520x520

Houdini3D_icon.png

🧵 Once and for all

Anonymous No. 847214

Is Houdini really THAT hard to understand?
I've been avoiding this bitch for a very long time.
I remember watching tuts when I was like 13-14? and said fuck it, and became a cinema4d fag. The rewatched tuts at 20-23,and still didn't pick uptmavh

Anonymous No. 847226

>>847214
Idk, it seems to be endless and I don't have a clear understanding of what is possible with it and what to even use it for. I guess you can do pretty much anything, but good luck with tutorials. The ones I followed though were pretty easy to understand, I guess I could get familiar with it at least in a particular niche relatively easily, especially since I have a background in coding, but I just don't know what to use it for. Since I'm focused on environments there could surely be some useful tools, but nothing that I really need. There aren't many times I had a thought like "man I wish I could automate this thing somehow". I even prefer doing a lot of stuff by hand.

Anonymous No. 847270

>Is Houdini really THAT hard to understand?
no, but also maybe and definitely yes.
if you do decide to try it, you need to commit to it for at least a year.
there will come a day when you'll need to define a transform matrix manually and you'll want to kill yourself, but it's not that hard and you'll get through it.

1918x794

Blade-Runner-168.jpg

Anonymous No. 847273

>>847214
>Is Houdini really THAT hard to understand?
No, and I don't know why people think it is.

On the surface, Houdini might be daunting, just like any industry DCC tool it's not going to easy to master, but it should be easy to learn.
Seriously, go watch a easy tutorial and you'll start to understand that it's not hard to learn at all. There's some aspects of Houdini that might be a bit complex or overly daunting, but all this stuff takes time to learn.

You wont become a Houdini master overnight, but if you spend a good year learning it, then you'll become proficient.

Anonymous No. 847315

>>847214
It's only multivariable calculus and linear algebra.

Anonymous No. 847350

>>847273
>>847270
thx

Anonymous No. 847353

Don't bother with it if you're a simple asset monkey.

Anonymous No. 847374

1 week in blender and you feel like you know what you're doing

1 week in Houdini and you still don't know what 70% of the tools do

Anonymous No. 847375

>>847214
If you have to ask here you are NGMI with Houdini anyway.

Anonymous No. 847377

>>847374
>70%
lol

2684x4096

80s Sigourney is ....jpg

Anonymous No. 847793

>>847214
>>847226
>>847270
>>847273
>>847315

Who's the Pavlovich of Houdini?

Anonymous No. 847805

>>847214
tried out Houdini once at work, it seemed easier to learn than the retarded Blender UI

Anonymous No. 847827

>>847793
doesn't work that way with houdini unfortunately. it's just too big and too much to cover. good news is there are a lot of pavlovich-tier instructors doing houdini stuff at all levels.

>rohan dalvi - official intro to houdini (intro)
>tim van helsdingen - howdini (intro)
>steve knipping - applied houdini series (dynamics)
>matt estella - thecgwiki (essential)

maybe just start with those and see how you fare.

Anonymous No. 847977

>>847827
unfortunately Sidefx is totally incompetent with their own masterclasses on youtube / vimeo, not releasing project files

Anonymous No. 847986

>>847977
not really an issue for me.
if i can't recreate a houdini set up from memory, i probably didn't understand it.

Anonymous No. 847988

>>847977
I am starting to feel like SideFX is botching this deliberately. While they do put some efforts in to make Houdini more approachable, they still keep it at a level where the person must show self-initiative and effort to become part of the club.
Undesirables are getting weeded out. You must show to be worthy by overcoming the hardship of the learning curve yourself or with a proper teacher. It cleanses the user base of mediocrity.
I can understand it, quality over quantity is a good approach to keep their reputation at a high level and the customers of H-artists satisfied.

Anonymous No. 847989

>>847986
Not putting up your files is bush league

Anonymous No. 847990

>>847214
Better question is what Houdini is FOR (besides muh VFX/procedural). So far to me it looked like a collections of systems working together (nodes), but besides aforementioned I’ve seen it used for procedural modelling and that’s about it (there are not as many generalised videos on it as on Blender and even Maya). There is also a kineFX but looks like it’s mostly used for retargeting atm.

Anonymous No. 847991

>>847990
it's nice that you added muh in front of VFX, otherwise i would have thought you were a moron

Anonymous No. 848112

>>847827
Is it worth it for a casual dude to try to learn? Don't have much time between uni, family and work and a basic understanding of 3DS Max. Have a small handful of hours a week to practice/learn

Anonymous No. 848131

>>848112
>Is it worth it for a casual dude to try to learn?
probably only if you really enjoy it and don't have a really hard time with it. as with most things, only way to know that is to try it out. if you find it very difficult and stressful, spend your time elsewhere.

Anonymous No. 848169

What's the correct way to polymodel in this thing? SOP or OBJ level with shelf tools? Polydraw? Every update adds a few neat modeling tools but the overall workflow is a mystery to me.

Anonymous No. 848182

>>848169
i barely do it, but all the nodes get dropped into sops, so if you need the controls they're over there.

you probably want to create nodes by tabbing in the 3d view itself. e.g. you don't get the edge loop option anywhere else (because it's just a macro for polysplit set to edgeloop). setting up some decent shortcuts would probably help, a lot.

Anonymous No. 848212

>>848182
I find the basic operations like edge loop, weld and extrude work well but other essential tools like extract and combine are just broken. Extract just duplicates and doesn't actually delete the geometry from the original object, with any kind of transform anywhere combine will just fuck up the resulting object. And of course the "delete history" option is hidden form the right click menu when you need it, and when you find it and use it on the wrong object it fucks up all the downstream objects dependent on it.

This is my experience working at the OBJ level so I wonder if inside SOPs is actually the way to go.

