Vellum Simulation for Beginners in Houdini
ฝัง
- เผยแพร่เมื่อ 2 ส.ค. 2024
- In this video I demonstrate how to setup Vellum simulation and how to inflate a vellum object in the middle of the simulation.
Download Houdini Files:
www.bubblepins.com/blog/vellum-simulation-cloth-sim-in-houdini
SOP Solver for Beginners in Houdini
• SOP Solver for Beginne...
Procedural Modeling for Beginners in Houdini
• Procedural Modeling fo...
#houdini
#sidefx
#physicalsimulation
One of the best Houdini channels around, keep the amazing work!
Thanks!
Super! In depth and great as always, thanks so much for the thorough introduction!
Thanks for watching! If you have any ideas for next topics, leave it in the comments. I might not get to it right away, but it'll be added to the list.
You explain everything and you don't go fast over the stuff that my slow mind can't handle, great tutorial! thanks!
Thanks for the comment!
Thank you bubble pins!! I like your tutorials.
Awesome content! Subbed on the first video I saw! now I need to binge-watch the rest!
Thanks for the sub!!
The best channel I found here. Thank you so much for your time, and effort.
Thanks for watching!
Thank you for your work on this video! You have just beautiful teaching skills, I think, you can be one of Side FX team really easy. Video was organised really great, I've found no place to cling to) Thank you for knowledge, that you're giving to us in so accessible way!
I wish!!! The would be like a dream come true.
Thanks for watching!
@@bubblepins May be I'll say nothing new but you can just choose one of way that you like in job list and apply for job. I saw one of responsibilities is a "Developing learning materials, such as written or video-based tutorials" and Required Skills that you're suitable so well. I saw it here sidefx.bamboohr.com/jobs/view.php?id=17 and here www.sidefx.com/careers/
You are the BEST Teacher in Houdini Lesson,Very detailed and clearly teaching style.
Thanks!
Nice clear and concise - thank you.
Thanks for watching!
So clear , thanks
Thank you for the mountain of knowledge!
Thanks for watching!
Great tutorial, thanks!
Thanks!
So good and useful. Thank you!
Thanks for the awesome comment!
Excellent tute. Thank you a bunch!
Thanks for watching!
I swear I didnot find like your way of teaching all over the internet
again :
for(int i=0; i
You're welcome a million times! Thanks for the super generous comment!
Aw, yes, just what I need, thank you
Thank you for the comment!
great tutorial. Thanks!
Thanks!
Great effort, thank you for enlighten me.
Thanks for the awesome comment!
Awesome video. Very nice presentation and crystal clear voice 👍👍
Thanks so much!
Thank you for these.
Thanks for the comment!
So direct, organized, clean.
The boxes you use to explain are great, very didatic, like all your other videos.
Do you work in some VFX house production? You seems like a pro and the dedication to make these videos...
Thanks a LOT and keep it up!
I wish! I've been a programmer for the last few years of my life and wanted to change my career. I started learning Houdini a couple years ago on and off. As of last year I've been grinding myself a little more focused on Houdini. I started doing these TH-cam tutorials around few months ago I think when a sub made a request for a topic. The feedback was very good and I started doing more. I super appreciate everyone giving all their time to watch my videos!
I might take this even further one day. I always dreamed of making a platform where everyone can work together remotely on vfx projects. And over time, these small groups of people can become mini-studios. Imagine the possibilities! Because it takes a lot of work to do 10sec of vfx, but imagine if we have 5 people doing a small part, we would have a very awesome vfx scene very easily.
But I've also been in group projects and I know the challenges of working with strangers, sometimes it doesn't work out. Has anyone seen that video by Jaiden on group projects? That's totally what I mean. But I wondered if I can make a platform that can tackle these challenges.
Sorry for the rant. I've been thinking about this for some time now.
@@bubblepins Nice, as a programmer, Houdini is four you, to get the best of it.
About the professional change, I understand you, theres a whole universe of photography, color, 3D especifications that make a lot of people give up.
I think facebook groups already do this, putting people together, and also live meetings.
But your idea is very very nice, imagine the artists has some score system, that others can see and choose from it.
Theres the size of the files too, like some caches, and the time thats use to transfer... so via internet can make it less feasible, more risky, even more for big studios.
Just trying to help you in this thinking...
No need for sorry, is always nice to talk with intelligent people.
Thanks for the answer and your videos! :)
@@fndpires You totally get me. Coding was very boring, but Houdini is so much more interesting! Thanks for listening to my idea. facebook groups, I didn't know that people were already doing this, cause I don't use facebook. My platform idea is a little raw and needs more thought into it.
Thanks again for listening and all the input you have given me!
@@bubblepins Yea, facebook groups are a thing, i participate in like 5, all just about Houdini and jobs related.
