other bones, it takes whatever bones are in front of it in the hierarchy with it.
File: texturemalfunc.png (1.23 MB, 634x540)
1.23 MB
1.23 MB PNG
noob here, using blender, when I go to render an image some of the faces aren't applying the texture. The same areas lose color whether I vertex paint or unwrap and use a .png texture. How do I fix this??
no it's actually because there is a model underneath. Hidden models will still render so you have to uncheck the render icon not just the eye icon.
Much appreciated!
File: weird_noise.png (1.14 MB, 1680x1050)
1.14 MB
1.14 MB PNG
why is my noise looking dumb? it looked good when it was just a one face plane but as soon as i added more faces it all went to shit. I thought Maybe it had to do with hard/creased edges but that wasnt it. then i thought maybe it had to do with texture coordinate/mapping issues but that didnt do anything either... I'm kinda stuck here.
SO i found it only screws up when i move my vertices/faces around...not when I make my mesh more dense. Some progress at least.
should I just start with Blender over 3ds max (have a free student license)? seems like there's a larger amount of tutorials targeted at newbies
k so found the solution if anybody cares :/
needed my texture coordinate node to be on object. after that it looks fine.
Okay so I've figured out how to properly export UVs outside blender to texture in another program, and holy shit it sucks. It's so time consuming. How the hell do you guys do it?
>export UV's
you don't export UV's, you simply exort the mesh file (like obj) and it contains the UV's inside
try 3Dcoat for your texturing experience, blender texturing is not so user friendly, go back to it when your'e a little more versed in texturing
If you have the choice between anything and Blender, choose the former. You can always fall back on Blender if you need to, but I tend not to even bother answering dumb questions if I see its UI in the cap.
what's the best 3D software for celshading/npr?
In terms of quality rather than how fast you can create one.

I'm too poor for Psoft Pencil+ 4.
I've seen people use softimage to render their scenes, should I try that?
Rigging kills me. I've never been able to create a proper rig.

Is it really that hard to rig for everyone? For those who are able to create a good rig: which program are you using? Any tips on not fucking things up?

I'm using 3ds Max, which I know it sucks. I've been told that Maya is best for this, but 3dsM it's the only program that I know how to use

Softimage + normal editor plugin (free)

Something that stills get updated?

I don't think anything can't beat that combo you're using.
The thing here is that you're getting here fucked not because you know shit, but because you're too afraid to experiment with other programs.

The point, You can use or Maya Or blender.

Blender is super easy and free, also it has a 3d max viewport controls layout already set up.

Maya has max controls too and rather than fast is literally instant, and even easier than blender, the only con aside money(which is not), maya sometimes crashes weirdly, try to not customize the ui either.

Good luck.
working on rigging in blender. Body rigging is easy, working on facial rigging. The mouth is convoluted right now, you cant just make a loop of connected bones to deform with, have to do workarounds. And learning resources are very sparse for facial rigging in Blender, just about all of the tutorials are for beginners and only covers simple skeletal movement.
im sure there are good tutorials for 3DSmax but surely most of them are paid
hey i was wondering if there were any good resources for learning how to model such as video series/books?
for what software?
Nuke, after effects or fusion?
Which one is the easiest to learn?
I've been reading that both nuke and fusion have free versions for personal\educational use, are there any downsides there?
What if im a 2d artist and only want to use 3D as reference material for my work. Is there any better solution for referencing people than Daz? it's not very anatomically articulate and the deformations are shabby.
a program called Anatomy 360. I have never used it but it's made for that purpose. I tried looking for a pirated copy but I have not found anything yet
File: bendy.png (114 KB, 1645x939)
114 KB
114 KB PNG
bendy bones aint bending at all
Hello, a question for the sculpt masters, if I'm doing a hard surface sculpt should i still follow the same workflow as dynamesh/zremesh/subdivision or just go straight into dynamesh and then retopo, i have done mostly organic sculpts but I'm in doubt if it's the same way.
Try out Blender with the ManuelBastioniLAB addon. Both free, the latter is a character creator of sorts, allowing you to create models with specific measurements and pose them after. Quite powerful, and I guess pretty accurate for anatomy. I'm using that right now as a more flexible mean of studying proportion and general pose referencing.
alt click on unwanted edge loop. hit GG, slide/move that edge loop to overlap another edge loop. repeat process until all unwanted edge loops overlap each other. press a /select all. remove doubles.
huh I favor dissolving edges but that's kind of creative. I never thought of doing that
You have to set them like they are one chain, without the intermediate control bones.
Also their settings are in the bone settings panel.
File: Render_1.png (515 KB, 960x540)
515 KB
515 KB PNG
I made this after following a Blender 101 course. I'm fairly new to the whole 3D thing as I just started literally a couple of weeks ago and seeing this rendered made me very happy. However I want to do mostly stuff for games and I know so far that this hard surface stuff doesn't work for that but it interested me. Should I keep going on learning hard surface if my objective is to make stuff for games? I mean if those are two very different disciplines or if learning stuff in hardsurface will work for low poly too and all that.

Also I plan on finishing the 101 courses but is there any other course worth seeing after that?
>I know so far that this hard surface stuff doesn't work for that

it does.
...what? hard surface is extremely useful for games. It's literally the first thing you should be learning for anything in 3D.
File: Untitled.jpg (157 KB, 1366x764)
157 KB
157 KB JPG
HELP! I can see my light source... Works fine but I can't point.

PD:Show cone is activate
PD2: I am a noob
The individual chains don't bend either.
>is there any other course worth seeing after that?

yes, cgcookie have tons of good courses, they are paid obviously.
but you can still get some of them on cgpeers
edit mode
Im not that anon but where can I find Softimage its no longer on autodesk
File: 1516150163529.png (142 KB, 430x490)
142 KB
142 KB PNG
You're lucky, actually I'm THAT anon, you have 3 ways.

-Rutor (torrent tracker, just search softimage 2015, use chrome for translate, it's on russian)
-or go into pages such as ebay or others that sell software keys, it's very unlikely that you can find one but thats all
any suggestions on creating cloth/blankets in blender that look like theres a weight being affected by it like picture?

maybe a cloth sim over a box while having a negative force field pulling towards the center and a turbulence force to create some folds...

any suggestions would be appreciated.
File: 1505530351850.jpg (1018 KB, 1920x1480)
1018 KB
1018 KB JPG
hi /ic/ retard here.

What are the different sub domain of 3d? I know some people specialize in animation, others in modeling, is that it?

I think I want to start learning 3d, there's more job than 2d and it seems fun.
use fractal factor before using the sim
File: sshot-123.jpg (15 KB, 650x317)
15 KB
can i make similar effects with blender ?
do i do this through camera settings or anything else ?
blender cycles btw
yeah go to color management and then select the "film" setting and then in the contrast presets you can get a list of all those hipster camera effects
File: 1453197608059.jpg (84 KB, 507x547)
84 KB
Hey so I need help getting a cape to move like this In blender making the cape flow in waves at the end having it spread wide like a cone and have the the same kinda height.


I've tried increasing the wind it only gets the height right, then I tried keying the wind strength to try and make more of a gust effect, I tried turbulence and even a force field but nothing seems to achieving this effect properly
bumping my question. Come on it's an easy one
Can anyone point me to a really good book on topology for animation?

I’d rather not just fumble with it until I figure something out. There’s gotta be a holy grail somewhere
Sculpting, modeling, rigging, animating, rendering, texturing
File: favicon.png (1 KB, 32x32)
1 KB
thanks anon. I thought sculpting and modeling were the same thing. And I thought rendering was automatic? I'll google those out
File: Screenshot_5.png (108 KB, 882x286)
108 KB
108 KB PNG
How would I make the handle of the back of a simple, rounded wooden shield?

This is what I have so far. I think I need to bend the middle or something so it ramps up from the left, then goes down from the right.

Im not too sure how to do that though, any ideas?
Sculpting is without regard to topology modeling is the latter.
you can take it into the composer and use the color-gradient node to change the colors of the image
there are several fields. generalist, programmer for shaders or for animation, animator for characters, grafx, simulations of physics, mechanical objects, etc, also architecture visualisation, also game asset creators, crime scene reconstructors, model builders for science or medicine, sculptors for toys, specialists for film and cinema productions.
and many many more

Anybody know the tool to accomplish this in Maya?
How should a low-poly game-ready model of this look like? I made a basic blocky shape of the main body of the bus, but now I don't know if I should start adding loops everywhere and put a subdiv modifier on top of it, or I should maybe chamfer the sides. The problem with chamfering is that if I chamfer all the sides on the top, the topology on the corners will get destroyed, but if I use subdiv, it will probably be too high-poly for the game environment. I have no clue, never modeled anything like this before. Please help.
File: busworkflow.jpg (47 KB, 1134x652)
47 KB
So, left chamfer, right subdiv. Both pretty much get fucked up though, so I bet I'm doing both workflows wrong.
look into retopology and baking
I know about both processes, but that isn't the answer to my question.
my guess is that the guys who made this bus use subdivision modeling.

iv seen way more complex models made in subd, its just a matter of experience. you need to start with a basic blockout and uniform topology, don't chamfer shit.
How would you do something like a movable (open-able) door with baked lightning?