Anonymous No. 848222

>>848212
unfortunately i don't think going into SOPs is going to help too much. you'll be able to see exactly why it's doing the weird stuff it is doing, but it's going to keep making those dumb choices. you can script your own shelf tools if you're so inclined to try and make it less insane.

you can check out modeller if you really need to keep polymodelling in houdini:
https://alexeyvanzhula.gumroad.com/

Anonymous No. 848223

>try houdini apprentice and its only available renderer, mantra
>create a scene with a single object with a principled shader and textures from substance
>render a 12 frame animation at a low resolution
>over 3 minutes to render and flickers like mad
>setup the same exact scene in blender cycles CPU and render in less than a minute and the frames dont flicker
>try various mantra pixel sampling quality settings but even extremely low quality scenes take multiple minutes.

Even 240x125 resolution at 2x2 or 1x1 sampling takes an eternity compared to blender (cpu). What have i done wrong? People say that mantra is powerful but just slow - am i experiencing it? How can it possibly be this slow at low res?

1282x721

806792744589.png

Anonymous No. 848224

>>848223
unless you need to learn mantra for some reason, just pirate fx and save yourself some pain

Anonymous No. 848226

>>848224
I only have 3.5gb of vram making my gpu useless and crash all the time. My gpu is too old for arnold gpu. I need a cpu renderer but the problem with mantra is that it looks worse in terms of flickering and instability while also taking a lot longer

Anonymous No. 848233

>>848226
arnold and vray cpu are still better options than mantra.
plus with fx you can export alembics to send over to blender.

Anonymous No. 848244

>>848233
Before taking the steps into external renderers I tried Karma. I already found multiple bugs. The IPR just wont refresh after you've set a key on a param, go to another frame and then set another key. I'm interested in Solaris because they say its the present and future. How is renderman?

Anonymous No. 848246

>>848244
Karma is beta.

Redshift has decent Solaris support. Renderman, I haven't tried, but should be the gold example. You can also have a look into AMD ProRender, and 3delight.

Anonymous No. 848267

>>848212
Often times people have a certain way of working in one application, and when they try to translate that workflow into another application it just ends in misery.
I'm having a hard time imagining what you're working on or how you're going about it, maybe you can give a little more info? Do you have a bunch of Objects at the /obj level and are you trying to concurrently do modeling work on them? If so I would not recommend that.

I would work in one SOP network where it's much easier to relate different parts of a larger model to one another, and then break things out into different objects via Object Merges or caching to disk. Again I'm not sure what the use case is but you can accomplish an "extract" by using the Split SOP or the Blast SOP set to Delete Non-Selected.
What do you mean by "delete history"? Instead cache files to disk and read them back in at key points in a modeling network if you don't want to wait for things to cook.
>>848223
Like anon said just pirate it or go Indie, so you can have access to more modern renderers. Mantra is very capable, but you won't get far with it unless you have a renderfarm at your disposal.
>>848244
Karma could be a nice option a little further down the road. The problem with Karma and Solaris is that USD is still very much a moving target. While the scene composition stuff has been stable for a while, Pixar is still making major architecture changes to things like Hydra, lights, materials and things related final rendering. Kind of hard to build something solid on top of that, which I feel explains *some* of the jank in Solaris.

651x786

hdelhistory.jpg

Anonymous No. 848276

>>848267
I'm not having problems with any specific project, they were just general observations about modeling in Houdini. For modeling tasks I still fall back to Maya and import to Houdini as .fbx but ideally I'd like to ditch other DCCs completely.

>Do you have a bunch of Objects at the /obj level and are you trying to concurrently do modeling work on them?

Yes, this is the standard approach in Maya and hopefully in the future Houdini too. Working inside the SOP context doesn't seem ideal, after a few minutes of intense modeling it's very tedious to zoom in and sift through 100 edits and polysplits to find some critical node like some sub-object's pivot. Doing housekeeping by putting dirty edits in subnets etc. is pointless busywork when in the end you'll lock the model's last node and delete everything upstream, you get clean organization for free in other 3d applications by working in the object context and continuously deleting history.

>>848222
Thanks for the suggestion, but I wouldn't learn to model with a plugin made by one guy who might just drop support on a whim or charges you over and over again just to keep the plugin functional in future versions of Houdini.

Anonymous No. 848371

>>848276
Fair enough, I think it's a difference in viewpoints since I'm so used to working in Houdini. Personally I have never had things that would be considered separate objects and needed/wanted to do modeling work on them together, but I know Maya and Softimage artists have inquired about that on the forums.

I do a little bit of modeling in Maya (when it's more convenient to just stay in Maya) but I have no desire to port workflows from there to Houdini. I *like* having a long-ass history of nodes, because I use it and pull stuff from the upstream network often. If I'm trying to organize things I stick them into network boxes (I don't like diving in and out of subnets).

216x174

stressed-rubs-eyes.gif

Anonymous No. 848390

>tab indents lines
>backspace removes one space, not the whole indent
please fix. it's painful every time.

Anonymous No. 848400

>>848390
if you're at the beginning of a line, Shift+Tab removes indentation

Anonymous No. 848445

>>847214

I'd certainly recommend it for anyone that has come up against the maya API when they wanted to script something a bit more complex.
doing complex shit is easy in Houdini.
doing run of the mill stuff is harder than it should be.

I'd love for them to get their rigging and animation tools up to the state of maya then I could use it for the majority of my work and not need to shunt alembics around all day.
Even then I'd always go back to maya or blender if I needed to do simple polymodeling things.

Anonymous No. 848464

>>848445
>doing complex shit is easy in Houdini.
>doing run of the mill stuff is harder than it should be.

This is where USD comes in. Set it up once and create variations.

630x399

1615749357393.jpg

Anonymous No. 848490

>>848445
>get their rigging and animation tools up

888x83

h19.jpg

Anonymous No. 848839

bros..

Anonymous No. 848855

>>848839
no sneak peek yet, though

Anonymous No. 849674

it seems that Labs Substance Archive is getting deprecated. I wonder if they'll introduce something better for Substance files, or just drop it

Anonymous No. 849695

>>849674
Either the Labs guys are releasing the fully live link between Houdini and Substance they teased a while back OR they're shifting the burden of development on Adobe since providing bridges seems to be their new thing (Substance >< Blender).

