Nice tutorial! Cheating was great)) I really like your attention for details in every video, yo're explaining just everything. And your support on this channel so awesome! It's quite hart to find person to ask. Thank you for that! I wish you all the best!
It’s low res at the beginning because it’s a longer video and it takes some time for the server to process. Just be a little patient people :) Awesome content Bubble Pins :)
I started a UA-cam Membership you can join. I offer exclusive content for members, but its not long tutorials, it's more like Tips and Trick videos that are like 2-4min long and Live Streams, but most of my members watch the replays of the Live Streams. ua-cam.com/channels/gSL2gPQragMHtVh3f2YUQA.htmljoin Thanks for the support!
Your RBD solver is trying to do convex shape vs. arbitrary mesh or VDB detection. It probably doesn't handle edge collisions well. You might have better results if the RBD pieces had a higher vertex density, especially on the interior regions. The detection against VDB is probably point based.
Another great tutorial. Thank you so much. About the video resolution... I have noticed that UA-cam loads the different versions of your high res video one by one. If I watch your video as soon as I get the notification from UA-cam, only the low res versions are available. Consequently, getting something better than 180 or 360 is a matter of time.
Thanks! I need to be more careful next time. Maybe I'll try uploading it ahead of time and releasing it 1 hour after it's uploaded, I think that should give UA-cam enough time to process the video.
@@pablocastelo7090 Patience was never my strength lol but I'm glad everyone jumps to watch my videos! I super appreciate it and everyone that takes the time to watch my videos, even though my videos are quite long.
The rbd solver has been already for a few versions of Houdini. hmmm RBD solvers can only be added inside dopnet nodes, with the exception of the new RBD Tool node which just confuses things. The Rbd solver used in this video can only be added inside a dopnet node. Hope this helps!
At 15:56 I dropped down a SOP network node and then I double clicked that node to go inside that node, that's why you see a new network. Hope that helps!
First: With this way, seperate to two dop, we don't need phase 1. Second: We can do in only one DOP. Don't need phase 3. If you want two object iteraction reality. Only work with POP grains and using Multi Solver so far. That's my idea. Hope anyone share the way to make two dop iteraction really work!
THat is true, I did try using POP too just to test at the time. If I remember correctly the POP does work with the Multi solver and vellum solver, so you can use that. I didn't try using POP grains though. But POP grains is slightly older and Vellum grains does have those newer fancy features like clumping effects for snow. I do find that Vellum grains is very very fast, I don't have anything to compare it to, because I don't work with POP grains that much. You do make an interesting point, if we could make different dop networks work together that would bring a lot of new possibilities. I have no idea how to approach that.
this is interesting, but... what about the actual interactions? it have to be some way to use the mutual relationships, but i still cant figure out how exactly it works. anyway thanx 4 the tutorial, this is a tricky approach
the issue is Vellum is a lot newer than RBD sims in Houdini and I believe they way they are implemented is different. This is going by some of the random bits I watched from SideFX's webinars. I think in the near future SideFX will make the Vellum work with the RBD using the multisolver, this is a total guess though. The approach shown in the video is a little more advanced. Try doing just the first two phases and I think that will really help understand the workflow, before adding the 3rd phase.
I'm planning on making some videos soon on the new Houdini 19 Vellum Unified Solver which allows us to do vellum, flip particles, grains, more all in one solver! Soon! It's on my to do list!
Good evening bubble pins. There is a scene in a curtain with simulations of particles, liquids, terrain, etc., all cached. I want to transfer everything to 3ds max, add materials and render in Corona. How to do it better? Using the Houdini Engine? Why in 3ds max, because the same object (quite simple) in Houdini Redschift renders 4 minutes, and in Max Corona renders 1 minute. I have two minutes of animation in Houdini and this is 2880 frames, so I want to render in Max render. Thank you in advance.
WOW Max Corona is so fast! Now I gotta check the rendering engine out when I get some time. To answer your question, I used 3d max 2 years ago for 2 weeks and gave up because that software was way too expensive for me. I have exported many things to Blender. If 3d max is similar to Blender, I would try to export it as FBX, but FBX doesn't always work, especially with animations. Then if FBX doesn't work, I would try exporting as Alembic. In my experience, Alembic export aren't very efficient, they cache the entire mesh and every point and every point transform as a new point in the next frame. So you lose all the sim data like velocity, force, impulse, etc. And you need all that for rendering. One way you can try is to attach all these attribute data into the Color and use that as a medium to transfer the data back and forth.
