In Maya how do i render an hdri/ibl when using sun and sky? Or, how do i just disable the sky in sun and sky from renders? It seems like like MR for Maya is missing a lot of the options that Max has.
Hey /3/, i made a large number of rapid fire props for a game i'm developing (pic related), and now i'm in the unexpected position of having to unwrap them, in order to put some grime and damage on there. Is there a tool that can do a rough unwrap and process a large number of fbx files? The thought of wasting days of development time on processing hundreds of props is depressing
>It's too hard so i won't do it unless you give me an easy way out
Do you not realise how fucking retarded this sounds?
Fine, have a shit game because you couldn't be bothered to put some effort into UV unwrapping.
Do you think studios have some magical one click program to do this shit for them?
The optimal way to do it is not be shit.
If you're going to polypaint the textures then your UV layout doesn't even matter as long as it's in 0-1 space, non flipped, and not overlapping, ergo use auto unwrap.
There are shortcuts in any medium. But I just asked a question, i don't get why everyone's getting so triggered :)
It doesn't sound retarded at all, especially since it's not a matter of difficulty, but of wasting lots of time on menial shit. If i have to choose between making a compromise, or not making anything at all, i know what i'm always gonna choose. The optimal way is the way that is time and cost-effective; i don't live in a dreamworld where time and resources are limitless, and i've gotta spend my resources in an efficient manner.
Thanks for the tip, i'd love to be able to batch it somehow though.
for future reference oyu should be unwrapping as you make each prop so you don't build a pile of tedious work that kills your motivation.
for now try using an automatic mapper, load them one at a time, smash the button, and save the default results. Maya's unwrap tool will probably make adequate maps. You oculd even load all of the meshes into one scene and use the tool on all of them at once. I also like zBrushes uv master but it take an extra step or two for every mesh.
How on earth have you not figured this out by testing it/experimentation yourself is completely fucking beyond me.
No, it does not save automatically. You have to save either the project or the tool, not the document.
trying to write an expression in Maya to make eye controls move slightly. Like flick from looking at one spot to a new spot, very slightly.
So far i've got this:
if ( frame % 5 == 0 );
Null_EyeRand.translateX = (noise(time*2)*0.2);
Null_EyeRand.translateY = (noise(time*2)*0.2);
the problem is noise creates a smooth movement which doesn't look like the minute movements of an eye changing focus. I can use rand, which jumps from value to value instead of transitioning, but it changes per frame which is way too fast.
Does anyone know how to write something more like what i'm after?
that isn't what i'm trying to do though. That would just make the eye movement too quick. I want it to jump to a value, hold that value for a duration, then jump to a new random value.
the rand function jumps to a random value but it won't hold for more than 1 frame.
i thought that's what "if ( frame % 5 == 0 );" did, but it still evaluates every frame instead of every 5th frame.
CAN SOMEONE HERE SAVE ME
I am not a /3/ person, I came here just looking for help because I need some model changed because I have to print it in an hours or so
all I need is the text on this STL removed, please
just need the thing to be flat/textless surface
been fucking with blender and other free software and can't figure out how to /3/ on basic level at all. not to mention that the mesh of the letters seems to be fucked in some strange way
in case anybody might need this in the future,
$randNum = rand (1, 10);
if ($randNum > 9.5)
objectName.attribute = rand(-0.2,0.2);
I'm only using this for tiny eye movements and keyframing over the top as well, but this effect would work great with bigger movement values for an insect or bird type creature.
market is give-up-now tier. Even if you have the skill to compete, you're in for a shit of a time competing for placement in studios, and if y some miracle you get in somewhere, enjoy your slave labour hours and no benefits.
You're best bet is trying to find an indie studio to work with.
You can't make a career out of it anymore. There are too many established artists who do it full-time now and even they are barely scrapping by and are at the whim of Valve's selection process.
Luxrender 1.5 just came out and when I tried using the Path OpenCL I the message in pic related.
Could this just be a problem with my GPUs? I am using two Radeon 6770s. Could a GPU upgrade solve this problem?
If i want to get into editing/using mocap data, what's the place to begin? Where can i get mocap data from? What software is best to process this data, Can you use any rig you like with the data or do you use a specific rig based on how the data was captured?
Read the error... "not enough device memory", which means GPU RAM. Your card is probably only 1GB, maybe 2GB, it's very easily to go over that limit.