501x334

210846577185.png

Anonymous No. 849761

>>849674
>>849695
lads, i'm rewiring the existing substance material node to work with arnold and v-ray. i've got both working, but i'm running into one last issue:

i want viewport previews and so i need to keep the existing principled shader wired in.

unfortunately, i can't figure out a way to get display / render flags to work here.
pic-related shows shader in viewport, but does not render it.

if i have display flag on the principled shader, but an output node wired into arnold, it'll render correctly, but break viewport shading.

is there any way to make this work without having to have a toggle-switch?

1164x363

rg.jpg

Anonymous No. 849764

>>849761
I would submit a bug report. Tested with 2 constant materials and your setup works.

Anonymous No. 849765

>>849764
t b h i'm not sure it is a bug.

more likely i'm doing something wrong / specific to the substance hda / intended behaviour inside hdas(?)

also i'm a pirate so i don't make bug reports

Anonymous No. 849770

>>849765
>what is houdini apprentice?

Anonymous No. 849771

>>849770
>what is me being a lazy cunt

Anonymous No. 849772

>>849770
houdini apprentice is useless because it doesnt let you use 3rd party renderer

Anonymous No. 849780

>>849772
>useless
i'm the pirate and even i think that's a bit harsh, m8

Anonymous No. 849790

>>849770
does that let you make houdini for unreal shit

Anonymous No. 849791

>>849790
https://www.sidefx.com/products/compare/

Anonymous No. 849889

>>847793
no such thing I'm afraid but Entagma will get you there

Anonymous No. 849891

Do nodes have a performance cost like in substance designer? doesn't seem like you can model complex stuff without having a mile long list of nodes, wondering if it has big effect on performance.

Anonymous No. 849896

>>849891
You can't do anything without nodes. Nodes are the point of Houdini.
Houdini went though some extensive optimization in that regard but at some point, with a large and complex node network, it will inevitably get slower.

1280x720

test_mantra.jpg

Anonymous No. 849897

>>849896


Sorry I was thinking more like for example, lets say I had box and added a polysplit node then set the number of loops to 10 then I have 2 nodes in the network but I could also have a box and 10 edit nodes each adding a loop cut now I have 11 nodes in the network, it's obviously better to use the more efficient network but is there a performance hit when in the end you end up with the same result and is it even worth taking into account.

Anonymous No. 849915

>>849897
Different guy here. Houdini is pretty efficient when it comes to dealing with large quantities of nodes, but yes, in the end it does have a performance impact in the sense that you are dealing with larger .hip file sizes, longer load and save times, and potentially cooking through longer chains of SOP nodes (if you are not making use of File Cache nodes to cache along the way). Your example was good, and it would be more efficient to do it in one node as opposed to ten if you can achieve the same result.
Generally speaking you would keep your modeling hip file separate, save the result to disk and then load from disk in your layout/animation scene.

Anonymous No. 850655

>>847991
So what is it for then?

Anonymous No. 850659

>>850655
making people with big brains feel like they're better than everyone else

764x551

1629157164646.jpg

Anonymous No. 851203

>>847214
The thing about Houdini is that it's wide as an ocean, deep as an ocean. People seem to think they need to master all of it like they would with Maya or ZBrush, but nobody does this it's just too much for a single person.

Also, to really master it you need to have both artistic and mathematical/engineering skills, needless to say the people who have both of those are very very rare, which is why people keep claiming you need 500IQ for Houdini (which isn't true as mentioned in the paragraph above, unless you want to excel at it, in which case it probably is).

Anonymous No. 851394

What are the go to flip sim tutorials/courses?
I'm mostly interested in small scale advertising things vs gushing rivers / big ocean tanks.

I've already checked out applied houdini liquids 1 and the few entagma ones they have on flips.

Anonymous No. 851395

>>849891
Yes, there is a perfomance cost, however Houdini's nodes are very light and you have more control over how they evaluate. Substance is an absolute dog compared to Houdini.

Anonymous No. 851426

>>851394
For small scale stuff look for tuts that explicitly mention "surface tension", IIRC it's a fairly recent addition to Houdini's FLIP solver and older stuff will not cover it.

Anonymous No. 851564

>>851394
Houdini small scale FLIP sims suck fucking ass. I work fulltime as FX artist at MPC.
You have to do shitton of RnD with every fucking parameter, you won't learn shit from tutorials.
Create or find a hip file with wedging and auto flipbook setup. Do lot of wedges with every parameter to your heart's content. And you'll start to understand every small single parameter drastically a simulation.

Anonymous No. 851573

>>851564
Ah fug, parameter tuning my least favourite thing.

>>851426
Thanks mang.

Anonymous No. 851674

>>851564
what's wedging?

Anonymous No. 851681

>>851674
https://www.sidefx.com/docs/houdini/tops/wedge.html

Anonymous No. 851693

>>851564
We found out 20+ years ago that going bigger scale in sims negates 90% of the problems.

Anonymous No. 852511

Will learning basic python help me with Houdini.

Anonymous No. 852518

What was in the leaked H19 e-mail? Any juicy deets?

Anonymous No. 852542

>>852518
nothing specific, just very vague news that they're improving almost everything

1080x1248

19.jpg

Anonymous No. 852563

>>852518
>Today's beta email was erroneously sent to all Houdini customers. It was meant to be an early access letter for Annual Upgrade Customers who will be getting a chance to work with Houdini 19 before it goes Gold.
>For Indie, Education and non-AUP customers there will be a longer wait as we put the finishing touches on this exciting release. There will be a presentation to introduce you to all of the new features and then you will receive your new licenses once the product is released in late October.
>We apologize for the accidental email. At least you now have some idea of what is coming in the new release and we look forward to getting it into your hands.
Sincerely,
Your SideFX Houdini Team
From reddit but people on other forums got this mail as well it seems.

Anonymous No. 852565

>>852563
I love goemetry

Anonymous No. 852597

>>852511
??? someone plz

Anonymous No. 852676

>>852563
>improvements to modeling and gömetry
I hope it's actual usability improvements and not just a couple of new niche nodes (chain sop lol).