No matter how I think I would do it, It would always end up with a visible "seam" between room A and room B lightning, or with bleeding through the door...
those are carriers. But thanks
Is there any real advantage to getting a 2tb hard drive over the 1tb hard drive for modeling?
hmm it depends
sculpting projects can reach up to 100mb per file
modeling projects are usually somewhere between 10-20mb, large scenes can reach 100mb
I asked this elsewhere but it's worth putting it here too:

should i be sculpting in perspective mode?

anything I do in orthographic mode ends up looking grotesquely malformed in perspective mode - what focal lens should I be using to sculpt in?

I'm not sculpting with any goal in mind (an animation/short film etc.) so sculpting to a hypothetical project's 'lens' is out of the question - most films use different lenses and perspectives anyway, so what would be the point in using just one?

I'm fairly confused so if anyone could explain their sculpting process with the above in mind I'd be very grateful

In short terms snap among each one while you're working, it's the proper way, like when you paint in something like Photoshop and you change sides to see inconsistencies.

Also if you're not sculpting for anything but (i suppose) just rendering, remember to experiment with cameras and angles before doing the designated render.
Not sure if this is the right thread but i'll ask anyways.

I'm in the early stages of learning how to sculpt and i'm wondering if i can use a ripped 3D model as a reference for sculpting that exact 3D model but from scratch. By loading up a mesh and it's normal map, is it possible to use that as a reference to sculpt?
Usually you do want to sculpt in a mode that offers a somewhat compressed field of view, since perspective can also trick you in terms of scale, making you, say, make the nose on a face too small. If using perspective you should sculpt at around 24-28° FoV which would be equal to a standard 85-100mm portrait lens.

I don't see why not, as long as you can keep the model up on a second monitor, as you would with any refrence.

For modeling, storage isn't an issue, but where textures are concerned, things can get out of hand if you want to have 4K or even bigger.
Got a weird issue in Zbrush, it started a while ago but it's only now starting to really fucking annoy me...

Whenever I hit Dynamesh, or ProjectAll or some similar calculation intensive process - Zbrush does this really strange thing where it momentarily boots me out of the application, as though I'd just hit alt+tab. It completes the task without issue, but I have to actually switch back into Zbrush after it gets done.
It gets weirder; I had task manager open when I did this once and noticed that Zbrush actually totally disappears for a moment when it does this - it's like Zbrush is actually crashing/quitting for a second - then comes back to life.

This isn't a life-changing problem or anything I know - but it is WAY annoying when I'm Dynameshing a sculpt maybe 1 or 2 times every minute.
I googled but didn't find anything. Anyone run into this one? Any ideas?
Are you using cracked version?, Specifically 4r8 p2?, If so download from cgpeers the one that says "with Keyshot bridge" that's the one you want, i had similar problems by using other crack.

If no it's just zbrush itself maybe you touched something and should re install or just restore config.

Now if that doesn't work, something tells me that all this has something to do with memory/ram/cpu usage since you say that task manager pops up, so go to preferences and calculate memory and do the optimize Threads thing.

Otherwise goodluck.
I'm supposed to interview for a job doing simple CAD drawings tomorrow. Most of it looks pretty basic but I'm nervous because I've never really used CAD before seriously, I know the basics more or less but are there any particularly good youtube tutorials / websites to check out?
Not sure if this is the right thread but i'll ask anyways.

I'm in the early stages of learning how to sculpt and i'm wondering if i can use a ripped 3D model as a reference for sculpting that exact 3D model but from scratch. By loading up a mesh and it's normal map, is it possible to use that as a reference to sculpt?

You already asked this and >>602356 answered, if you don't know how to do it, you should tell us what program you use so we can give you an answer.
Completely honest question regarded towards modelers and texture friendos. How did you get over the whole, "There's more than one way to do it, and no 100% right way" Kind of thing for each of your fields?

It's the one thing I think is holding me back right now, I'm so used to there only being one way to do something, so it's taking me a while to open up to the idea that I can just make what ever I want to make my way, or even just in a way I know how. This isn't referencing efficiency. But the idea behind it.
Actually you shouldn't *open up" if you're already used to something that you like, you should open if the way you're doing things doesn't feel alright or is not at your proper speed, not if you see other people doing it.

Think about this like if we're all in an MMO with different classes and skills.

Also remember is what YOU find right, do you light good topology or fast results?

Would you rather paint red your draw with caution, slowly, trying to stay inside it or throw a fucking bucket of paint and then clean what's outside the draw?
Step 1: figure what field you really want to focus on (game assets, high-detail, archvis, etc)
Step 2: narrow down a workflow for that asset type that lets you make a finished object with the least amount of hassle
Step 3: immediately get your feet wet and start making stuff. If you lack your own ideas or concept, copy an existing one.
Step 4: keep doing it until you're really fast at it. Once you're a modeling ninja, try for a job, speed and efficiency is usually valued above all.

There's still going to be a set of modeling types that you'll have to know, for instance, I use mainly edge and spline extrusion with booleans for hard surfaces. But some shapes are just easier to create when you know the right modeling type, and pick the right method when the need arises, so I'm not entirely against using subdiv surfaces if I need to. What I'm saying is, it pays to know several methods even if you focus on just a few. I hate baking normal maps with a passion, so I usually do everything I can to not use them other than for details.
I've got a problem where X-mirroring constantly breaks in Blender any time I touch the armature in pose mode. Anybody know if there's an easy way to unfuck Blender or if a different modeling program will let me actually let me test out the deformation on a vertex group without having to constantly revert my save file after every test?
I'm on a tight budget and interested to get into 3d.

My current build is 2500K and R9 380X, and I'm uncomfortable with the sculpting and rendering performance I am getting.
With current GPU and RAM prices bullshit I could only afford a high end gpu or a ruzen platform, not both.
What will improve my 3d workload performance more? A Ryzen R7 1700 of a GPU like 1070ti/Vega 56?

AFAIK a GPU does rendering better than a Ryzen CPU by like 30-50% but what about viewport and sculpting? Are they GPU or CPU limited tasks?

Am I going to be better off If I replace the platform but keep the old mediocre GPU or should I stick a new GPU into my old platform?

Enjoy an anime lady I just denoised and tweaked from a shit quality scan for your attention.
Man, I try fuckin EVERYTHING to get results. I mean *everything*
And when I say I 'try everything' I don't just mean testing out bits of software (which I do, obviously, but that's not what I'm getting at right now). I actually studied graphic design at a particularly 'concepty' out-there kinda college and it gave me an attitude towards the computer that there is always a way you can hack something or mis-use something to arrive at a unique result - and I carry this forward into 3D; sometimes it works, sometimes it doesn't.

Case in point - driving home last night I was struck by an idea that I've spent all day trying to get working and I've only just now resolved that it ain't gonna fly, (which is why I'm here).
I've used Vue Xstream a few times and been really impressed by the level of detail and quality you can get really easily with the procedural tools - but trying to export anything just doesn't work - all those procedural materials and samplers and noise generators work within the Vue environment and aren't intended to be portable.
SO - i realised last night; if I set up a Vue scene; (just a few rocks and things), if I rendered them from all angles with an animated camera - I could then push all those renders into a Photogrammetry processing tool and voila - I'd have my beautiful Vue procedurally generated textures and geometry all there in a high-res photogram.

It worked - yup. But in the end it wasn't worth the render time.

My point is - you've got standardised tools, yes - we all have them. But its down to you as a creative person to exploit them and mis-use them and use them in new ways. THAT is what makes shit interesting, and what makes you attractive to employers because it proves you've got a brain and enthusiasm and a sense of resourcefulness.
is there a general rule for intersecting topology? like if I work on a high poly model with lots of details, I know I can get away with multiple meshes cramped into one 'object'
but if I try to retopologize the whole thing, is it better to create one mesh, where every detail is connected and ignore a higher polycount or could I get away with 'lots' of lowpoly-objects forming one big mesh?
File: Iroha-zaka2.jpg (36 KB, 390x260)
36 KB
Whats the proper workflow to model and texture a road like pic related. My biggest concern is how to set up UV maps for those white stripes as they are getting more wider as they go. Sure I could do it by hand, but I am wondering if theres any better ways to get this done and smoothly. I am using blender btw
Last time I did a road, I used a bunch of UV channels:
Channel 1: A simple planar, square tiling map. Just a top-down, square planar map. This is for a repeating texture, a good quality tilable asphalt texture.
Channel 2: A carefully laid-up map of the entire object. Again taken from a top-down projection, but this time keeping the actual shape of the road. This is for paint-work and details in the road like your lines etc, and also bits of dirt etc.
Channel 3: Similar to 2, but this time; straighten out all the islands so they are perfect rectangular shapes - even on the bends. This will obviously cause stretching - correct. But that's the point. In this channel, you can put textures with perfectly straight lines like tire-tracks and other marks that follow the curve of the road. On this straightened out map those straight textures will bend to follow the curve of the road.