Also, AMD's OpenCL drivers are notoriously shitty, I would definitely recommend upgrading from that low-tier 6770. Get a Nvidia 660/760/860 or higher. Don't go with any of the x50 or lower types, those are mid-range and lower. 760Ti is great value for money right now.
In Zbrush, is it possible to apply a texture on an already existing polypaint character?
Like say I have a polypainted character that has all of their shadow and light painted on them, but I have the finer detail on a texture image. Am I able to just apply like the PSD to character and KEEP the shading, and ADD the details on top?
You can have a texture on and toggle it off to see the polypaint. But if you apply the texture as polypaint, it will overwrite the existing polypaint data... What you want to do is Create Texture from Polypaint, and then blend that texture with the other in Photoshop so you only have the 1 and apply that one.
in Max, how do I select things so the type-in affects them all individually. the type-in boxes should be completely blank instead of blacked out numbers, and if i were to type, for example 3 into the X type-in, the vertexes would all be moved to 3 on top of each other, instead of moving them by the center of my selection. I've had it before but i don't know how to go back to it.
tl;dr how do I type-in values to move things all to the same value instead of the center of the selection (like in the image)
so i've learned that it works this way with multiple objects, but not at the sub-object level, where it's based off the tripod in the center of the selection.
is there any way make this work at the sub-object level or is this simply the fate I'm doomed to live?
Why do some of the materials change after I drag and drop new materials in the ZMaterials folder? It's like they're overwriting the old ones despite the names being different. Loading them directly through ZBrush doesn't fix the material either.
Because you're doing something you're not supposed to, putting materials in the ZMaterials folder...
The "ZStartup" folder is where you're supposed to put all your custom content, as it treat it differently when put there. So put your materials in ZStartup/Materials, and they won't affect other materials.
I'm not sure how you're causing that then, because I duplicate and rename materials all the time and never run into such a problem.
Also, what you're pointing to in that last image isn't even a material swatch... It's one of the textures used in the z7 material, notice how the background is colored instead of transparent?
Is that a custom menu you made?
The problem might be that I'm trying to work with over 200 materials. Instead of dropping them all into the Materials folder at once, I'm going through 15 at a time and getting rid of whichever ones I don't want. Tedious, but at least I'm not running into the problem anymore.
Anybody know of a way to cover up intersecting meshes like this? only thing i can think of is masking the reflection, but you still end up with a pretty obvious seam.
Intersecting geometry will create very visible jaggies on your screen at the intersection unless you have high anti-aliasing, since your vertex shader can't interpolate that transition normally like it would with a continuous surface.
Nope, there's no way really apart from perhaps creating a transparency transition there or mapping the shading to be killed off before the intersection. Your normal map can help hide it as well if you paint/bake a transition or more natural looking crease there.
Is this just a small bit? Or are you planning on instancing this across the mesh? Because I'd say just slice an edge above it and connect it if it's for games, you don't need to worry about shit topo in an area like that since you won't be smoothing it nor deforming it much.
I know what's happening, I'm wondering what the mesh is. It looks like a small detail which won't ever be fullscreened. It's hard to judge how much must be done on a cropped image like this. It could be really noticeable or completely negligible. We don't even know if it's animated or static.
Is there a plugin (possible for free) compatible with Photoshop cs2 that lets me paint black and white heightmaps and make them into normalmaps that I can merge with already existing baked normal maps for a model to draw on detail like patches, engraving, stamps etc.
just paint a black and white height map, then use nvidia normal map filter (free PS plugin) to turn it into a normal map, then Overlay that normal map over your existing normal map, duplicate/merge them then use normal map filter to normalize the normals (because overlay brings the ranges outside of normal map spectrum).
Okay, thanks, I guess thats the way to go, but is there something that lets me paint directly onto the normal map and it automatically updates it in realtime into normal map info? Just wondering.
>Is it hard to learn Blender?
No, there are a handful of quirks that you have to disable in the Options menu or solve by enabling free add-ons. Turn on "Dynamic Search bar" add-on for one, and whatever the newest iteration is of Polystrips or the like for quad draw (although I think it's a bit more powerful, in that you can define a distance and then set the distribution of quads), also the add-on called "F2" is really nice
Take the time to make a proper Start-Up file for Blender to load every time with your UI tweaks, render / viewport settings, and so on is worth while.