Anonymous No. 852681

>>849891
It has an effect on scene file size (relatively minor) and on scene load and save times. If you take a complex model from start to finish you will end up with a ton of nodes, and cooking that network from start to finish ona fresh scene load would potentially take time.

>>852676
From a recent job posting we know they are wanting to rewrite the viewport in Vulkan. Hopefully that gets rid of some of the buggy viewport behaviors and occasional funkyess with selection highlighting and tool initialization.

Anonymous No. 852690

put axiom in dops you coward

Anonymous No. 852701

>>852681
>From a recent job posting
Not for H19, that's for sure.

1207x920

fuckblender.jpg

Anonymous No. 852782

I made a donut in houdini.

Anonymous No. 852790

bag of new tutorials to keep u boys fed:

HelloLuxx: Houdini In Bloom
https://www.bilibili.com/video/BV1dh411W72


CGMA: Procedural Modelling For Production In Houdini
pt1: https://www.bilibili.com/video/BV1h44y1479H
pt2: https://www.bilibili.com/video/BV1NR4y1p7Za
pt3: https://www.bilibili.com/video/BV1yb4y1y7UZ
pt4: https://www.bilibili.com/video/BV1bh411n7K2
pt5: https://www.bilibili.com/video/BV1A64y187fQ
pt6: https://www.bilibili.com/video/BV1bh411n77F

Anonymous No. 852812

>>852790
oh snap

bloom is not even on cgpeers
thanks houdianon

Anonymous No. 852821

>>852812
it just popped up as a recommendation. nice surprise.
can't find it on the website that's in the watermarks. guessing it's probably in a wechat group or something like that?
make me curious about what else is hidden in the chinesenet

Anonymous No. 852824

>>852790
noice, thanks

Anonymous No. 853036

I'm so sad for you Houdini bros. So much time invested in learning that convoluted byzantine mess, and know Blender introduces geometry mixed and BTFOs poor Houdini. My condolences, bros.
(totally not bait)

Anonymous No. 853621

Rebelway: Math For FX Artists
https://www.bilibili.com/video/BV1tL411s7pd

Anonymous No. 853661

>>853036
learn you VEX boyo, it's good for you.
>>853621
chinese are quick, thanks anon

Anonymous No. 853666

>>853036
>I'm so sad for you Houdini bros.
I'm so sad for you delusional Blendlets.
>So much time invested in learning that convoluted byzantine mess, and know Blender introduces geometry mixed and BTFOs poor Houdini.
implying it does BTFO Houdini is the manifestation of your delusion.
>(totally not bait)
it totally is great bait, it made me laugh out loud and I answered
10/10
On a serious note though: Learning and comprehending Geometry nodes made me also understand some concepts in Houdini I avoided in the past. Learning procedural nodes in Houdini or Blender will train your brain either way, so no matter what you use, you benefit from it.
The first program I learned that used nodes was Substance Designer, after doing so Nuke and every other program that used Nodes appeared easier, it made me loose my fear of nodes and my Brain upgraded.

Anonymous No. 853828

>adam swaab
why do i instantly feel sleepy

501x312

e1410ad62bf8e1f75....jpg

Anonymous No. 854117

>mfw you can start event based timers in solver sops

Anonymous No. 854127

Heres the thing about Houdini that you people don't want to realize : it's good at proceduralism, not simulations. ALL the sims look the same and you can tell rigt away that they're fake.

Anonymous No. 854128

>>854127
Thanks for letting us know, king

Anonymous No. 854460

>>854127
>ALL the sims look the same and you can tell rigt away that they're fake.
What do you mean

Anonymous No. 856187

this thread will stay alive till h19 is out

250x250

1610227837109.gif

Anonymous No. 856360

H19 reveal - 18th 18:00 CET

https://www.viewconference.it/article/565/houdini-hive

Karma XPU - 19th 09:00 CET

https://www.viewconference.it/article/660/karma-xpu-materialx-and-vellum-grains-in-houdini-19

Vellum Fluids - 19th 10:15 CET

https://www.viewconference.it/article/661/using-vellum-fluids-and-crowd-updates-in-houdini-19

Anonymous No. 856363

>>856360
>karma
sucks that my subscription money is being wasted on this

Anonymous No. 856368

>>856363
eh, if it gets usable enough, it might mean one less subscription (Redshift) from me

Anonymous No. 856399

>>856360
>Vellum fluids
Are they just going to jam every single solver ever published into Houdini? I appreciate the variety but now we have:
-POP fluids
-FLIP fluids
-Vellum fluids
-Pyro (technically a fluid solver)

I hope at some point SideFX looks into consolidating solvers, either via robust inter-solver communication nodes (for instance the multisolver fails with RBD + Vellum) or just doing everything in Vellum since XPBD tech is pretty versatile.

Anonymous No. 856402

>>856360
these sound cool too:

Shockwaves Tool & Fire Tornado in Houdini 19
https://www.viewconference.it/article/663/shockwaves-tool-and-fire-tornado-in-houdini-19
"This talk will highlight the process for the Fire Tornado seen in the Sneak Peek, which makes heavy use of the new Volume Deform workflow"

PolySlice Tool + Asset Workflow and File Caching
https://www.viewconference.it/article/665/polyslice-tool-asset-workflow-and-file-caching
"new digital asset creation workflow to help technical artist with they asset management"

Anonymous No. 856431

Just upload the teaser, leafs

Anonymous No. 856454

>>856368
I don't think its reasonable to expect a infantile toy renderer like Karma to compete with established renderers like redshift.

Anonymous No. 856522

They completely dropped unity development and the unreal integration looks like shit

Anonymous No. 856524

My fellow Houdini giganiggachads

https://www.youtube.com/watch?v=FRpkUZmR6ro

650x779

1608129400569.png

Anonymous No. 856525

>>856524
YES

4HGNN

Anonymous No. 856528

>>856524
So modeling still sucks, too bad.