You're welcome.
I'm assuming you're talking Zbrush right?
Keep your HP as one mesh for as long as you can, if your machine will allow it. Save that out as your bake source.
When it comes time to re-top:
1. make a clone of the HP, then decimate, then Zremesh to get your topology nice.
2. Go around the mesh logically working out which piece need to be kept together, and with your mask brush paint up these pieces and go 'Group masked' keep doing this till your LP Zremesh is divided up logically into groups.
3. Export this low-poly into 3DS Max - the .obj importer will show that the mesh is going to come in as a bunch of separate meshes (these are all your groups). This is what you want.
4. Lay out the UVs for each of these meshes and assign a unique material to each one as well. At this point it's a good idea to make sure everything is named up nicely.
5. Select all these individual meshes and go 'export selected' and send them all out as one FBX.
6. In Marmoset, set up bakers for each chunk of the LP group, bring in your HP and bake it down to onto each LP chunk, one piece at a time.

You, also, are welcome.
i think your polygons are not welded (yet) while in mirror mode
File: hurtubise.png (535 KB, 1182x660)
535 KB
535 KB PNG
thanks for the detailed answer, kind anon, I'm relatively new to this and really appreciate your help!
is there any downside of having multiple meshes as fbx? As a beginner pic related looked kinda awkward, having the meshes just 'colide' rather than stichting them all up to be one
but I guess that's irrelevant if you can save some polys?
>is there any downside of having multiple meshes as fbx?
Nope. It's what FBX is intended for - FBX is like a container format - animation, materials, textures, and multiple meshes all contained neatly in one file.

>>As a beginner pic related looked kinda awkward
It's fine, but it depends on if your camera gonna be really super-close to that object? I mean, like within 18 inches, staring straight at it, as the main point of focus? Because if so - you'd probably want it to look a little nicer. If it's just a side-prop, what you have right there is fine.
Secondly - what you're describing as 'meshes' colliding with each other are just different parts of the same object. Separate meshes are literally unique, individually selectable objects in a scene. What you have there are just chunks of polygons in the same mesh that don't happen to all be welded to each other in some way. This is fine. An object can have individual elements.

Relax - you're overthinking things.
yeah, that makes sense.. kinda nice actually, that you're not forced to weld everything up into one tight clump of connected polygons
anyway, thanks dude!
oh big thanks, i ll look into how to do this exactly now
For sculpting, a nice CPU is all that really matters if you're using ZBrush, as it doesn't utilize the GPU at all. Rendering is a different problem, because the most efficient price/performance comes from GPU rendering, and even a decidedly high-end CPU won't be as satisfactory for fast iteration, but it's not like a 1700 is that bad.
A CPU upgrade is probably the most universally useful as it'll benefit texturing, baking, and any kind of physics or particle sims. A fast GPU is more useful if you're doing a lot of rendering or working with game engines.
Thanks I think about working with Blender for now, but I guess it's gotta be similar. What about drawing high poly wireframe and wiewport? Will a 1700 pull it with my old shitty GPU?

Let's say I want to make a simple-ish 3d product showcase video for a small business.
Is my best course of action to get a 1700 to model it on my side and then render it in amazon cloud?

How much would amazon cloud cost to rent?
I really appreciate you guys getting back with me on this. Actually makes me happy to see that going outside the box is a normal thing in this field. Again not something I'm used to in my life, as everything has always been exactly by the books.

Time go to fiddle with shit, thanks again.
>I think about working with Blender for now
As far as I know Blender uses polygons for modeling, so you do actually want a fast GPU there, same with Mudbox and maybe 3D Coat, although I heard it uses voxels. If you're hardware limited, ZBrush is by far the most workable option because of the way it's designed. If you're using Blender, you'll have to be more aggressive with breaking your model up into chunks.

>What about drawing high poly wireframe and wiewport? Will a 1700 pull it with my old shitty GPU?
Viewport rendering doesn't require much, my computer has a really old card and it's perfectly fine for manipulating things in the viewport. as long as you don't run out of VRAM, performance will be acceptable regardless of what GPU you have unless it's really fucking bad.

>Let's say I want to make a simple-ish 3d product showcase video for a small business.
>Is my best course of action to get a 1700 to model it on my side and then render it in amazon cloud?
Can't comment on Amazon as I haven't used or looked in to it. For the product showcase it depends on how high-quality you want the render to be and how fast you need the result. The 380x might still be decent with a GPU renderer fwiw. You should also try out Keyshot, which is a fairly popular program for product showcases.
Is there a place where I can find a 3D female model like pic related with bones so I can pose her in blender?
Tf3dm has lots of them
File: 1505859442957.jpg (89 KB, 792x819)
89 KB
Question here:

What yields faster renders:
4 PCs like pic related, Xeon processors with 22 cores and a Quadro GPU.
or 2 PCs with Dual Xeon processors for a total of 44 Cores.

Im about to start a big project and was asked to research what would be the best for the render pcs, with a budget of about $5500~6000 for each computer, what could be the best PC to build?

We are going to work on After Effects, 3ds Max, Cinema4D and going to do some Liquid Simulations.
File: file.png (187 KB, 1920x1080)
187 KB
187 KB PNG
Does anyone know what's going on with my 3ds Max UI? I am at a loss. The buttons are super large and the command panel covers one third of the screen. Many of the other panels are completely unusable. I don't want to risk fucking it up even further by trying to fix something I don't understand.

But I want to understand. I want to fix it. I want to start modelling again.

Is it a Windows error? A holy sign of God? Am I retarded? Or should I just give up now and learn Modo or something instead?
Pretty big project to be asking for advice on /3/.
So here's the thing, if you'll be using a GPU renderer, then it almost doesn't matter what platform it's based on as long as it has a mobo available with 4x 16-lane PCI-E slots to max out your GPU count, and cram them to the brim with 1080 Ti's if you can actually find any, let alone for a reasonable price.
If you'll be using a CPU renderer, then you need to optimize the amount of efficiency you get out of each chip, and the TR 1950x pretty much rips everything else up in that aspect. There is literally no reason to buy Xeons for rendering when TR and Epyc exist. The 1950x scores 3000CB, the 32-core Epic is 5600CB, and you can have a dual-Epic system for like 11k rendering power in a single box. The dual Xeons are rated for about 5800CB by comparison, so a single 7551P will do as well at half the cost, and you don't have to rely on an older platform, that's also less secure due to the recent bug revelation. V4 is Haswell IIRC, and I think the patch hurts its performance, so it'll be even less good than what you'll see on existing benchmarks.
Also you don't need quadros, they don't bring anything to the table unless you're using CAD software that requires some very specific optimizations, otherwise almost any GPU is fine.
With that said, Max and C4D, as do most modeling packages, prefer faster CPUs over more cores, as does AE unless Adobe finally got around to rewriting their threading engine to actually use more than like 8-10 cores. If it's the latest version of AE then do be aware that it's essentially broken and won't even use the second CPU.
And if you're doing liquid sims, if it's in Houdini, then yes it will take good advantage of more CPU cores, however it also chews through RAM like it doesn't exist depending on the scale and detail level of what you'll be attempting, so get like 128GB just to be safe, and on top of that upgrade to an NVME SSD like the 960 Pro, because if the memory overflows performance will tank like a motherfucker. It may even be reasonable to suggest getting an Optane 900p as it's been shown to drastically reduce simulation times when memory is at a premium, and by a not insignificant margin - you could be looking at a 4x reduction in sim time vs a basic SATA SSD if your memory caps.
Is there any downside on using AMD CPU over INTEL? Ive heard that AMD comes with issues in terms of compatibility over software and plugins.
The only discrepancy is performance in poorly-threaded programs, where Intel does enjoy a decent lead, however, since we're talking about CPUs on the level of server-grade hardware, this isn't even a consideration because Intel server chips are aggressively voltage regulated for stability and give up that advantage. On the contrary, as long as the cooling on a Ryzen chip is sufficient, it can ramp all of its cores to the rated turbo clock very easily, which is why it's so much faster than Xeons.
Oh ok then it will work perfectly as long as everything is tidied up.

I went ahead and took up your suggestions and made another list following my bosses budget. The thing about my boss is that he really believes that the most expensive every card or processor is the better, so im always trying to aim for the most cost effective builds without going overkill with those ultra expensive Quadro or Xeon processors.