My pleasure to talk with you.
All the best, cya! :)
I'm falling in love with your voice
She appreciate it, really, but if everyone writes it under the every video, she just change it! So, let's be smart;)
@@yaroslavhoncharenko106 LOL Don't worry I'm not going to change my voice. I really do appreciate all the comments! Thank you all for spending the time to watch the videos!
Thank you!!
Thanks for watching!
Really nice intro to vellum. Also i am only noticing now that you've joined the 4k monitor club. Good on ya' . Makes reading geo spreadsheet easier :D .
Yeah it comes in soooo handy for viewing a lot of data on the monitor.
Sadly I can't record the videos at 4k rez that would be awesome! It slows down my PC to a point where it affects Houdini. The monitor was on sale at the time and my old monitor was giving up, so I thought why not.
@@bubblepins what gpu and recording software you using?
@@bubblepins but good call on picking up the 4k when u had the chance.
@@psychedlicavfx6646 I have a rtx 2080 (not a titan) and I just use the nvidia screen recording that comes with the driver
@@bubblepins I soooo knew it. RTX architecture is generally decent but nvidia's driver and software support is still lacking. I would highly suggest that you switch to OBS 64-bit (free), most people know it as streaming only software but its an equally good screen-recorder and extremely efficient and easy to use. I have posted the settings I use to record my tuts here imgur . com/a/0avXqyF (remove spaces). It records with enough quality for visibly lossless edits in Premiere. Hope to see your next tut recorded on OBS :) And keep up the good work!
Great, just found your channel, amazing tutorials! Do you know how to do soft body sims on animated objects, for ex. a character with mocap which reacts as softbody to its own movements? Thank you!
You mean like FEM? I haven't looked into FEM yet. Would be interesting, it's going to be added onto my list of topics, but it's a long list.
bubble pins yeah, maybe FEM, I’m a beginner and don’t know very much. Does Fem support that kind of deformations (mocap+softbody altogether)?
Super helpful, thanks for the video.
Would appreciate if you make more tutorials on houdini
Thank you again:)
THanks for the comment!
I've started doing a lot of Live Stream tutorials instead of these heavily edited versions of tutorials. It allows me to get more video content out quicker and I offer heavily edited content to members only, but this is temporary. Until I can find a way to automate or procedurize some of the heavy tasks of the editing process... I even started looking into OpenCV, which seems a little futile for UI computer vision. Everything is work in progress.
@@bubblepins
That's great how can I join your Live stream.
Please do let me know.
Thanks for the reply :)
@@kghosh30 Check my Community Tab often, because that's where I post the time/date for the Live Streams!
I have 2 types of Live Streams that I host, I will post a message on my Community Tab before each Live Stream. I have tutorial Live Streams that is more planned and detailed tutorial demonstrate specific techniques that I announce a couple days before.
The second type is Casual Live Streams where I just do unplanned unscripted Q&A from the public and show case some stuff I am currently working on. The Casual Live Stream is announced just a couple hours ahead of time also on my Community Tab.
I had a dream about a very Vellumous snake lol. It was green and flat, then it inflated when I got near it. Good job it was just a dream, there is no Anti-Vellum.
I was just picturing that in my head with inflating teeth! It would look scary until the teeth starts to inflate.
You have a nice voice!
Thanks so much!
Thank you for the tutorial!
I am trying to create the simular simulation with another geometry that I created. Inflation works well but the geometry bounces so hight when it touches the ground. Do you have any tip how can I reduce / get rid of the bounce?
You can try the Velocity Damping parameter in the Vellum Solver node, it's under the Forces tab. See if that helps!
Can you get into the Presure and weld constraint . For stitching together pieces of materials to inflate them .
I actually just finished a mini-series on Vellum Clothing this past month. I weld the clothing patches together and all, but I haven't gone over Pressure yet.
A valuable tip for anyone who wants to inflate a complicated model: the best way to get good physics is to make an object deflate and then retime this simulation by reversing the timeline. That way, you keep 100% of the object's original shape. This can be used in any other 3D software
Good tip! this can also be done in the editing software too.
exactly. I recently switched from cinema 4d to houdini, because I needed to use more powerful software in the company I work for. Your videos have helped me a lot, you explain clearly, I love that. :3
@@vertexiom awesome! That's so great! Thanks for letting me know! I get very happy hearing how this helped u professionally!
At 34:05 when you subdivided the Taurus before you baked the sim, it looked awesome with all the detail, but indeed this was very slow. Is there a way to optimize it, so you will get the same result at a faster speed?