498x498

ryan-gosling-clap....gif

Anonymous No. 856559

>>856524
>Those vellum updates

640x539

5d66f1e5703d8ffa8....jpg

Anonymous No. 856586

>>856524
I'm excited

121x121

download.jpg

Anonymous No. 856588

>>856524
Using Houdini since release 9, can't get exited for anything anymore.

Anonymous No. 856591

>>856588
What path should I follow to master this software? I got into it cause of procedural assets creation for games but the more I use it the more I fucking love it and want to make it my main 3D tool.

Anonymous No. 856592

>>856591
>master houdini
become a technical director (TD)
granted that you are proficient in coding and a bit autistic

Anonymous No. 856614

So happy I started to learn this beast
It's painful but I'm going to make it, you too anons, you too.

393x360

hmmm.jpg

Anonymous No. 856622

>>856592
That doesn't tell me what learning path I should follow

Anonymous No. 856624

>>856622
applied houdini

Anonymous No. 856745

>motion graphic-y karma render
>1m30s/frame @ 1080p
>no hardware specified
why. they tell you what they used in the render straight after.

and that vellum fluids tim van helsdingen render also looks really fuzzy and i can't tell if it's rendered at 720p or if it's just timmy being shit at render and comp

>vellum
>native support for xpdb rigid bodies and plasticity
barely mentioned in the teaser so curious to see how robust this is going to be

>improved vex wrangling ux
yes, good

525x455

3dapu.jpg

Anonymous No. 856884

New enthusiast fag here. Should I bother with the learning curve with Houdini or should I take easier route with Blender?

Anonymous No. 856885

>>856884
Good luck finding work as a blender artist

Anonymous No. 856897

>>856884
houdini is not great if you're new to /3/
start somewhere else and come back in 8-12 months

Anonymous No. 856912

>>856884
If you’re patient yes. I did it and it made other software really easy to get into afterwards. But patience is the key, I’m not a pro with it but if follow some high quality tutorials, you’ll be able to make some cool stuff once you start to learn its node system. But don’t expect to be proficient for a long time

Anonymous No. 856933

>>856912
post specs of your rig

Anonymous No. 856945

>>856884
in either one you need to learn the basics first, normals, UV, edge flow, local and global positions, subdivision modelling etc.

that would be much easier in a traditional poly modeler imo

Anonymous No. 856974

>>856524
I just need PRMan to give an update to support H19 day 1 but I know it wont happen

Anonymous No. 856977

I wish i were smart enough for Houdini

It seems comfy as long as you know how to code and do other difficult stuff

Anonymous No. 856981

>>856933
You don't need a war machine to start getting into Houdini but it's true that once you'll ger further than the basic stuff, your computer will start to scream

Anonymous No. 856986

>>856977

>It seems comfy as long as you know how to code

This is greatly exaggerated I think. There is simple code but you can do a lot of it you don't need to code for simple tasks. Take your time to learn the basics and then the simple code will follow.

But I just started Houdini a week ago and it isn't that bad but I'm coming from a background of Cinema 4D. I just want to use Houdini to do some sims. I understand what people meant by "it will just click" because if you have any experience with After Effects or any other program you will know how it feels to instinctively do simple tasks. Just keep using it when you have the time until the magic moment because it WILL happen.

Also, always start with the basics. Never ever skip to fun-looking tutorials. Wasted too much of my life skipping in Cinema 4D.

Anonymous No. 856997

why is everyone jumping into Houdini now?
especially motion designers

did I waste my time learning c4d?

894x1200

1634097347155.jpg

Anonymous No. 856998

>>856997
>why is everyone jumping into Houdini now?
/stage/ is OP now

Anonymous No. 857032

>>856997
Probably got tired of needing a plugin to do everything in c4d.

Anonymous No. 857133

rigids III 2.0
https://www.bilibili.com/video/BV1Yb4y1h7oX

Anonymous No. 857139

>>857133
>https://www.bilibili.com/video/BV1Yb4y1h7oX
thanks anon, china numba wan

Anonymous No. 857150

>>856986
What tutorials/etc would you recommend to start with?

Anonymous No. 857152

>>857150
try the cgforge and applied houdini ones

Anonymous No. 857164

/3/ is seething that LOPS is first lmao

Anonymous No. 857165

it's up
https://www.youtube.com/watch?v=hTfmtiz9qSI

Anonymous No. 857166

>>857164
I WILL NEVER LEARN /STAGE/
YOU CAN'T MAKE ME

Anonymous No. 857168

>>857166
>using /obj/ for overall scene management
dont do this

Anonymous No. 857169

bro i just want to make some fire, smoke etc for my unreal engine game
don't care about your karma

Anonymous No. 857170

>>857168
I HAVE THREE ELEMENTS IN MY SCENE, BRAD. I DON'T NEED A WHOLE NEW CONTEXT TO MANAGE IT.

Anonymous No. 857171

>>857170
you have to deal with various iterations so you should switch as well as /obj/ being unsuitable for multiple people working on the same scene

Anonymous No. 857174

>>857166
YOU WILL USE USD
AND YOU'LL BE HAPPY

Anonymous No. 857187

why do we still have to wait 9 days
fuck

Anonymous No. 857218

You can't just blueball me by teasing a best-in-class quad remesher. Fingers crossed H19.5 is finally *the* modeling update.

Anonymous No. 857241

>>847214
I started out with 0 knowledge and being a shelf tool fag and within 2 years of self study and using it at previous work managed to land a job using houdini. So it's definitely possible with some determination and the right study materials.

Anonymous No. 857255

>>857187
1-2 days more for a crack to appear
>>857241
are you making 6 figures?

Anonymous No. 857257

>>857218
>*the* modeling update
neverever

Anonymous No. 858374

chinaman too powerful
https://www.youtube.com/watch?v=lVnrrsUWvkw

Anonymous No. 858375

Anyone know a top-tier Discord server of Houdini users?
Also, has anyone ever made some documentations on how to build a city procedurally ? I mean, I can make a procedural terrain, houses, buildings and roads but how to glue that all together for a game?