So far pic related is what i got. The idea is to get 3 of these pcs linked up into a network and have them go on Render Farm duty, doing renders while we work on other PC. The Project is somewhat big but still it isn't so extremely focused on getting things like hyper realistic lightning or things like that. Its more about getting huge motion graphic renders for a mapping projection we are going to do on a football stadium.

My boss insists on pumping more money into super expensive processors so thats why im asking around if thats even worth it if we could just get several powerful pcs for the renders and it will be more than enough to get everything done in time.

Funny enough, when i mentioned my boss that the GPUs were getting really expensive due to the bitcoin mining thing he went ahead and asked if we could do some mining too while the PCs werent being used to get some of the investment back, which i found hilarious but very plausible.
File: 1499989299879.png (62 KB, 763x828)
62 KB
Oops forgot my picture
Ah, I know bosses like that, but still, you have to be mindful of what actually improves performance, even if price is no object. Funny as it is, a plain 8700k is the best CPU for "realtime" tasks like modeling, animating and other things with no obvious use for massive amounts of cores. By a lot even, like 25% faster than the 1950X would be for hands-on usage scenarios, it's only the hands-off stuff where the 1950x will fly.

A few adjustments, get the Noctua U14S TR4, not only will it cool better than a liquid cooler because of the custom cold plate, but I also have a thing against putting water into 24/7 systems. Ryzen isn't a hot architecture even when you have 16+ cores crammed in, so it doesn't even need water, but if you're insistent, the Enermax 360 TR4 is the only option because it's custom made for it.
The Asus Zenith is also a halo product and probably not the best pick, esp if you read the reviews on Newegg, apparently it has very poor stability. The Asrock Taichi is a better pick IMO, or perhaps the Fatality pro gaming as it has integrated 10Gb Ethernet.
For memory, absolutely do not get less than 2666 speed, because Ryzen in particular is affected by it, and also memory intensive tasks in general. 2666 is the fastest native speed of DDR4 and that's what you want. I'd suggest more had I an idea of how reliable it is to push even 3000Mhz over 64Gb of RAM. 3200 is technically the sweet spot to aim for, should it be feasible.
Finally, a minor point, but while the 750D is a nice case, get the variant with the mesh front as it greatly aids ventilation. Or any case with a mesh front for that matter.
For the PSU, stick with EVGA even for the high wattage units, they're just great and not overly expensive. Seasonic is the Rolls of PSUs though, if you want something high-end in that aspect.
One extra thing, if you go to pugetsystems.com, they have articles on the site detailing the performance of high end hardware like this in various 3D software and render engines, it might be worth your while to check it out.
Thanks a lot for the recommendations, really helped me a lot, will get a final list going and show it to the boss, now it's a matter of getting into Amazon to get the buylist going. Hopefully theres GPUs still lying around with all the Bitcoin mining fest.
Yeah. I just realized you have two 1080 Ti's listed here >>602554, but I didn't quite get whether you're decided on CPU or GPU rendering, you really don't need a second card unless you intend to render stuff, or perform any other special type of processing that can utilize multiple GPUs.
If CPU rendering, the render boxes don't even need cards that fast, just give them something basic for video out. If GPU rendering, you may consider stepping down to the 1920X or even 1900X as the CPU doesn't do that much, but keep the 4x16 PCI feature of TR. speaking of which, 1000w is notreally enough if you ever intend to have four 1080 class cards in a system, I'd get 1200w just to be safe. It's not even about whether or not it can handle the power draw, it's just that for extended use you generally don't want your PSU supplying close to its limit.
Its going to be a mix of CPU and GPU rendering that's why i got the 2 cards going so it can handle both as the situation needs.
I got a larger psu too in the final list too on the final list, i also read some of the articles and oh boy, money can surely still buy a lot of more performance but the cost jumps a HUGE amount. The Dual $3k Xeon's can provide an insane 180% boost in Cinema4D rendering but it would cost a fuckton of money to get those.
Better to have 2 ThreadRipper PCs than to have a single Dual Xeon PC imo. The cost is crazy.
File: 0078-0765.webmhd.webm (990 KB, 1280x720)
990 KB
Anyone get a weird problem with Blender where new objects won't render?
Everything else that was there previously renders fine, but all new objects wont render. They're not hidden or anything. I've been using Blender for years and I've never had something like this happen before. Saving and reloading doesn't work either.

Here's a webm of what's going on.
There's no compositor stuff going on, so it's not that.
It's a brand new object that's placed and it doesn't show up at all on the render, but shows up fine in the render preview.

I don't know what's going on.
(posted this in the wrong thread, because I wasn't paying attention)
Give the new object a material
Nevermind, I'm a fucking idiot.
Looks like I had one of the render layers excluded by accident.
Don't really know how that happened.
File: DSC04725.jpg (1.38 MB, 2500x2500)
1.38 MB
1.38 MB JPG
Some good advice there, the X399 stuff makes for a great workstation platform. We just built a couple new Threadripper render nodes at work, pic is my TR build (MSI Pro Carbon board, 1950x, 32GB 3200MHz HyperX, 2x GTX 1070, EVGA 1200W P2, 750D Airflow).
I don't think i've posted in this thread before. But anyways i'm using zbrush but i'm still at the beginning stages of learning how to sculpt and currently learning human anatomy.
Why do you need to press F in order to pan around with Spacebar in Photoshop?
why can't you do that in a window?
Damnnnn... was it worth the grand for the 1950x? Also, why did you chose the x399? If you got 2 1070’s and an amazing processor , why cut back on the motherboard ?
No regrets about the processor. A lot of stuff in Houdini is multi-threaded, so I've been very happy with it. As far as the motherboard, the feature set between the different mobos is almost identical, save for a few things like 10GB ethernet. I'm not a huge fan of MSI, but I don't really have any complaints about the motherboard itself (now that the BIOS are a little more mature). The one thing I skimped out on is memory. I'm regretting only having 32GB and prices are even worse now. The kit I purchased was $340, that same kit is now $506.
How did your interview go?
Ahh I see.. my gaming ass didn’t even think about Houdini.
My first board was msi and I switched to asus for my newest build .. I like it a lot better for a myriad of reasons.
I hear ya on memory, did you get 2 sticks of 16? .. when ddr5 ram comes out the prices will go down and you could always add on. Though I hear doing it separately can affect the performance of the ram-like if you buy separate packages they don’t sync as well or something
File: 1508825507199.jpg (114 KB, 968x706)
114 KB
114 KB JPG
Serious question: How do I keep eyes from looking bad like in the right? It has become an irrational fear of mine.
1. You have just tried resetting the UI, right? It's somewhere under the 'customize' menu.
2. You may have a preference enabled for 'enable large buttons' somewhere. Try toggling that on and off.
3. You aren't going to fuck anything up by altering the UI, don't be daft. If you're really that worried, you can save the current UI layout if you want - Customize>Customize UI, chose save.
4. If all else fails, delete your prefs folder (again, don't worry, this is not gonna wreck anything - but if it makes you feel better, just move your old one to the desktop or something instead of deleting it). When Max restarts, it creates a whole new set of preferences with default values.

Close 3ds max
Navigate to the following path (Windows7/8).
C:\Users\%username%\AppData\Local\Autodesk\3dsMax\2015 – 64bit\ENU
Rename the ENU folder to ENUOld.
Launch the 3ds max

5. >>Or should I just give up now and learn Modo or something instead?
Do what you want. If you're ok trying to learn modelling with an application that has probably just 5% of the support community, learning resources and 3rd party developers as Max - then go for it. You do have a legal licence for Modo right? Because if not.... well, good luck getting a job.
Bake the lighting properly.
I have the same fear. Ive made heads that look better than the one on the right but I cant keep these images from gnawing at the back of my mind.
Not him, but...
>I hear ya on memory, did you get 2 sticks of 16?
As I've said in my extend posts earlier, Ryzen is pretty sensitive to RAM, and since TR4 is quad-channel, you really really want four sticks of RAM in, regardless of how much memory it is, or else you could lose up to a monstrous 40% performance in anything that is memory-dependent.
Because of this, think of the amount of RAM you need now (across four sticks), and know that in terms of adding extra sticks, you can only add twice that much.

>Though I hear doing it separately can affect the performance of the ram-like if you buy separate packages they don’t sync as well or something
With mixed RAM, you get shoved to the lowest denominator in your system, so if you have say two sticks of 3000 CAS15 and and two sticks of 2666 CAS16, the faster sticks will be hard locked to the specs of the slower ones, so for all intents and purposes you're just wasting the extra money you spent on them not being able to run at their rated speed. Otherwise, all RAM is supposed to be compatible with the baseline set forth by the organization that handles the DDR standard, so you're never SOL. Do note that all the speed ratings on RAM sticks refer to the XMP profile stored in their rom, if you actually want them to run that fast remember to engage XMP in the BIOS or else they'll still be running at base DDR4 speeds like 2400 or something.