Other random useful tips: Ctrl Arrow keys to cycle through the different UI layouts. My recommendation is on Default to press the + key so that it makes a Default.001, then go and edit that or rename it or whatever (just so that you have the Default untouched), for some dumb reason there's no "revert UI to default settings"
Other tip: You can split blender windows by clicking and dragging on the little transparent plastic tab looking thing in the corner, to merge windows click and drag the border to the right such that you see a -> icon. Basically watch Andrew Price's video on UI basics, it makes sense once you see it but trying to describe it is hard
I don't know about Maya, but blender is the exact opposite of Max.
Worse at render, better at modelling.
It's very keyboard-based and you're probably going to fuck up the ui with some hotkey.
>No i finished school = no more maya/max
u wot? It's fine to spend a month or two with Blender to see what it's like (the free Add-Ons are extremely important), but why can't you continue to use maya/max? Just keep using your student license til you get hired
What exactly makes it "worse at render"? You don't even have to use Cycles, there's plug-ins for a bunch of other renderers.
I don't know why "hotkey based" is a detriment, you learned the hotkeys for basic tasks in Maya, right? Right? I mean, you're not a shitter that opens a menu every time he needs to do something, right?
You can find everything in Blender by just pressing Spacebar (*with* Dynamic Search Bar Add-On enabled) and typing some of the name, or you can search through the context specific menus (Dynamic Search Bar is still better for this, the Add-On groups all the menus together), and if you *REALLY* have to use your menu bullshit, there are Add-Ons that copy the menu systems from other programs, again all for free.
in maya I can rotate my eyes fine just using the rotate tool, but when I attach them to an aim constraint and move the constraint the eyes do some kind of weird 180 degree rotation when rotated from left to right or top to bottom
what's the cause of this?
I have a spline in Maya and i want to create a braid and conform it to the spline. It's already a giant pain in the ass just to extrude a shape along a spline in Maya, especially a closed spline.
Does anybody know of a convenient way to do that, or should i do it in Max? I haven't used Max for years so i still have 2014.
Hey do any of yous know how rendering hair works in Blender? Because in the rendered viewport, I get nice smooth hair but when I render the frame with f12, it goes all low poly and shit. I've got a half fix, I can change the primitives under "cycles hair rendering" to curved segments, but that makes the hair curly slightly.
I just what my proper renders to look like the viewport renders.
How do you work inside of your models and in crevices? Working in wire-frames gets confusing if your area is complicated and has a lot of vertices around that you could accidentally hit.
Like, if you wanted to detail the insides of a character's mouth, their armpits, and their taint. These are all very hard-to-reach areas.
When I use OpenCL rendering I get the same render speeds whether I use just my CPU or my CPU+GPU. Is there something I'm not doing right? Also when I try to render with just the GPU I'm getting 0 samples per second like it was never using my GPU to begin with. Is this a problem with my GPU or the renderer? And is there a fix for this?
GPU: GTX 750ti
Renderer: Luxrender 1.5
How do game engines deal with free moving/'dynamic' clothing or in this case, necklace thing
I was considering adding dynamic splines, parenting joints to the splines, and skinning the feathers to the joints.
I don't really plan to put it in an engine, just wondering what the procedure would be.
I guess i could always just skin it to the body's skeleton also.
why do the knees get completely fucked and not bend forward when I move the foot plat bone? how can I solve this?
woops sorry the webm got a little fucked, is there any way to fix this without rotating the knee?
select your IK end effector and check under 'IK Solver Properties' in the 'Motion' panel.
The box labeled "IK Solver Plane" has a box labeled "Pick Target:".
Click this button and select a helper object or whatever you wanna use as a target. Now the knee will always look in this direction.
Also when building a leg make sure that the hip knee and ankle lies in the same plane if you want it's natural bend to point where you imagine.
A real human skeleton do not line up like this, but the IK algorithms are very simple and require this exactness to turn out right.
hmm, that is odd, the only problem Ive had with this is path using the classic api, since I use the luxcore api I never really ran into this problem.
Id suggest to look at all the documentation o lux and see if there is a solution.
It seems rotating the leg object "fixed" it, but do you think this would cause problems later?
In maya, how can I increase the max influences in a skin cluster without causing the entire envelope to be recalculated?