Anonymous No. 858377

>>858375
think procedural is the big one

Anonymous No. 858429

Just a friendly reminder: Houdini 19 is out TOMORROW (27th).
sidefx.com/download

Anonymous No. 858440

>>858429
i'm a little bit annoyed that they haven't uploaded any of the talks from the view conference yet.

Anonymous No. 858442

>>858440
Yeah, would have been nice material to digest before the launch.

Anonymous No. 858449

>>858440
>>858442
I assume the VIEW conference imposed an embargo, it doesn't make much sense otherwise

Anonymous No. 858538

What is the difference between the Karma MaterialX as mentioned in the 10/19 reveal and MaterialX Lama in renderman?

1208x473

5754236516.jpg

Anonymous No. 858590

it's up

Anonymous No. 858593

it's out

389x395

1441126005.jpg

Anonymous No. 858598

>HoudiniLicenseServer service stopped.
>Reading sesinetd.exe failed. Something is wrong.
>Successfully restarted HoudiniLicenseServer

Anonymous No. 858609

How long do Steam Indie peasants have to wait?

Anonymous No. 858622

>>858598
is the crack out yet?

Anonymous No. 858641

>>858622
nah and no expectation that it'll happen any time soon. XFORCE, who do the crack, haven't been active since april.

3296x4389

MtlX.png

Anonymous No. 858661

>>858538
Honestly I didn't even know Lama was related to MaterialX until the Renderman 24 release. I can't recall Pixar ever mentioning MaterialX prior to the Rman24 launch whenever they would talk about Lama.
But from my limited understanding: MaterialX Lama is basically ILM wrapping a bunch of MaterialX shading functionality into a handful of high-level nodes that they made available to Renderman's shading system. Houdini's MaterialX support looks to be much more granular, more of a "reference" implementation maybe.

Anonymous No. 858695

>>858661
...if you dont get lama, wont your renders be below industry standard?

Anonymous No. 858703

>>858641
did they get busted?

Anonymous No. 858704

>>858703
i'm sure they live on, on the darkweb

Anonymous No. 858786

Why does SideFx not recognize unity as a viable game engine anymore and only support unreal?

Anonymous No. 858824

get in here, it's live
https://www.youtube.com/watch?v=IyWseMa1yZk

Anonymous No. 858825

>>858703
>>858704
just as likely they died of old age.
they've been active for 2 decades.

i think they're also responsible for the current autodesk patcher.

317x123

385172199860.png

Anonymous No. 858832

Anonymous No. 858854

>>858832
Blender derangement syndrome

Anonymous No. 859135

>>856986
>Wasted too much of my life skipping in Cinema 4D.
So much this. Wasted my first year in 3D/Cinema just blindly following advancet tuts, having no idea why and what im doing.

Anonymous No. 859136

>>857032
>>856997
This
Also late C4D updates (r23/4/5) are really lacking. Its like im paying for nothing. Also i feel like a retard building new workstation with 3970x and learning Houdini could somehow justify buying 32cores.

Anonymous No. 859138

>>847214
Shit this thread made me download houdini. Been using procedural tools for a while now and i am used to shader programming and programming in general. Didn't had the time and motivation to learn houdini yet. It's time to step the game up.

Anonymous No. 859143

>>859138
Forget about houdini. The prices are too high.

Anonymous No. 859144

>>859143
It's 400 for a 2 year rental.
That's like 16 dollars a month.
It's one of the better deals out there.

Anonymous No. 859145

>>859144
you can only use that license if you earn under 100k a year with it and just as importantly you can only have 3 licenses concurrently

558x305

flat,800x800,075,f.jpg

Anonymous No. 859146

>>859145
you're bringing in over a 100k and you can't expense an fx license?

Anonymous No. 859147

>>859146
I need multiple licenses

Anonymous No. 859148

>>859147
no you don't. you need to restructure your business and hire back your employees as indie permalances.

Anonymous No. 859161

>>859148
>hire back your employees as indie permalances.
you can't use the Indie license that way

Anonymous No. 859162

>>859161
it's a good thing your freelancers are bringing their own indie licenses with them then isn't it?

Anonymous No. 859203

>>859145
very few earn 100k a year as freelancers, so it's a very good deal

Anonymous No. 859276

>>859162
you can't share Indie Houdini assets with entities that lack the right to use Indie. what you can share is renders, geo caches, that sort of thing, but that's not enough for the scenarios we're discussing.

Anonymous No. 859410

still no crack

Anonymous No. 859453

>>859410
the boys the on the forum cracked marvellous 11 so there's hope.

Anonymous No. 860197

>>859410
still no crack

Anonymous No. 860317

>>859410
>>860197
get a JOB

Anonymous No. 860367

Any place where I can get houdini project files?

Anonymous No. 860372

>>860367
you can look through SideFX and OdForce forums

Anonymous No. 860381

>>860367
rumour has it that if you post a houdini problem matt estella appears a few hours later and drops a hip file with a solution

Anonymous No. 860400

>>860317
still no crack

Anonymous No. 860764

>>860372
>>860381
there are literally 0 project files on there.

Anonymous No. 860769

>>860764
Brainlet filtered.

Anonymous No. 860773

>>860764
https://forums.odforce.net/topic/25173-atoms-link-page/

try not being a retard. youre welcome

541x498

1635991809845.jpg

Anonymous No. 860775

>>860764

Anonymous No. 860802

>>860400
there won't be a crack boyo
>>860773
epic stuff, thanks

678x422

1453683255261.gif

Anonymous No. 860835

Do you guys know some fine repo with some Houdini scripts that could be used by most users? I'm specifically interested in modeling stuff but everything is welcome

453x327

1617114130906.png

Anonymous No. 860837

Are any of you using Karma ?

Anonymous No. 860853

wait, i can just use python 3 in the python 2.7 builds?

Anonymous No. 860901

>>860853
I don't think so.

Anonymous No. 860904

>>860837
Not yet, but I will once its close to feature complete and does render everything. Looks promising so far.