Regarding motherboard choice, there's nothing really "low-end" about any of the TR4 boards, they have to be top end by design just to support all the features it can pull off. Even the simplest TR4 board will impress you more than anything on a consumer platform.
Been learning SketchUp for a while and have gotten pretty decent at it. Is this feasible as a main modelling program or just a stepping stone to a "better" one? Personally I prefer SketchUp to other programs, but there's quite a few limitations so I'm wondering.
Try Fusion360 next. You'll find it very similar.
1. Yes, first thing I did.
2. Well, the "Use Large Toolbar Buttons" option is turned on right now, but turning it off doesn't change anything other than the size of the toolbar buttons.
3. I use the default UI layout, so that's not the problem.
The thing is, I've already tried a bunch of things that I thought could help - changing Windows scaling settings, overriding high DPI scaling behaviour, deleting configuration files... I've even uninstalled every Autodesk program and reinstalled them, and it had no effect.
I'm starting to think this is a system issue, and I'm afraid that I have to delve into regedit or something to fix it.
4. I tried this too. Unfortunately, this doesn't help either.
5. uhh..

Thanks for the suggestions, though. I really appreciate that you took the time to try and help me.
Do you have a lot of plugin dependencies? I'd suggest trying out Max 2018 as it uses an updated interface, maybe it'll render correctly.
No, I don't have any plugins. It's pirated though, if that wasn't obvious.

I'll see if I can find the 2018 version. Thanks again.

I just noticed that i did posted the exact same question earlier, i realized i tried posting a few days ago an i was constantly getting a "Connection error" and my post wasn't showing up :0

Either ways thank you!
File: file.png (182 KB, 1920x1080)
182 KB
182 KB PNG
I just installed the 2018 version, and I still have the same problem. Am I cursed? What do I do senpais
Huh, weird that they are that big. I don't have the answer but when you mentioned it, I have to also say that Max 2018's UI is fucking horrible. In 2017 you could have pulled the right border to the left and get 2 fixed rows. Now, it's "responsive", but the buttons get stretched in weird ways and sometimes you have to pull even more to get 2 rows displayed, which means it wastes even more space now for no reason. Terrible, just terrible.
>What do I do senpais
git blender
What's your dpi scaling set to? I think it's a Windows scaling issue and not to do with Max at all.
Why would it happen only to a single program, though?
File: file.png (204 KB, 1920x1080)
204 KB
204 KB PNG
I agree.
It's set to 100%.
It looks more or less the same at 125% except now I feel like a toddler who needs big buttons and big text everywhere.

I have one other scaling issue, though. I don't know if it's related, but the text in installation programs is so ridiculously tiny that it's unreadable.
in blender, i need to convert like 50 actions into 1 action.

i should also mention this animation has no keyframes, but after adding 2 keyframes (one in the start and one in the end) i can't copy paste those keyframes into a new actions. it tells me there is no f-curve to paste into
File: modified butt size.png (192 KB, 1071x611)
192 KB
192 KB PNG
Greeting, i am new to blender and 3d modeling in general.
My question is about how to make that bikini bottom fit the butt of the model, the model itself is a modified model with a bigger butt than the original.
nvm fixed.
i just figured i can bake nla strips
snapping, shrinkwrap etc
File: blender problem.jpg (325 KB, 1305x914)
325 KB
325 KB JPG
really noob question here, I usually work with just a mesh and then export to another program to add bones etc, but after using blender its hard to go back, how do I make it so I only see the mesh ? I have tried the eye symbol in the top right panel to no avail. I vaguely know about visibility layers / groups on the lower panel but have never used them
click on the plus symbol in the armature and then hide the armature
armature tab on properties window has its own set of visibility layers for your armature, just click to a different layer and you'll only see mesh.
I got it now thanks, I selected mesh then pressed M to moved it to a new layer and as you said it did not show the other things

Now I have to figure out how to seperate each part of the clothes etc because it imported it as one object but with every vertex group

That makes me think, when I make a new model should I define vertex groups for each area of the body ?
bumping for help
should I just rig it and manually do the cape wave
File: Cape.png (65 KB, 849x600)
65 KB
Pin the outer edge of the cape. That'll give it the stiffness it needs to stay upright, no matter how stupid the shape is. I used a 0.4 weight in this image.
If you want literally only the mesh, then you can turn on "only render" it'll get rid of everything that isn't the mesh though, even the grid lines and cursor.
If you hit L over your object it'll select an entire section based off this setting.
Alternatively, if your model came with vertex groups you can use the vertex groups subtab in the vertex data tab to select them based on that.
Using maya.
Is there a way to copy skin weights from a SINGLE joint, from one mesh to another? Same exact meshes, but different weights. I want to essentially "combine" the two and keep certain weights from each.
I know how to copy all the weights, but doing so overrides everything. Even if I lock the joints I don't want to change it still decides to override the original weights. I just want a single joint from another mesh copied over to my original mesh, without ruining the rest of the weighting.
Question, When Animating in blender. Is there any tips to prevent clipping, Specially when I somes to fingers interacting with objects.
I could rattle off a laundry list of all the dumb little failings: bizarre makeup, unappealing face shape, startled expression, bad reflections on the eyes, no shadows from the lids, no sense of depth on the eyes, constricted pupils, no visible corona ciliaris (the little wiggly ridges on your iris) making her iris look very flat, clammy skin, manface + mancut, ugly LOD hair on a close up shot, and so on that's not going to alleviate your fear.
You can't just assemble a list of "do not"s, you need to improve your core knowledge / understanding.

So, as I said, your (very sensible) fear is borne of a lack of understanding. 3D is about simulation and approximation, the "best" and most realistic (if realistic is what your piece calls for, obviously) simulation is going to be the heaviest and least optimal.
Study the human eye, the parts that make it up so you can understand why the physical material behaves the way it does when light hits it.
Once you understand that, you can then set about the task of optimizing your simulation (into an approximation) for game engine (or animation) purposes. The one on the right looks like shit because, besides just the whole thing being an abomination, the shortcuts taken for optimization purposes failed in small subtle ways. It's that accumulation of little errors that makes the viewer think "this looks like shit" even if they can't point out what those errors are (and this is true for all 3D).
Incidentally, besides being more technically excellent, the one on the left is just stylized enough that you mentally give it more leeway.

You should feel good about that fear, when you know you're doing bad work but you do it anyway because you have to (for a deadline, to keep your job, to appeal to your boss's shit taste), you've sold out and you haven't become complacent. Complacent is similar except it's where you stop listening to the little voice in your head asking "Can I do this better?" because you "made it".

for some reason when i use knife tool when i click on one edge and try to connect to other egde the second edge is like invisible to the snap thing knife has..
You can make a copy of the two meshes (duplicate-special and keep weights), then delete the faces you dont want. Then you ctrl-click those two meshes and then your target mesh and transfer the weights.

You can't really choose a single bone to transfer because the weights might not add to 1.
you can select 2 edges and click J, it will connct
meant to say verticies*
File: render8.png (519 KB, 1291x915)
519 KB
519 KB PNG
I'm a fucking idiot who decided to work on something complicated for his first project. What are the techniques that I need to know so that I may have an idea on how to incorporate the feathery ornament to the helmet?
for stylized look sculpting a piece of mesh with dynatopo would be acceptable

but for more realistic look with real hairs you should try the particle emitter. plus, you have a special mode in blender that lets you control the shape of the hair
and your final (and best, in my opinion) is to make hair with special curves.

at first, it will require you to do some tinkering but the good thing is you can save all of your hair types and use them later

Thanks my dude, I'll look into it once I get back to work tonight.

You can make an extra rig that helps with positioning, and just use that to prevent clipping.
Related video shows an example where you don't have to manually mess with every bone in order to position things.
Just look up IK control rigs or something along those lines.
File: 84b84fea56[1].jpg (202 KB, 1096x660)
202 KB
202 KB JPG
What the FUCK is happening to my leg?!
That's a common problem. It's caused by the weighting on the upper portion of the ass being too heavy on the hip/root bone for llateral bending like that to look right. The easiest fix is to use corrective blendshapes.
Using Zbrush, how do I sculpt a character with arms crossing :
Sculpt in T pose or A pose and then use Zsphere rigging to pose it with arms crossed or try to sculpt it directly with arms posed (i find it tedious and way too dificult)
if the character wont gonna move, and her final pose is crossed hands, sculpt in crossed position,
because whun you'll posevher (which is kinda hassle already) you'll have to resculpt arms and clothes slightly.
Use corrective bones, partition the weight between 3 bones, instead of just two :
1) the pelvis
2) the upper leg bone
3) a third auxiliary bone that rotates 50% of the upper leg bone
To weight things properly first distribute the weight between the pelvis and the leg then distribute the weight between the pelvis + leg and 3rd bone
File: object in question.jpg (65 KB, 813x756)
65 KB
Hello /3/, I want to built a circular shaped building with different sections in Blender, but I don't know how to approach it properly.