As of now, I don't know of any. The main thing is that the multisolver doesn't work with the vellum and rbd or maybe I just don't know how to get it to work together. Hope that helps!
@@bubblepins i put cloth on my car , and created a scene in mountains, to reveal it , by dragging the cloth backward? how to do that? i am sharing my hip file too, forums.odforce.net/topic/48539-vellum-cloth/ here u can check my hip file , if u want :)
Not that I know of. This is the way I came up with because the multi-solver didn't work for me. Usually, in order to combine different types of simulations, you would use the multi-solver, but I couldn't get that to work with vellum & rbd. This multi-phase workflow can be logically simplified a little bit more, but with the cost of more nodes to setup. Try this. Set up the Vellum sim. Then file cache it. Then set up a rigid body sim separately like how you do it normally in a different geo node. Then just do the same thing in the video, throw down a static object node inside the rigid body simulation and just link it to the separate Vellum sim file cache. This way involves more nodes to setup, but it's easier to understand.
Interesting. Do you know how would i go to instance (multiple different) vellum objects? And after that, it would be good to replace each with proxy object?
hmmm.. I think you can use the merge to combine different objects, but give them a different group name and when you apply the vellum constraints you can filter it by group name so the constraints are only applied to certain objects. Say you wanted object A to be have more stiffness than object B. As for the proxy, I think proxy is always a good way to optimize the scene and speed up the cooking performance. You can try what I did in the last Live Stream on Vellum Skin, I used a point deform to transfer from the proxy Vellum Solver object back to the original object. Hope that helps!
@@bubblepins I haven't checked the last stream, so definitely will! By the way, yes, in constraints, the object affected can be set by choosing with a group. But i was rather thinking on different objects as different meshes. Imagine having 2 types of flower, one Orchid flower and a Rose flower...so two different (like a box and sphere). So by this, i was thinking of instancing them on a grid so that i can simulate bunch of them falling through the air. For that, i believe 3 things are needed. 1. To be scattered through a dop net, probably with scatter to points in vellum source (not continuous), 2. To be scattered (1.) but as different objects, for example two flowers (or two basic prim. box and sphere), that are distributed randomly on the scattered points. 3. And to be able to correctly replace these simulated objects (that by now come out from a simulation DOP net, and have been scattered on a grid) with the proxy meshes (for example, you mentioned with the point deform). My guess is with a copytopoints, but can't get it to work currently. It seems straightforward, but i had problems with the setup, mainly with step 2. and 3. Im going to see the livestream...
@@NextroneOfficial You have an interesting idea there! It sort of reminds me of a scrap paper scene I did for my Members-Only videos. I used the Vellum Solver to model some scrap paper and have the scrap paper dumped into a garbage can using the Vellum Solver so it would end up all over the ground near the garbage can. The thing is my scrap paper scene was very light, because all the paper were just plane 2D grid at low poly count. As for the petals for flowers, I think the biggest challenge may be poly count and performance. The copy to points may not be too suitable for animation, but who knows I might be wrong I'll need to test it, but copy to points can definitely be used for the starting position. Proxy objects used during animation, I would stick with point deform. If I get time after I finish off the Vellum Clothing mini series I'll try the petal scene out!
@@bubblepins Hey! I looked through the stream. I believe the way you approach making tutorials is unique, and that one can really enjoy all the bits and pieces of the explanation that you carefully try to convey to the audience. I haven't find the particular case in it (i think i haven't missed it). Right, but it's a bit different over here, scrap paper is a plane and flowers are closed 3D objects :-) Also, i do not intend to add separate vellum constraints to the petals...just a vellum per whole flower is totally fine because they would merely collide while flying, and i think it wouldn't make a difference since there would be many of the instanced and different ones and the eye wouldn't pick the difference in realism i think.. so no ground plane would be included in a shot, just the fall through. Anyway, it's a bit tricky setup.
@@NextroneOfficial The scrap paper scene is in a different video. Have you considered using textures for the petals instead of a 3D objects? This will speed up the performance, and collisions should still work because you can add Vellum thickness to the Vellum Solver. But I'm interested in your idea, because I always wanted to procedural model flower petals. This could be very cool for 3D printing them out too!
The ball just drops from free fall (affected by the gravity force) and I pinned all four corners of the cloth so it stays in position. Do you have a particular issue you're having trouble with?