Anonymous No. 861073

>>860837
Just test-renders at the moment but like anon said it looks promising. I was surprised that it supports caustics already.
The nice thing is that it will support things like ocean spectra and other odds and ends that 3rd party renderers typically don't. Plus you neve have to wait for the renderer to catch up to the Houdini build you want to use.
The Karma LOP parameters though are pretty bad from a UI/UX perspective, SideFX should re-evaluate how some of those tabs are organized.

Anonymous No. 861079

>>860904
Sidefx said its feature complete in the keynote.

>>861073
>Plus you neve have to wait for the renderer to catch up to the Houdini build you want to use.
Renderman 24.2 is coming out any day now and will support H19 according to official posts by pixar q&a people on their support forum

1444x812

renderman.jpg

Anonymous No. 861088

>>861079
I thought they said production ready, not feature complete. There are still some important features marked as Planned for Future Release:
https://www.sidefx.com/faq/question/karma-vs-mantra/

>>861079
There is no guarantee that 3rd party renderers will immediately support a new production build or daily build. Also renderman sucks butt.

Anonymous No. 861105

>>860901
ah i'm just getting started with python scripting and naively moved some py3 code over to py2 houdini and it just werked. i guess i just got lucky.

just trying to build myself a little material library python panel.

Anonymous No. 861108

>>861088
>Also renderman sucks butt.
materialx lama nodes justify using it. There is no going back once you start using lama

Anonymous No. 861111

>>861108
>materialx lama
qrd?

1159x801

1634254493538.jpg

Anonymous No. 861281

>>861111
lama gives you simple energy conservation for layered material rendering. It's easy, it's way better than nilla material x, and ILM uses it for everything.

>MaterialX Lama is a state of the art material layering system developed at Industrial Light & Magic that introduces a modular approach to building material networks and includes recent shading R&D work on energy conservation and dispersion.

>The MaterialX Lama project represents ongoing R&D into materials at ILM, and represents one facet of the larger MaterialX project.

Anonymous No. 861336

>>861281
i guess what i mean is, how is this different than energy-preserving PBR materials of literally any other renderer

or is it just that renderman was so far behind its exciting for them to have caught up a bit?

Anonymous No. 861338

>>861336
with traditional materials you only get energy conservation in one material thats hardcoded this way - say, principled in X renderer. What happens when you want a custom setup that goes beyond principled? You're stuck having to code and its hard for the artists to make changes

With Lama, your setups will conserve energy across a vast node setup with the included lama nodes by using lama blending layers. It solves the energy conservation problem without having to write code and the artists can do it. You combine the lama nodes and blend them with the lama blending nodes to go above and beyond what you can achieve with principled in X renderer. It's worth it and you can demo for free with NC...

Anonymous No. 861365

>>861338
got it,thanks anon

Anonymous No. 861369

>>861281
>PBR
>energy conservation
Meme nonsense that means zero in actual production. Our studio (Amsterdam, 24 people) uses Arnold and we still tweak materials on a per-shot base, else most would look like total garbage.
PBR was a nice pipe dream but nothing of it matters in real life production environment.

1280x720

cringe.jpg

Anonymous No. 861376

>>861088
>Also renderman sucks butt
Get a load of this midwit here

Anonymous No. 861397

>>861369
>Meme nonsense that means zero in actual production.
Come on now, shut up with that nonsense.
Yes you still tweak materials on a per-shot base, like it was always done because VFX supes and VFX companies are paid for near perfectionism and the customers are demanding.
But you don't tweak ALL materials or have to build complete different shaders from scratch just because you have to render a scene in day and nighttime scenarios and everything completely breaks if you fiddle around with the light a little.
A good PBR metal shader looks like metal, no matter in what kind of illumination scenario you throw it, compared to the past where it would almost always look like shit and you have to dial in values that are completely arbitrary and removed from reality.

Anonymous No. 861398

>>861397
>dial in values that are completely arbitrary and removed from reality.
Around half of our work involves NPR, so this is largely still the case.
Concerning the "real" PBR stuff; it's often the case that customers demand a "larger than life" look, almost in the realm of hyper realism. In that case we actually DO go beyond physically plausible material parameters.
I'm just a bit weary of industry buzzwords.

Anonymous No. 861404

>>861398
>I'm just a bit weary of industry buzzwords.
Fair enough, but you just went overboard.
When I look at my own development than I can say 100% that before PBR all of my texture/shader work was absolutely crap and I hated doing it, and it wasn't just me, it simply was shit back then.
Since PBR and proper tools (which came simultaneously) are around I have become good, able to do photorealism and I love doing it, it has become piss easy and fast without much guesswork involved.
PBR is an godsend and it took out all the unnecessary, arbitrary and stupid complexity and made the workflow easy so that one can concentrate on the art-side.
None of that applies to NPR though, but it also was never meant to be.

Anonymous No. 861414

>>861404
Whatever you say. Coming from painting and photography, I was perfectly capable of producing photorealistic work with old material workflows, even when I had to write my own Renderman shaders for it. Good day. :)

Anonymous No. 861426

>>861414
>I was perfectly capable of producing photorealistic work with old material workflows
Sure, it was entirely possible but not as easy or efficient.
>even when I had to write my own Renderman shaders for it.
I can't do that and I don't think it should be necessary.
Have a nice day too.

1024x704

1624574772215.png

Anonymous No. 861432

>>861369
> Our studio (Amsterdam, 24 people) uses Arnold and we still tweak materials on a per-shot base, else most would look like total garbage.
sounds like you dont understand PBR workflow


>>861414
>I was perfectly capable of producing photorealistic work with old material workflows, even when I had to write my own Renderman shaders for it.
RSL is really easy and I wrote shaders for it for years, but reyes doesn't really trace too well at all, so the results were never "realistic". Pic related is a "photo realistic" render using reyes found on wikipedia. Note how terrible it looks despite being photoreal (like the work you achieved xD). What improves your rendering for real though is raytracing and the benefits of PBR and lama ;).