I could just create it with the screw modifier but then I could not add in windows, doors and other details. When I tried to model 20° sections I quickly found out that editing the mesh is a nightmare because the transformation orientation is different for every section and not part of any grid due to the curved nature. Lastly I thought maybe the Simple Deform modifier would be a better approch, but at this point I don't get it to do exactly what I want.

Can any experienced gentleman tell me where it all went wrong?
Has anyone ever rendered a fender bender with blender render?
I can think of a couple of ways to do it. The easiest would be an Array followed by Bend 360deg modifier
File: example.png (478 KB, 797x843)
478 KB
478 KB PNG
I don't have the best pic that shows this but, overall it depends on what you want and your workflow.

pic related (I think) uses both dynamesh and posing/no dynamesh.
First off, the blue mark shows what the model will look like when dynamesh is applied in high resolution.
Then the red mark shows what the mesh will look like if pose/put meshes next to each other.
Thanks, by looking at your image and experimenting I found my error. I had to rotate my object correctly before the Simple Deform would actually work doing what I want. Still kind of weird, guess I'm just not used to this feature yet.
In blender, how do I connect two vertices without the edge appearing in the wireframe mode? I can use F but it will appear as a line.
why you don't want it to appear in wireframe mode?
Because of my autism.
I did managed to get rid of it.
Ok so I think I get it, I’ve seen some rigs with the double knee bone and I’m guessing that’s what you are saying. I think that’s how I’m gonna have to do this because of the anotonmy.

Quick question though, how do I know if I’m having a topology problem, a joint placement problem, or a skinning problem?

How do I know which one is the cause of my deformations looking like shit ?

Also do you have any reference on where to place these extra bones? I’ve seen the knee one but I’ve never seen a rig with the hip bones, do you have an example ?
I strongly suggest you start in T or A pose, there’s plenty of tools to set up the rigging such as using z spheres as a rig. You can then apply your pose to an animation layer and then pose your model and save that to a second layer.

This way you can work back and forth between your posed model and your standard pose.

Why this is benefits you:

1. Symmetry is usable from start to finish with your work
2. You can nail down proportions easier
3. You can work on those hard to reach places easier such as folding the arms or in the armpit area
4. Adding and working with clothing becomes much easier
I want to buy Zbrush and use it on my desktop that is not plugged into internet
I can plug it in for the activation, but it's normally plugged off. My question is:

Can I use Zbrush while not connected to the internet, or does it require users to be connected at all times?

I checked their site and I quote

>Each copy receives a unique serial number which is assigned to a single artist. ZBrush may be activated by internet, email or phone. Internet deactivation is also available from within the software, making it easy to move ZBrush between computers. In the case of hardware failure, a license repair system allows removal of the lost activation so that it can be reinstalled on a new computer.

kinda makes it sound like I would need to be connected to the internet at all times for the software to recognize I am working on the activated machine.

Can someone confirm if I would be able to use Zbrush offline?

ZBrush 4R8 single user license that is.
cracked 4r7 works offline, don't know about 4r8 tho
Yeah, 4R8 totally works offline, at least the cracked version on my Windows VM does.
well, a cracked version needs to work offline because you actually don't want it to connect to the internet.
thanks anyway
What program is this guy using? I don't recognize the UI.
It's Lightwave. People thought it was dead for a while, but it actually received an update recently.
>Lightwave isn't dead
Guess now we only get to make fun of XSI.
I want to start learning animation inside Maya. What tutorials and/or books can you advise me implying I have no prior knowelege or experience in animation. So basicly I want to know form where to learn principles of animation AND animating in Maya specifically. Is there a learning path that you went through?
File: unknown.png (9 KB, 648x610)
9 KB
So, I have a more or less question.
Lets I made a tentacle model. And I want to follow along a curve with different control/pull points.

does anyone have a tutorial that has what I want that I could follow.

here is my really crude mspaint to help someone understand.

(The tip bones will go down towards the middle point first, but the further I move my right to the left the tip point will go towards the far left point. )
Retopo and baking. From what I understand you make a real rough low poly version of your model, and baking it a certain way will preserve some amount of detail, including some complex curves and depth.
I haven’t gotten there myself, but a few of the class videos I’ve watched have done something along those lines, both within blender and in Substance.
there is usually something called a curve modifier but you have to use it carefully or it will wreck your model.
basically if your topology is clean you will have less problems
Hmm, I don't want it effecting the mesh. I'd like it to only affect the armature.
there are a few ways to do this

first of all im gonna ask you some questions

1)do you plan on it being a continuous animation where the armature can ONLY follow a path
2)do you plan to use it only as a guide so you can easily control the rig through a small number of points

so consider the following:

1) use an IK Rig first, this will give you more control over your rig
2) look into constraints, namely "follow path" constraint and "spline IK" constraint

here is the tutorial:

warning: it might be outdated so it best you look some terms using the searchbar

good luck!
File: unknown.png (37 KB, 836x546)
37 KB
So I was thinking along the lines more like, IF I were push the rig model forward, it was take the curve positions and I and depending where the rig is along the curve it would update the rig.

This way I have less to animate but while keeping a very nice curve flow.
i found this
there are also lots of good answers on stack exchange
Why in blender sometimes when I extrude some verticles just refuse to extrude, even if I select all with A and use remove duplicates?
Every animator should start with http://www.theanimatorssurvivalkit.com

You may want to go with https://www.youtube.com/watch?v=WHsKm9VaMK4
It's a 4 parter, so just watch the beginning up to 5:30 to see if that's what you want.

It's not a spline IK, but I believe that's a better tentacle/tail/hair rig.
ok, this may sound crazy, but it looks to me like it could be done with a rigged lattice.
You move a lattice forward, and deform the lattice using a rig
Then you animate the mesh going through the lattice
File: 03 (1).jpg (198 KB, 1280x905)
198 KB
198 KB JPG
2D fag here
What aspects should I focus on for 3D?

Pic not mine obv.
3dcoat, mari or substance painter?
tryin to assign automatic weights to some mouth bones. All the bones selected, assign the weights, the weights get made for the upper lip bones, but nothing for the lower lip bones. I cant figure it out.
npr - non-photorealistic rendering

Emphasis on materials, and not hand painting textures

A lot of post processing magic

Normal editing (and by that I mean editing Normals)

There is a lot. You didn't write what you want to do, so I assume you are one of those people who go to 3d to recreate 2d

people generally say 3d-coat is better for hand painted textures and SP for PBR

3d-Coat is also good for retopo
and I think it is the best sculpting software FOR LINUX.

If not Linux, then Zbrush of course.
File: anisotropy.webm (2.84 MB, 900x400)
2.84 MB
2.84 MB WEBM
Is it possible to achieve anisotropic highlights in Blender?
I've tried plugging similar textures into anistropic shaders, specifically their rotation inputs and couldn't achieve something similar.
Is there a way to transfer/copy the vertex of a model to another with similar shape and size? I am user blender.
Is there any good realtime renderer for Blender or compatible with Blender better than Eevee?

I tried rendering my scene on Eevee top settings and I am generally satisfied except I have high frequency details getting aliased even with x16 fullsuze sample AA.
Does AA get better than this?
Will super resolurion fix it?

I cant afford rendering that scene with Cycles ray tracer.
Didnt you forget the normals?
evee is relatively new, matter of fact its not even out yet
blender guru have a tutorial
>evee is relatively new, matter of fact its not even out yet

Yes I used the bleeding edge unstable build and it kinda worked after crashing on me a couple times.

It looked fucking good i was extremely surprised and I cant wait for the release because it's FUCKING GOOD.

Is there any better realtime Anti Aliasing I can use for free or should I just render it in 4k or larger and then downscale?
ue4 is the best real time renderer but it will take you ages to d/l it with all the updates.
you can also use marmoset but its paid.
there is also octane which is paid too but i would use it for real-time animation if that's what your'e looking for
So, I think is okay, I can't really get it work with multiple bones. I'll still be testing.

This sadly does look cool, but the end goal is to the tentacle to go in the same end goal every time, but have a different out start/mid.

this sounds interesting. But I don't know much about this mad science. Do you have an example I could follow?

So if I come up with a solution, I'll try to post my results here.
I got that idea after I remembered https://vimeo.com/46062006

part of interest at about 7:20

I guess you could maybe animate the lattice in a way you want your tentacle to move, and then just animate the tentacle going through it.

I honestly don't know what you are trying to achieve thou.

If you want the tentacle to go to the same point every time, I guess you just don't edit the end part of the lattice.

It would help us to help you if you explained more what you are making. Chances are someone done something similar in the past, and we may know better ways
Has anyone here worked with the Character rig tag in Cinema 4D (they couldn't have picked a shittier, more confusing name for it)?

I'm trying to figure out a way to get it exported out of Cinema 4D to an FBX file by making a skin rig for it, but it's extremely clunky to duplicate (it's forbidden unless you break it first) and remove all the excess controllers.