I have a rig with max 2 influences, but I've discovered my face rig is gonna need a max of 3, but when i set it to 3 influences it recalculates the whole thing and fucks it up.
duplicate your skinned mesh
freeze transforms/del history
bind the duplicate to the skeleton with the new settings you want
select original mesh with skin weighted correctly, select duplicate mesh, skin > copy skin weights option box
influence 1: one to one
influence 2: closest to joint
influence 3: label
Actually i just thought there's probably a simpler solution. You should just be able to lock the skin weights in the paint skin weights tool. Just click the padlock next to each bone in the tool.
That previous method is good for when you cant delete non deformer history on a skin mesh, i.e some UV operations.
took some searching but i actually managed it
still you can't solidify it cleanly after the whole process
I'm a little stuck on this. Making a thing for soldier so I copied his head to speeden the process. What I'm doing doesn't need a detailed nose or mouth so I removed that to save polies. But now I'm kinda stuck on how to fill the mask. Any suggestions?
This will be helpful: https://youtu.be/MpsURXZvcsU
In your case you will leave holes instead of that pattern, but it's close enough.
I know he's using 3DS Max, but I've used his tips and techniques in Blender without a problem. Just focus on what he's doing and try to translate that into Blender.
I'm fiddling with SketchUp.
I've figured how to export outline-only, but how about "filling only"?
Pic related, solid colors with the object's color.
A way to export shadows only would also be nice.
i just added a placeholder texture on my model - its not all kinds of fucked up.
It's fine with the default lambert, but adding a text with an image from file does this weird 'see-through-to-oither-faces thing
That's because it's auto-connecting the alpha channel to the transparency on your material, and since your viewport transparency setting is on the default basic version, then you get this sort of poor transparency sorting even when transparency is 0%. Set your VP2.0 sorting order to "Depth Peeling", or right click the transparency channel on the material and "break connection".
Mirror Skin Weights, what the fuck?
I have a symmetrical mesh and I'm trying to mirror the weights for the face joints. Both "closest point on surface" and "Closest Component" yeild the same result. Skin weights which seemingly skip every other vertex. I dont understand what the problem is. The mesh is symmetrical and so is the skeleton. You'd think the "closest component" would find the exact mirroed version of every vertex, and that closest POS would be the same. Yet it's not.
The mesh is high density, and being these are the face joints the weights really need to be exactly perfect.
If I absolutely have to I'll load an older version that hasnt been smoothed, but I'd like to not have to do that.
ok I seem to have fixed the problem but I dont really understand how.
Before I was selecting only the verts on the head in order to avoid any unwanted hiccups through out the mesh. I accidentally retried with the whole mesh selected and it worked as expected. Is it just not possible to select specific verts for skin weight mirroring? What if part of my mesh isn't symmetrical?
is the Intro to ZBrush and Character Design good enough for someone that wants to get into 3D sculpting but might later on want to retopo for usability, or should i try to understand topology and edgeflow through polyediting before trying to do sculpts? i'm mainly interested in game ready characters, but i'm not sure where i would have to start, be it the optimizing or the character detailing.
whenever i animate something, i render it out to a series of image files then sequence them together in a video editing program. however, after i have my video file, i still have all the frames of my animation saved as images.
should i keep these? why? how long should i keep them? i don't do anything with them, so they're just taking space on my computer; but i don't know if i will ever need them for something. what would i ever need them for?
>What programs can open .ma or .mb files besides Maya?
>Are there any programs that can be used to strip metadata out of .fbx files?
How does one find the worldspace coordinates of an object in Maya, in order to mirror geometry?
In 3DS Max is gives XYZ coordinates for every object, so one can simply copy them and reverse their position in the world.
How does knowing world-space coordinates help you mirror something? That doesn't make sense... What matters when mirroring an object is where your PIVOT point lies within the object. If you want to mirror an object so it connects with itself, then your pivot point should be positioned at the point you want to mirror across.
On your keyboard, just tap "Insert", hold down "V" for vertex snapping, drag the pivot to your desired mirror line, hit insert to confirm the pivot change and now do your mirror.
It takes only a few seconds, no world-space values to deal with. Perhaps mirror only happens across the center of WorldSpace in 3DS Max, but in Maya the mirror function happens across the local-space center. And you can easily find world-space center by just looking at your grid. Hold "x" to snap to a point on the grid, like the center.
But if you just want to know the worldspace position regardless, you can find in the same area you see your local-space transforms in the Attribute Editor, it's just below that under "Pivots>World Space".