Anonymous No. 861490

>>861432
>>861426
Whatever you say, stupid American. Good day neger. :)

Anonymous No. 861491

>>861432
Your displacement is faulty, neger. :)

I don't take advice from stupid American kiddies...and neger worshipers. :)

Anonymous No. 861528

i fucking hate pyside, but i have done what i needed to do.
i just needed to say fuck you to someone because of how dumb this shit was.
fuck you. thank you for reading my blog.

Anonymous No. 861529

>>861490
>>861491
fuck off, Cris

Anonymous No. 861586

>>861529
AmeriNeger :DDDD

Anonymous No. 861620

>>861586
I'm not American. Go play with mirrors, Indian.

Anonymous No. 861656

Is Houdini 19 uncrackable now?

Anonymous No. 861660

>>861656
give it time, I think with 18 it also took a while

Anonymous No. 861666

>>861656
the keygen probably just needs an update to target new license server version.

only problem? the group that makes the keygen is inactive. so someone needs to reverse the keygen.

Anonymous No. 861817

>>861656
>>861660
>>861666
houdini 18 took 6 months to get cracked

Anonymous No. 861828

>>861620
I'm Dutch, negermutt.

Anonymous No. 861834

>>861828
Sure you are, Ricardo.

Anonymous No. 862842

>>861656
whats the point on cracking it?

Unless you really need the full license, just use apprentice.

670x720

houdonut.jpg

Anonymous No. 862855

ive started trying to learn houdini. made a donut from a tutorial. baby steps lol. I really want to make a building generator but I know I have a lot to learn.

Anonymous No. 862888

>>862842
can't use third party renders

Anonymous No. 862889

>>862842
Have fun rendering a complex project on a single node with Mantra, kek

Anonymous No. 862892

>>862888
>cant use redshift
>cant export fbx or alembic
>can only use usdnc
>watermarks, making sprites unusable
>720p

Anonymous No. 862893

>>862889
>>862888
whos going to be rendering inside houdini? Im willing to bet $100 you aint going to be using it for anything outside of tutorial work.

Anonymous No. 862894

>>862893
Karma is quite good.

Anonymous No. 862895

>>862892
>redSHIT
>export as obj
>who cares
>make sprites in any other program. simple.
>720p is good enough for test renders, and thats all you're going to be doing with it.

862x447

955167148723.png

Anonymous No. 862897

piracychads win again

Anonymous No. 862898

>>862895
trollpost / 10

Anonymous No. 862899

>>862898
no rebuttal / 10

Anonymous No. 862900

>>862899
YOU have no rebutall, troll. You have to use redshift. YOUR rebutall was "redshit"

LMAO

0/10

Anonymous No. 862901

>>862900
>>862899
>>862898
break it up lads, the chinks have cracked it.

Anonymous No. 862906

>>862900
still no counterargument

redshift is god awful, i'd rather use VRAY.

Anonymous No. 862908

>>862906
You are throwing your time away on what can easily be made up in post. You are dumb.

https://www.youtube.com/watch?v=myg-VbapLno

Anonymous No. 862910

>>862908
>Actually has the audacity to post this blendlet loving fag
jeusus fucking christ, get off this board NOW

Anonymous No. 862911

>>862910
>no counterargument

Anonymous No. 862912

>>847270
Are transform matrices really that hard to understand? It's literally just a very shortened form of a linear equation system. In fact you don't even have to know linear equation systems at all, just read up on how transforms get put into e.g. a rotation matrix and do exactly what you're told to do.

Anonymous No. 862913

>>862908
I have to remind you, posting Blender Guru on a Houdini thread is against the law.

Anonymous No. 862915

>>862913
hurr durr slowdini doesnt accept third party renderers

hurr durr use mantra

lmfao

Anonymous No. 862916

>>862915
What are you babbling about? Of course that you can use third-party engines with Houdini.

Anonymous No. 862926

>>862897
oh shit, the chinaman came through, Xièxiè nǐ

looks like xforce got beat, they must be american

Anonymous No. 862931

>>862926
didn't they just modify the XFORCE crack? also apparently uploading that is against the rules on CGPricks

Anonymous No. 862935

>>862911
heres my counterargument: my cock

Anonymous No. 862940

>>862931
CGPricks just deleted it

Anonymous No. 862942

I just wanted me summomo updated animation tools, good thing they themself said that they are working on them. Well then, more time to get into Houdini.

Anonymous No. 862954

>>862935
my cock is larger. your argument is invalid

Anonymous No. 862966

>>862926
>>862931
>>862940
i kinda wanna join this chinaman's qq group now to see what else they've got

is there anyway to get into those without downloading an app?

510x600

file.png

Anonymous No. 863094

>>862954
nuh uh

mines bigger.

now run along

1466x1964

h19karma.png

Anonymous No. 863712

What do you guys think of Karma on H19? Still a toy or actually viable?

1280x720

baking a donut.webm

Anonymous No. 864223

Anonymous No. 864240

>>848224
Is redshift that fast?

Anonymous No. 864326

>>864240
Yes, biased renderers usually do that

Anonymous No. 864327

>>864223
source, please?

Anonymous No. 864427

is the group selection on attribute adjust float and attribute adjust vector completely broken on 19.0.383 or is it just me

Anonymous No. 864969

>>864327
https://www.youtube.com/watch?v=pkaJKZpDG5E
https://www.youtube.com/watch?v=nDTkjbEwdOM
https://www.youtube.com/watch?v=oPXyWsmBnD4
https://www.youtube.com/watch?v=uhFix4DsZB8

Anonymous No. 865015

>>864240
Yes, Redshift is based as fuck and nothing else compares despite what the Cycles and VRay shills say.

But this case is a little biased (no pun intended), Redshift shines really hard on volumetrics. It's not going to have a lead that massive in every use case.

Anonymous No. 865016

>>864969
thanks

Anonymous No. 865024

>>865015
in fairness to those other renderers, that screencap is from a 3 year old video. i don't think arnold was even running on gpu then.
redshift is still probably the fastest for most situations tho.

Anonymous No. 865115

>>865015
I take VRay over Redshift any day.