There's got to be a way to export the rig, but because of the horrible name of the rig tag, I can't find any information about it without getting it confused with character rigs in general.
Trying to make short-cut grass in realtime. Can't find something that looks good from top-down and doesn't tank FPS. Any suggestions?
File: jose-olmedo-cocina-95-01.jpg (405 KB, 1920x1920)
405 KB
405 KB JPG
are those isometric room renders that people make are they modeled any differently than for a normal render or is just all in the camera settings?
>pic example related
Unless you want to go for some Escher-esque shenanigans, then no, just set up an ortho camera, or in absence one, a far-away camera with a really long focal length.
File: BOSS2.jpg (21 KB, 320x240)
21 KB
I want to recreate the shading/lighting that you see in old pre-rendered video game graphics, like duke nukem. donkey kong country, or myst. Especially pic related. What should I do to get this look?
File: unknown.png (126 KB, 401x459)
126 KB
126 KB PNG
so here is the concept.

So, I have 2 tentacles, and they'll be putting their 'fingers' in the holes and the spinning the thing.

so the holes will be sorta random on which ones they will ender, and grab from.

SO basically, I WOULD have really liked to put empties on all the holes, and then have a tentacles path be parented to the holes, and then the rest could work.
Anybody have a good place where you can learn MODO like shortcuts and other basics? Been a few years since i used a 3d-program.
File: unknown.png (55 KB, 454x293)
55 KB
so as the wheel spins, it would collect the tentacle. rather than me animate the tentacle I would just animate its path and since its parented to the hole, I wouldn't worry about clipping and scrubbing through 100+ frames .
>lambert diffuse
>blinn specular
>specular level to 0.2
>spotlight or point light
>no anti-aliasing or 5
>low resolution, preferably below 720
>can use either bump maps or normal maps
>rendered in none-physical rendering engine
Use a basic shader model like phong, make use of ray-traced shadows, render with a ray-tracer instead of a path-tracer, and save the image in a format with a reduced color palette, and at a low resolution of course. Just do things as primitively as possible.

oldfag here

HASH Animation Master(used to create the models/sprites in DN3D, Shadow Warrior and Blood. A nice and cheap tool to have back in the mid 90s) only had Phong shading back then.

There was some touchup involved, mostly cleanup work and palletizing.

No raytracing shadows, only shadow maps from spotlights. Grayscale bump maps, as no normal maps existed back then. Paletted textures and procedural textures, the later often in bump inputs.

Ironic...the sprites are low-res, but the models themselves had very dense topology. Animation Master does not support polygons in any form, it was ALL modelled with NURBS/patches.
File: hash.jpg (215 KB, 1320x488)
215 KB
215 KB JPG
It's still being developed/sold. The UI hasn't changed that much since the early 00's.


^ this was the first release I used. It came with a VHS tape(but no stylish tie :| )
ok, Look into Animating with changing influence on constraints.

You want multiple holes to be set as constraints of the last bone (the tip of the tentacle) to be set as 0 influence, and as the wheel spins, you change the closest holes influence to 1.
You may probably use a constraint like distance, or location, but I think a track to constraint would work best.

You may put a bone in a hole for the tentacle to copy the angle of it too to prevent some clipping with another constraint.

As for the rest of the tentacle, the easiest way to complete the look would be setting up a pole target for the rest of the bones and animating that.

I think there is also a good way of getting that result with a spline IK.
Like having a Spline starting at the ground going through a tentacle to a hole, and having a tentacle with armature follow that path. You then need to control where the end of the spline goes with a target, and manually changing the end of the spline. Animating of the tentacle could be done with stretching of the armature across the path, I'm not sure if that would work thou.

As for the clipping goes, the smaller the holes, the more likely there is gonna be some clipping you will need to fix. I guess putting tubes in holes is quite hard... look at all the clipping in 3d vaginal penetration porn.
What are good guides for making 3d videogayme assets?
I checked the sticky, it had two links for this and they were both 404 dead as fuck
how do i figure out the dimensions of buildings in my city to model them without measuring them myself? using OSM and Gmaps i can only find rough measurements for the building's footprint but not the height.
Can you use decimated .obj's (talking about assets not characters) from Zbrush in a Maya scene for a short animation without retopologizing it? (like a statue or a tree or whatever)
File: jojo_c01.jpg (291 KB, 1000x563)
291 KB
291 KB JPG
Nips love Lightwave
>Is it going to deform?
Use Zremesh
>It is not
Use decimate
File: WHAT.png (129 KB, 322x604)
129 KB
129 KB PNG
What the fuck happened to my cycles render quality? The scale resolution is set to 70%.
File: DTZ25qOV4AYMlHl.jpg (74 KB, 1199x668)
74 KB
So I used to use after effects to compile renders when I was in college, but now that I'm out my Adobe cc has expired and I really would prefer not to spend money monthly. How many of you (if any) use a cracked version of after effects or the adobe cloud? Do any of you use alternatives?

Blackmagic Fusion and Foundry's NUKE
If you're not making money, just download CC directly from adobe and crack it.
They only care about people who make money using their software. They can't litigate after every starving faggot.
why bake shadow maps? if you use none-physical render engine it takes a few seconds to render shadows.
No, you misunderstood me.


Shadow mapping as in set your lights to use shadow mapping with a specific map size, e.g. 1024. Nobody used raytraced shadows in animation in the early-to-mid 90s. Too slow. This was limited to studios.
File: shadow_map_texels.png (120 KB, 766x371)
120 KB
120 KB PNG
It's obvious downside of this
1. no proper half-transparent shadows without some hacks 2. instead of render time it ate away memory
3. small map size resulted in blurry/glitchy shadows
File: Hair wip.png (479 KB, 805x760)
479 KB
479 KB PNG
What's the best way of sculpting blocky hair like that?

It's my first time sculpting a full character, so everything is new to me.
Should I just keep increasing resolution and keep doing what I've been doing?

I decided to have a break and watch some tutorials on youtube, but I also thought it would be a good idea to post here, since I may get a reply from much more experienced people.
Another question, more to do with character design than just 3d:
How do you develop consistent art style?
I get a feeling that when I make another character and put them next to each other, they are going to look off.
Do I just choose some features and keep them relatively the same in different models to keep the familiar look?
you can use curves which looks more smooth and you can control each stand.
i had no problems with sculpting other than you need to really refine it
File: 1517228450959.png (349 KB, 720x405)
349 KB
349 KB PNG
I heard that advanced software like substance painter etc have built in DRM features that mark the created content with metadata and whatnot maybe invisible watermarks (you tell me)

I cant afford substance painter.
How do I make sure the models I made with Blender and SP are 100% clean of any possible DRM features etc?
I obviously mean that I cant afford a license and have pirated it.
when i get into weight paint, i always have troubles with the mirror function, the weights aren't completely simmetrical,
is there a way to copy the weights settings of one side of the mesh and past it on the other ?
what do you think of this ?
i can't get it to work correctly, am i doing something wrong or is there better alternatives that i don't know of ?
it's for average poly characters
by the way, i'm using blender
File: 1493151628840.gif (44 KB, 725x503)
44 KB
I hope that image helps

So do you think keeping every strand separate is a better approach?
watch yanal sosak youtube channel he got plenty of good methods for hair
In my opinion fusion\nuke are far superior to adobe simply because they are node based, this guy(biased towards node though) covered layer vs node pretty extensively:
Both Nuke and Fusion also have legal free options, with Fusion being slightly outdated but completely free and Nuke non commercial being somewhat restricted but up-to date, so you don't even have to bother with cracks
File: 1b5.jpg (23 KB, 600x600)
23 KB
Sorry stupid question, should've read the Terms and Conditions in the nailed thread.
I can't believe nobody here did this at least once - literally how are you supposed to create a game-ready building/house with both the exterior and interior?

I want to create a small farm house, so I'm not even sure if modular is the way to go since not many pieces would be repeating. BUT, how do you do it, modular or not? If you have a 2 by 2 wall on the outside, it wouldn't fit on the inside because of the offset. And if you model everything as a single piece, how would you do interior then? I've seen one UE4 environment where the guy basically had exterior and interior as 2 pieces, and it seemed like he used a solidify/shell modifier to get the inside walls, and then separate them from the exterior. And then he just made them fit in the engine. But you don't get culling that way. Seems like a clunky way of doing things.