I'm trying to convert a 3D animation file from a proprietary, unsupported format to one with more support.
The animation file gives me 30 evenly timed keyframes (0.1333 seconds per keyframe) quaternion rotations for every joint in the bone model. So that's about 4 seconds of animation.
The animation is supposed to loop smoothly.
What file formats have the closest, most similar implementations?
I want to translate it with minimal changes to the original data.
I will also probably be using the assimp library.
Guess I will use this thread instead of creating a new one.
I'm totally not used to this board, but I thought maybe you guys could help me there.
I'm extracting some Borderlands 2 models, and I ran into something I never encountered before. I did some colage to expose the problem.
On the left, you have the character I wanted to get ( Lilith ). Next, you have what I extracted from the game. Notice the textures that have obviously not the same colors. I checked in the textures availables, I got the usual normals and these red/green things you can see there.
The files was named with a Comp extension I never saw, others where classed with D for Diffuse and N for Normals. I checked some other characters, they also got these kind of files, some with Color Comp extension, but I don't get what it is anyway. I thought maybe some kind of Color Balance or Palette but I really don't know where I could put that. I'm totally not used to Color Management anyway, if that's what it is.
Obviously it's two textures compressed in one. So after getting the coverage right, I tried to apply them. The right one reminded me of some sort of alpha map, but I could not understand why it would be color-coded. Besides, the fact that all the hair are in the red part made me think it was also use for an additional color ( But why is the eye in green and the lips in blue then ? ). So I tried to do a layered material and mix the base diffuse and these two textures separately with any mode combination, with a certain success as they effectively changed the parts that was not like the result I needed, which are for the face, the hair, lips and eyes. But I couldn't get the right thing after a lot of tests, so I guess it's not there.
I don't even understand why they did something like this as Lilith has only one skin through all the game.
Anybody have a clue on how I need to use this ? I highly doubt I can get the result desired with only this, but there is no other material.
Most likely each channel of the texture (RGB and maybe A too) is a different map, like a mask for something, gloss or ambient occlusion etc.
It's up to you to find out what they are, if you can get some info about the shader they used for the characters in the game that could tell you what you want to know.
Yeah I thought of this too this night.
I ran into some materials informations coming from a UDK file it seems, called Material Instance Constant. Unfortunately, opening it with Notepad just told me the name of the Diffuse and Normal files, nothing useful. And I don't have Unreal Engine installed, nor I know how to use it.
By searching for this Material Instance Constant, I learned that it's used to modify textures on an instance basis with some parameters. So they probably could use this kind of material to modify the color on the fly with values instead of textures ( Which would explain why there is no other texture used ) ?
So I considered the left part to be some kind of lightmap as >>498090 said, AO and all, which are pretty much details so they are not important for now, I will figure this out later.
The right part seems to be a mask then, and the channels could be used to define 3 different colors. But some things doesn't fit, like the armband on the right which is on green-channel but should have the same color than those on the red-channel, or some parts of the pants which should be on the blue channel but are not...
Oh, this thing is giving me a headache already. I think I will just modify the diffuse texture directly by hand and whatever. Thought it would take me more time to do this than understand how it is working, but I guess not.
If anyone have some UE experience and can tell me what's going on, I will gladly take it. But in the meantime I will just get around for now.
Thanks for trying to help guys, I appreciate.
>Mfw absolutely all the characters are done like this and I will have to recolor all textures I want to use
Thanks a lot, now at least I know what they are. So the mask is more than probably here to define color replacement groups then ( Even if some things are wrong but anyway ).
And the other is a curvature map ? What's that ? Never heard of this either back in 3D school nor at my work. This said, I'm an animator, so maybe it was too special to cover it in general course. So basically it's used to store one kind of information per channel, or does it have a more general use ?
I will install UE and see if I can open a .mat file, maybe I will get more infos than what Notepad could tell me.
Thanks again guys !
Yeah, of course I tried Google and I got on the Polycount wiki. But let say that the informations here were not really developed, and there is absolutely zero other interesting/official result/explaination ( Not without searching on some forum posts ). Like, on which connection is it used, how does it work and that kind of thing.
But I don't really care about it, I was just finding odd that I never saw this kind of map anywhere else and was a little curious. That's not important anyway.
Question on making models for a game.