You can use the vellum post processing which can do the subdivisions much more efficiently, and happens after the vellum sim. I threw down the subdivide near the end of the video to show why not to use it lol. And to demonstrate the difference between using subdivide and vellum post-processing. Because post-processing happens after the sim, it doesn't affect the sim. But the subdivide is different in many ways, it's slower and very inefficient, and it also affects the outcome of the simulation compared to when you use the vellum post processing.
This could be useful for someone that's looking for a intense accurate visual effect of a high poly torus, however for simple scenes I think you will be better off using the vellum post-process node.
Hope this helps!
@@bubblepins Thank you!
Great tutorial. The only minor complain, sometimes you switch to the next node too fast and it's not possible to see the last parameter's value you've set on the previous node. Could you also make a tutorial on the sound effects you have in your example videos? They are fantastically in sync with the visuals and I'd like to learn any specific technic for this. Also, thanks for not putting background music on your video. Thanks for your great work!
Thanks for the tip, I'll keep that in mind of the next tutorial! The sound effects are painfully done frame by frame, but luckily the animations are just a few seconds so it's not too bad. I use adobe premiere and that allows me to scroll by the video frame by frame and even stretch the audio to make it fit! LOL I could try making a blog post maybe, because I"m not sure if there's enough to make a video tutorial on, but let me try and see!
@@bubblepins I wasn't expecting that and was thinking there is some Houdini magic involved. Now, I'm thinking it would be actually cool to involve some magic and generate midi or even sounds from Houdini. I'll look into it eventually.
@@bubblepins kudos for extra effort with sounds :)
Well explained for noobs
Hi, I am really new to Houdini and been following your tutorials. I've encountered this problem with the warning on vellum solver in this video, written as Open CL exception: cLBuildProgram (-43); it's not the first time I had this problem, the other time when I did FLIP simulation, it pops with the same warning to my other nodes. I have no idea how to solve this as I seek for so many ways to help. Can anyone tell me what's happening here? Seems like it happens to most of the simulations I do, really frustrated...
I think you ran out of VRAM. How much ram does your GPU have? You can try turning off the OpenCL for the solvers, but that would use CPU to calculate things and will be a lot slower.
Hope that helps!
@@bubblepins My graphics is Radeon Pro 580 8GB, is this enough to create dynamics?
@@bubblepins Yeah I actually tried to turn off OpenCL already as was told by the other VFX artists, also set my OpenCL device to be CPU based instead of GPU in Miscellaneous under preferences but not getting any luck :/
@@twodumbocreatives2627 hmmm...
Try using the headlight for the 3D viewport and disable the materials on the 3D viewport. I suspect Houdini is still trying to use the GPU to displace the image onto the 3D viewport which gets updated while you cook the simulation.
This is a total guess though.
Hope that helps!
@@bubblepinsReally appreciated with your effort typing such an explainable message! However I just tried this method but still not working ... hmmm wondering what it indicates by the number -43 for this error, I tried to research it but not getting any answers for it. It seems like this problem can't be solved at the moment after I asked groups of VFX artists and some forums, but put this aside, say if I can't use vellum solver (I know for the really basics, vellum or other types of dynamics need solver to work), which other functions or different ways to replace or do the similar effect as this?
do have tutorial on RBD, pyro, flip ?
Yeah, I try to create a playlist for every category I can think of for all the Houdini tutorials that I have:
RBD Videos: th-cam.com/play/PLqovuMX-teVMgI6FV7AOEP4_t5d9E2J6Y.html
Pyro Videos: th-cam.com/play/PLqovuMX-teVO_wfN_5LE9NSA-9G41Xb9v.html
FLIP/Ocean Videos: th-cam.com/play/PLqovuMX-teVMjNozQVSn8OFclScdK5lUF.html
Or you can check out my animated playlist that may be easier to see the categories (scroll down on the page a little):
bubblepins.com/membership-guide
Hope that helps!
@@bubblepins ♥️♥️
how do you get your final renders to get realistic sound?
It's all thanks to TH-cam's audio library sound effects and some Adobe Premiere used for timing the sound effects and video clips!
@@bubblepins thanks!
Hi bubble,
Im trying to simulte hair using vellum, everything simulate and works fine except attributes I painted in my static mesh (For clumping,etc), It works on the first frames but then It looks like they stay in world space and shifts through the guides till i cant see them anymore...
Any Idea how to fix this? Ive try adding rest to everything i could :c
Thank in advance.
I think I know what's going on, but it's a total guess. I think the paint attribute node needs to be moved up in your network setup. Because it sounds like it's painting the simulated hair. You need to paint the hair before the simulation or right after the hair is created.
Hope that helps!
@@bubblepins Hi bubble,
Thanks for your quick reply, Im actually paiting my attribs before the simulation :( , Im using houdini 19, i was able to make it work in houdini 18.5..