I've googled this a lot and nobody on forums ever provided a clear, simple solution to this. I know that you could also have walls as blocks, don't about that one...
you won't get a clear answer because every studio does it differently.
some people like to have big chunks of walls, some people like to atomize it into small Pisces and then build it in later.
find the best way for it to work, but you have to do it in such a way that everything has to be laid down on a piece of paper/have folders for everything otherwise its gonna be a mess.

open a new scene and turn on grid snapping, start messing around with cubes and planes, measure things etc. also make sure to include the measurements in each object name or something to help you decide how things are gonna be done.
if you want some section to be broken down or be different then you have to include that too.
make sure you have a complete concept file with everything you want to make + at least dozen or more reference images
>Every animator should start with http://www.theanimatorssurvivalkit.com
and after I know basics of animation what tutorials would you advise watching for say animating walk cycle in maya (I don't have money to enroll on costly cources).
File: bug.jpg (318 KB, 1681x1050)
318 KB
318 KB JPG
Is this a bug or something I did wrong ? just trying to make a simple thing, the texture won't show up
mix shader should be before diffuse
File: the mask.jpg (449 KB, 1920x2160)
449 KB
449 KB JPG
I got a question / discussion topic that I want to make throw at you guys.

I got an idea for a The Mask-style of 3D character who looks "normal," but makes extreme cartoon facial expressions. The dream would be to make a single character who can then, say, have his jaw drop to the floor and have his eyes dart out like The Mask in the pic I provided, but I would have to sneak in A LOT of extra geometry, lots of additional edge loops so that such a character's face could possibly stretch to such an extreme without looking like complete garbage. However, realistically, I think I'd have to make two models. One of The Mask's regular face and his regular expressions and then an entirely different model that I would switch to for these extreme cartoon expressions.

Is there a way to do both with just one model and have it look good? What are your thoughts?
not that guy, but you pretty much sold me on nuke (or fusion, I need some videos on that one)
You see how your yellow outputs are plugged into green inputs? That.
I think you need https://docs.blender.org/manual/en/dev/render/cycles/nodes/types/input/texture_coordinate.html

using UV to plug into your Image textures.

I advice enabling node wrangler for getting those set ups with virtually one click

You could sculpt different extreme expressions and use them as blend shapes together with switching normal maps for sculpted detail.

You are always going to be limited by how dense is going to be your low poly retopo mesh, no matter how much detail you are going to bake for the blend shapes

So it is possible, but it's going to look good depending on how dense your actual mesh is.

here's a top search example of what I am talking about https://www.youtube.com/watch?v=kWS5Ymb_PXM
how do i get faster at sculpting?

I like and know how to, but generally take too much time on the base
how are the tutorials found on udemy? final day of some sale and it reviews well outside the site https://www.udemy.com/blendertutorial/
cgcookie have the best blender tutorials, the ones on other sites go through only the basics
File: mod.png (109 KB, 816x658)
109 KB
109 KB PNG
I was able to dump the in-game models for DBFZ (which uses UE4), but I'm not sure how to import the models and their textures into UE4, because the models are .PSK files, which I am unfamiliar with.

Not seeing much in the way of tutorials when I search online. Can anyone offer any insight as to how to accomplish this?

Pic slightly related; an untextured game model in Umodel's viewer.
well if it's a base you need to speed up with, why not just make one base mesh and then just editing for every project

Another advice to speed up making a base mesh is to use more primitives.

Another advice from me is using much lower resolution for blocking out the big, basic forms.

Memorize your shortcuts and use them

and of course get more experienced, and don't get distracted
So I dont play videogames and mostly use Blender. Could drop PS for Krita and GIMP too.

How do I go full GNU Pinus?
I'm getting sick of latest Windows 10 developments bugs and shit
>launch pc
>cpu load 100% for 5 minutes loading bloatware and bloatware accessoroes
>launch firefox
>it doesnt fucking work
use brave and disable updates
your pc probably comes with alot of bloat like anti-virus stuff so delete that
Brave is just a Chromium fork with pre-installed adblock, thats it.
Chrome is fucking shit because
1 it renders images in sRGB color space ignoring all color profiles and display calibrations
2 some webm files dont work on it while they do work in FF
So fuck Chrome it sucks ass.

As for bloatware and shit its mostly launchers and other shit from proprietary software (Adobe center, Autodesk Center, dicks and shit center, bloatware addon community hub, windows defender is checking them)
>turn off updates
So I get my credit card stolen by spectre and my hdd encrypted by wannacry?
>Thinking he's protecting on newer software that literally everyone is running
Just go so old that nobody gives a shit about it anymore.
Do you see any viruses for Windows 3.1? Didn't think so.
File: figure6.jpg (16 KB, 407x260)
16 KB
Are there any renderers that have GOOD glare and glow based upon the camera properties rather than some simple shitty raster blurry post processing?
There's a psk import plugin for blender.
You should be then able to export it as a .fbx for unreal.

You might be able to just load the .uasset in ue4 by placing it in the project file, I'm not sure.
So I was looking at job openings at some game studio, doesn't matter what and one of the openings that caught my eye was for a hard surface modeling job or something. Super basic shit.
The thing though was that a requirement was proficient programming skills. Is this common in the industry? Its probably a dumb question.

Art is something that seems to be hard for programmers to grasp. So if you can do both you can do fancy shaders/code/whatever without having to try to explain to a programmer (who probably doesn't understand) what you want.
m8 i used to have lots of updates and the newest anti-viruses.
then i realized 99.9% of viruses come from shit tier browsers like IE.
i rather have a secure browser than 360 security pack
yes it is common for small studious.
usually you are called "the 3d guy" and you ar taking care of ordering work from freelancers then implement it to unity and do basic programming on it.
I wonder if editing a bunch of preset values in an established shader constitutes as "scripting experience" in the 3D modeling world.
1.Why is /3/ a blue board? Wouldn't it be better to make this board red like the other art boards?
2.I once saw post on /ic/ that said that told another anon that going in to 3D art because they couldn't draw was a meme and that 3d requires you to also be good at art if you want to get good at 3d. Exactly how good at art do you have to be to excel in 3d?
3. Who's the best solo 3D animator? Just asking out of curiosity.
>Exactly how good at art do you have to be to excel in 3d?
After starting with a very simple zsphere tutorial for zbrush a few years ago I would have to say absolutely zero percent. https://www.youtube.com/watch?v=rR3b1JfE0qc
You dont have to to be good at art but if you are its a good help.
I pretty much knew anatomy, materials light reflection shenaningangs and had sense of aesthetic and composition before opening a 3d editor so i could make cool looking shit after googling "How do I do X?"

You will just have to learn that with 3d rather than 2d media
What are these pink and blue lines in my imported model when I select it on blender? They make it harder to see.
Why would you do that? Pros do this in post. More control.
>Why is /3/ blue
Because idiots who don't understand art set this board up
>3D to get around lack of 2D skills is a meme
Sorta. If you have good 2D skills, it helps. But it's not at all a requirement. I can't draw straight lines due to lack of control, but I'm at least passable as a 3D artist.
>Who's the best solo 3D animator
This highly depends on opinion.
So THIS is the power of DonutMaker™
I would like to set up a bounty for a certain rig type I cannot make, nor can find a tutorial on. Where would I do that?
What's the best way to Retopologize in Blender for beginners?

Downloaded the tool Retopoflow but it doesn't really make practical sense to me considering you can't make straight lines.
Doing it manually seems somewhat tedious but I assume that is the best possible way? Manual+Mirror Modifier is best for Retopology?
Is this really the best option or is it just the first video you found
its the best option, especially when you don't want to use plugins that later become problematic because of bugs

there is one thing he forgot which is merge function, but you can learn it yourself, assign a quick key for it
If I didn't want the interiors of my building to be visible at all, I'd just use modular planes with reflective glass windows. But what if I want it so you can see inside of those flats a bit, like in pic related?

Would using full blocks as modular pieces be too much for performance? What if I just want a standalone scene, since I'm not making a full game, though?

Because if I created the pieces as boxes, I could literally place some spotlights with different colors in each one of them, and you would be able to see a bit of the ceiling, which is enough. I also thought about using textures on windows which simulate that, but I don't think the result would be equally good.
File: photo2large.jpg (1005 KB, 3264x2176)
1005 KB
1005 KB JPG
Does oxidated surface work as metallic or not?
Stop thinking in terms of 'metal vs. non-metal' and start thinking in terms of 'electric vs. dielectric'.

'Metalness' is an simplified term - what it actually relates to is whether or not the surface conducts electricity. It's not just metals that conduct electricity - water does, for example.

If something conducts electricity well (eg. gold) then apply a high metalness value.
If it electricity not so well - lower that value accordingly.
I know that the coating itself is dielectric, but it is very thin and I dont know how transparent it is so does it work as if the whole surface was dielectric or is lit like metal covered in lacquer?
Here's a radical suggestion:

Try it with a higher metalness value.... THEN try it with a low metalness value.... THEN (stay with me now, we're about to hit the payoff....)


Your mind just got blown. Genius just happened. Etc.
File: 1502282602612.jpg (47 KB, 500x500)
47 KB
where can i find vagina references for 3d modelling (besides porn)? i don't have access to real life pussy
What is the best program for a beginner to use for industrial design?
Do you know of any pose libraries that would be good for drawing practice with the Lab?

Delete Post: [File Only] Style:
[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.