So you make the stuff in Maya and then you export it to your desired engine like Unity? Is that how it works? Or how does one go on creating characters, weapons or whatever for a videogame?
Engines and every modeling software uses .obj files. That's not the hard part, having workable, rigged characters being able to be exported and imported into an engine.
It's all specific.
Depends what type of game you want to make. Realistic/fancy graphics, or more simplified. If you want realistic, then you're going to need a good sculpting program like ZBrush so you can create highly detailed models that just aren't possible with poly-modeling alone. And then you bake those details to a normal-map so the game engine can fake those details for you.
Basically, the main things you want to learn these days are Maya/ZBrush/Substance Painter/xNormal/Photoshop. that's the "core" game dev toolset.
My friend asked for help building a computer to run Blender on a relatively low budget. I found a relatively cheap card ($150) with 4GB VRAM but it only has 640 CUDA cores. Should I look for a 2GB card with more cores? Or would 2GB not be enough to render animations and landscapes?
for soft image you would still need to be a student.
>all autodesk programs are free for students.
blender interface is shitty, but i mean... you are using GMAX so it cant be much worse. and the program overall is better.
Clara.io is cool ( its in browser, and isnt too bad )
and sketchup is also decent
>for soft image you would still need to be a student.
You don't need to be a student to use Autodesk's "Educational" versions. They're for learning the program, not just students, thus "Educational". So OP can use them for free to learn.
Start to finish on what?
From the default cube to a full length movie? No.
From the default cube to an animatable character? Quite a few.
From the default cube to a spaceship/car/whatever? All over the place.
From a collection of objects to an animation? Yes.
From a raw film clip to one with special effects and post processing? Yep.
From a series of film clips to a final video? Sure thing.
Youtube has a search function, and google has a video search. Plus there's quite a few blender blogs and sites for specific areas of interest.
Blender's quite a bit different than it was 10 years ago. It's still "nonstandard," but the interface was completely reworked in 2.5 and it's much easier to work with.
Worth a try if you haven't used it since 2.4. It still takes getting used to, though.
Complete noob here. I want to learn how to use Maya--I read through the sticky and checked out the various links. Digital-tutors seemed the most structured to me, but their free accounts don't allow dowloads of the project files. Will this be a problem or will my total beginner self be okay without them?
If they're important, is there a better free tutorial series to follow? Or a well-organized youtube playlist or something? Am I better off with a book?
Thanks in advanced for answering my dumb questions.
So I projection map this cube but on the lowpoly I get this weird hard edge.
leftcube is lowpoly
for what purpose?
>downloading a program on my shit internet that i might use once every few months if that
>some kind anon spending 10 seconds downloading a model, and already having Blender installed.
so you chose to waste time for the possibility of someone helping you instead of doing it yourself.
This one goes out to you anon, stay warm out there
Blender 32-bit is only 78.7mb. At 48kb/s, you will be able to download it in approximately 3.7 hours.
Then you'll be able to open any other Blend files you come across, as well as many Valve game models.
Ok /3/ need some help.
anyone know of a video where it shows a computer rendering a scene from blocky to sharp on youtube? im trying to explain something to someone and i cant fucking find one.
You mean 3D printed objects or prints of 3D renders?
In both cases, yes. Shapeways for the first, any print shop for the second, assuming you do your own rendering. If you don't, there are render farms available for that.
Here's another laugh for you guys, I have no idea how to close this hole with quads. :^)
A polygon with more than 4 sides
You're on the internet do you require spoonfeeding for every little thing? I'm not trying to front here, but come on man, you wasted so much time when you could've found it out yourself in an instant.
He literally asked what's an ngon, he devoted a post for that. That's like asking what's paper in a board meeting in a printing company.
If anything, your implication is just inaccurate, I'm not saying, stop asking questions in a questions thread, I'm saying stop relying so much on the thread when it's clearly not as smart of a move than just simply looking up the term on the internet.
fill face function and start making edges with a knife
- you can select 2 vertices's and create an edge by clicking "J"
-clean up the mesh before you do any complex operations, click a to select all and remove doubles
-use the knife tool by clicking "K" then while in knife tool mode click "C" to cut a straight line across an axis
-you can shape everything to quads, large or small but triangles are ok too
The trick is not to close it. It's not closed on a real boot. It's just a layer of fabric with thickness that then has a sole sewn on.