Am I missing something? :C
Thanks
I think i made it work by conecting my defomed mesh in both rest skin and animated skin inputs in the hair gen node.. That is not how it is supposed to work right? 🤔
Thanks!
@@JoaquinPeMM Are you using the guides or vellum solver to move the hair? But either way, what you can do is take the node where you have the paint attributes put a timeshift and freeze it at frame 1 (just in case if its moving, because I'm not sure if it is or not). Then use a point wrangle to transfer the attribute back to the simulated hair with some VEX code like this (make sure the painted attributes is connected to the 2nd input of the wrangle):
float attributeValue = pointattrib(1, "attrName", @ptnum, 0);
f@recreatedAttrName = attributeValue;
This will take all the attributes from the 2nd input of the wrangle and transfer it to the 1st input. It's a bit of forced and not as elegant as I would like, but this should at least get you going.
I did have a quick live stream on hair before, but I havent had a chance to put out a tutorial on vellum hair yet, which should be in the coming topics soon.
You can check out this old live stream on hair for the time being, but it's rather roughly done and it uses hair guides instead of the vellum solver:
th-cam.com/video/05lCOcrNikU/w-d-xo.html
Hope this helps! I've been wanting to do a hair tutorial soon, I'll make it higher on the list.
@@JoaquinPeMM The rest skin should be still, like freeze it on frame 1 with a timeshift node. I'm not sure, but this is how I use it for hair guides, sorry I thought you were using the vellum solver at the beginning.
Hey bubble, how can i inflate torus in DOP,
Do you mean the Dopnet? If I"m not misunderstanding, all the stuff starting 16:14 where I add the forces, you can do the same to the Vellum dopnet by adding the same nodes but hook it up to the Vellum Solver inputs. I believe the Particle Forces Input should do the trick. The Vellum Blend node is also available in the Dopnet.
Hope that helps!
@@bubblepins Yess in Dopnet, I did the torus animation in blendshape and hook it up directly in vellumcloth°geo and °constrain input, and out with 2 nulls Geo and Constrain, in DOP it's not inflating, Also i hook the Vellumrestblend in Vellumsolver>in Particles source, don't know what wrong i m doing 😅, sry for disturbing you
@@bhosaleshubham301 no worries! Let me try it myself in a couple of days later. It's because I'm currently super busy preparing for the Live Stream tomorrow, otherwise I'd just did a quick test myself. Let me get back to you on this in a few days.
very good thanks, but no need for sounds effects in the short videos that appear here and there. They are actually a little annoying :)
awww...
How do we connect vellum soft body to animated character??
You can use the Pin Points parameter to pin down selected points from the geometry to the animation when you enable the checkbox Pin to Animation. This option will be available for Vellum Configure nodes, like Vellum Configure Cloth and so on.
Hope that helps!
pins aight!! I'll try 🙌🏻
@@bhushankandalkar3104 Have Fun!
@@bubblepins hey I did it and i want to simulate only the character head and hand and in houdini 18 there is no paint node what to do then??
@@bhushankandalkar3104 If you only want to simulate the head and hand, you can physically split the geometry before feeding it into the vellum solver. I actually demonstrated this in a recent Live Stream about Vellum Antennas for a bug character.
OR you can use the Groups parameter in the Vellum configure clothing node at the top to turn only the head and hand geometry into a vellum clothing before feeding it all into the vellum solver, I believe that should simulate only the head and hand as vellum clothing.
Hope that helps!
using H 18.5 but my geo is not inflating like your's
Can you describe for me a little what is looks like on your side? Here's a few things to try out too, try increasing the number values to something insanely high just to see if it's affecting the scene at all. This will debug the setup to see if the particular parameter is affecting the geometry at all.
I got the one torus working, but I immediately wanted more fun and tried to duplicate them with CopyToPoints onto the grid, but the fun did not come, unfortunately. I see that those other clones do not get the blendshape data but what I'm missing here, how to pass the blendshape info to all the copies of the toruses?
The exact network would be also nice, but I'm more interested in how to think, what's the logic I'm not getting. Thanks!
www.dropbox.com/s/p119q0459ivk3w2/blendshape_to_copies.png?dl=0
I think I know what's the issue (well a lucky guess). THe blendshape node works by recognizing the geometry by its exact point numbers or by points. There is only 1 torus connected to your blendshape and there are an array of toruses connected to the vellum, it may not work. Try connecting the blendshape after the array of torus. You need to blend them all together or copy again after you blend. Then vellum can recognize this torus to that torus. An exact match. Does it work with just 1 torus? If so, copy all torus, including the blendshape ones.
Hope that helps!
MALAYALI POLI AANU... its not WALOOM "WAY ..LLAM" means water in Malayalam Language
Everything would with Houdini would be a lot of easier if people didn't use the STUPID SOPS VOPS DOPS crap.