Since you're new, you probably aren't aware that not all parts of your mesh need to be physically connected. You can simply have sub-parts that are penetrating eachother.
anyway, apply a subdivision surface modifier, just to see if it's all ok and smooth. Maybe you have fuked up some loops and don't even know it, happens sometimes if you extrude 2 times the same loop.
Rookie here, I've a couple of questions about UV mapping in Blender.
First off, pieces 1 and 2 are exactly the same, so how come their UV maps aren't mirrored?
Second, why does piece 3 keep leaning? I tried two different versions of Blender and various different marked seams and it always leans to the side. Pieces 4, 5 and the unnumbered one do too, but it's less noticable.
I know I can rotate the islands, but that'd be far from precise. I can only imagine the frustration of someone working for days on a high poly model only for Blender to give him shit at some point.
>First off, pieces 1 and 2 are exactly the same, so how come their UV maps aren't mirrored?
because you need to have the same seams
>Second, why does piece 3 keep leaning?
its not leaning, it tries to keep the islands close as possible
Just use Sculptris instead for learning to sculpt. It handles millions of polys without issue, unlike Blender, and is just a better tool for sculpting in general. Then when ya wanna move on up, you can go pirate ZBrush.
They have the same seams and the pieces were mirrored prior to marking them.
>its not leaning, it tries to keep the islands close as possible
And Blender turns the islands in order to pack them as tight as it can? The same piece gives me the same issue even when I erase everything but it before unwrapping.
It's probably just Blender attempting to pack them better. Can you disable scaling/rotation in it?
For such a simple object I'd personally just turn on snapping and snap the verts of one to the other to get a perfect overlap.
1) I'm autistic when it comes to asymmetry
2) It'll be easier to do if everything is nice and orderly
3) I'm trying to learn as I go. I don't really want to put a problem that'll be plaguing me in every project in the back burner.
>Do you actually know of anyone that learned on blender and got good?
You're asking the wrong question. It's just software. Whether or not you're any good is up to your art skill and you can practice that with real life clay.
Whether or not someone draws with a super expensive pencil or a piece of literal feces is irrelevant. The end result will reflect the artistic ability, even though the medium will cap the ability a little.
Blender's sculpting capabilities are still fairly new compared to some other software packages. It only got dyntopo a couple years back (I believe it's roughly equivalent to dynamesh, although I'm not sure since I've never used ZBrush). It's had sculpting in general for a while, but the workflow was a lot different than it is today.
So while there probably are some people who "got good" using Blender's sculpting, you probably won't find a lot of them yet.
There are a number of people on blenderartists that do very nice sculpts, but I have no idea what they started on.
All that said, it doesn't really matter what you learn on. Learning the interface of a new sculpting app is easy compared to learning to sculpt in general.
So /3/, How would i go about making a cloak with ripples like this? I was considering drawing the hem-line as a spline, extruding upwards, and scaling the top inwards, then possibly relaxing parts of the mesh.
Does anybody know of an easier / more effective way? Preferably using Maya, also have Max.
I'm trying to create a fog effect in maya's hardware 2.0 renderer but this fluid container keeps casting a shadow despite having casts shadows unchecked. How do I stop this thing from casting shadows or if that isn't possible then what's the most effective way to create volumetric fog in the hardware 2.0 renderer?
Also why the fuck does glow not work in hardware 2.0?
Marvelous Designer is by far the easier way to make this, though the design is simple enough that you could just do an nCloth simulation in Maya using a sheet of fabric with a hole for the neck in it. Place it around the neck and let gravity take it down. You can save the simulated shape as the base shape.
I've tried this. It isn't very effective since there's no real way to force the ripples and folds. I'm trying to place collision geometry to force it to drape in a specific way, then delete bits i don't like, but it's a pain in the ass. I've never used marvelous, is it a bitch to torrent / learn?
Yes, that would definitely work. You can use a simplified version of the mesh as a deformation lattice for the folded version, and then simulate with that. So you retain the folds, but have it simulate nicely over the body.
I am following the Joan of Arc Character modelling tutorial and something unexpected happens when I try to continue creating the thigh after I extrude and try to give some proper form.
First time this ever happens to me, how is this called and why does it happen?
Yes, as suggested by the Tutorial itself.
After Mirroring the other leg, adding the edges, I need complete the rest of the body, extruding and giving form etc etc, should I remove those edges at the top giving me 4 squares instead of 1? can that be done?
No I mean, it looks like THE WHOLE mesh is being extruded, instead of the top 4 polys...
If you wanna upload the scene file I can probably take a look at it and tell you what went wrong.
And no, you don't want to remove those edge at the top, you'd end up with an n-gon and break your edge-flow.
So I recently came around to finding Mike Inel's 3D adventure time
I dont know what, but the idea that this was done by mostly 1 man, and to come out looking nice like this has filled me with DETERMINATION to try and learn 3D to make a video game (and maybe porn, gotta love GOOD 3D porn)
Problem is I know jack fucking shit, besides watching The Boogie making the Midna model for a while
What is the best starting point for me?
When making a model you use a Wire Frame for the structure and then a tool assist to round out the edges?
The T-Pose is the best pose to make sure the model comes out the most clean?
I am video editing in Blender, but when I want to output it for upload video streaming settings the quality it produces is not up to my expectations. Especially darker areas and dark gradients tend to have pretty bad fragmentation. Now I also have to keep in mind the file sizes should be too big. I maxed out all settings way higher than recommended and now I have more than one GB file size for a not even 2 minutes clip. Anyone has expirience with blender video output settings?
I wanted to compress some video clips for upload to youtube.
But I looked at the recorded video clips and they werent so great quality to begin with, so it isnt Blenders fault.
I recorded some stuff with Open Broadcaster Software in Unity engine with 60fps and already tried to get a balance between minimal frame stuttering and picture quality, might want to get a capture card in the future.
Does anyone know if it's bad for game performance to use floating geometry? For unity in particular.
Just wondering because it obviously is much easier to make new objects within a mesh for details and such, instead of add more and more faces to a mesh, making it more complex, harder to work with, and more high poly.
No, a mesh having parts that aren't physically connected to the rest isn't going to incur some extra processing. If you're saving polys/verts by doing it, you save on performance.
But you have to remember that you won't be able to create smooth shading in those areas where the two surfaces penetrate, it will always be a 100% sharp edge. So you should only do that for parts that would actually be separated in real-life.
>But you have to remember that you won't be able to create smooth shading in those areas where the two surfaces penetrate, it will always be a 100% sharp edge. So you should only do that for parts that would actually be separated in real-life.
Ah, didn't think about this. Still is useful, though, especially for car modeling which is what I'm working on now.
How do you disable this cage smooth preview, whatever its called, in Maya?
I accidentally removed the icon from the toolbar and I cant find it now.
Already pressed 1, but this doesnt do the trick.
Can anybody tell me how and pls remind me the name of the tool that does this cage smooth preview?
I'm trying to use turbosmooth to achieve smooth mesh, but I need some of the corners to be sharp. So I add additional control loops to make corners sharp, but those loops stretch through the whole mesh (I use SwiftLoop or Connect) and as a result I get that annoying line across the mesh. Is there any way to control a corner smoothness without messing up the rest of the model?
Insert edge loop
But with edge flow
Wow it's fucking magical isn't it
Or you can just delete the part of the edge you don't want and merge it to a vertex point closer than to have ti go along the entire mesh you fucking retard
Its alright, but you preferably want a Nvidia card for 3D work. Nearly all 3D software is developed around Nvidia cards, it's really the only kind the industry uses as a whole. Go for a GTX 660/760/960 or higher if you can.
Is there any good tutorials on creating a basemesh to have correct topology?
All I can find are old tutorials or people who are using blender.
Has anyone got any references on how to get correct shoulder topology?
I'm trying to follow a confusing tutorial.
>correct shoulder topology
make a few disembodied shoulders for simplicity
rig it, see where it deforms worst
add edge loops in the correct orientation
best solution is the one that still looks like a shoulder after you deform it in a drastic direction.
Just downloaded an iphone 6 obj that I have to destroy with a bullet hit by the end of the month for a short film.
The mesh is fine for still images but royally fucked for dynamics stuff. Bullet crashes maya and ncloth is really slow. What options do I have ? will houdini take it ?
>browse random portfolios from my country
>mainly just architecture visualization
how does clients find this shit acceptable? it was actually paid for
Also its budget
If they're luxury condos, yeah, you may want photoreal, but if they're middle class apartments or housing just showing off what it looks like in 3D space
(Not that the design in the picture is good either, but still)