@@bubblepins I have rigid body solver setup for free fall of the ball and vellum solver for cloth and both solvers are plugged into the merge with Mutual relationship the ball kinda interacts but it weirdly passes through tearing the cloth in a weird way .........If u can provide a file of the setup it would be a great help as I am working on a similar project and I can just replace my model with yours .....Thanks for your help!
@@gameswamy1465 The thing is Vellum object doesn't really work with the rbd object in the dopnet with the mutual relationship or even with the multi-solver. Or at least I never got it working that way. I did both simulations separately even with the sphere dropping onto the cloth, using the technique I showed in this video. I know there are a lot of comments on below here saying that they want a "true" interaction, but Vellum isn't designed the same way RBD solver was designed. This may change in future Houdini versions. For now, the technique in this video was what I came up with as a work around for anyone that needed this. Hope this helps!
@@bubblepins I found a workaround which worked for me .........I solved the sphere as a VELLUM SOFT BODY and not RBD and changed the settings for the sphere that it wont deform much.... And increased mass etc.......so this works if we don't want fracture ......I think for fracture RBD we have to use your way .....but if just and interaction in this simple free fall this was the workaround I came up with .........Thank for you help....👍👍👍
![[UA] NAVI проти G2 Esports | Blast Premier Fall Final](http://i.ytimg.com/vi/QWlIm4FGESQ/mqdefault.jpg)
Need a "TRUE" interaction
This video was too amazing that I need to sleep to cook and process it while sleeping. Jokes aside, wow that's mindblowing!!
Thanks!
Thanks. I really wanted to know how vellum and RBD interact. Your tutorial is amazing~
Nice tutorial! Cheating was great)) I really like your attention for details in every video, yo're explaining just everything. And your support on this channel so awesome! It's quite hart to find person to ask. Thank you for that! I wish you all the best!
Amazing work. You make Houdini look less intimidating. Thank you:)
Thanks!
This is amazing like you read my mind, needed this so much!
Thanks for watching! Hope this helps you.
It’s low res at the beginning because it’s a longer video and it takes some time for the server to process. Just be a little patient people :) Awesome content Bubble Pins :)
Thanks for watching! I'll probably wait an hr after uploading next time to give UA-cam time to process the video and avoid confusion next time.
You don't have to use Vellum source, you can only use Vellum Object it has identical settings as the vellum source.
to quote a stranger from a Facebook group: "bubble pins is awesome!"
Thanks so much!
Hey can i suport you in any way? these are the best tutorials i have found so far. becouse you actually explain what you do. THXXXXXXXX
I started a UA-cam Membership you can join. I offer exclusive content for members, but its not long tutorials, it's more like Tips and Trick videos that are like 2-4min long and Live Streams, but most of my members watch the replays of the Live Streams.
ua-cam.com/channels/gSL2gPQragMHtVh3f2YUQA.htmljoin
Thanks for the support!
Your RBD solver is trying to do convex shape vs. arbitrary mesh or VDB detection. It probably doesn't handle edge collisions well. You might have better results if the RBD pieces had a higher vertex density, especially on the interior regions. The detection against VDB is probably point based.
Another great tutorial. Thank you so much. About the video resolution... I have noticed that UA-cam loads the different versions of your high res video one by one. If I watch your video as soon as I get the notification from UA-cam, only the low res versions are available. Consequently, getting something better than 180 or 360 is a matter of time.
Thanks! I need to be more careful next time. Maybe I'll try uploading it ahead of time and releasing it 1 hour after it's uploaded, I think that should give UA-cam enough time to process the video.
@@bubblepins Yes, that is a possibility. Another option is to be patient as a viewer. :-)
@@pablocastelo7090 Patience was never my strength lol but I'm glad everyone jumps to watch my videos! I super appreciate it and everyone that takes the time to watch my videos, even though my videos are quite long.
@@bubblepins Well, obviously, everybody is so eager to watch your videos that can´t wait for them. hehehe. I can´t blame them.
@@pablocastelo7090 Awesome!
great video, thx
Thanks for watching!
Hi to all
bubble pins - incredible lessons
I have a question, how to add Rigid Body solver houdini 17.5.173?
The rbd solver has been already for a few versions of Houdini. hmmm RBD solvers can only be added inside dopnet nodes, with the exception of the new RBD Tool node which just confuses things. The Rbd solver used in this video can only be added inside a dopnet node.
Hope this helps!
I have already tried, and indeed, inside a dopnet node.
thank you very much for saving me)
16:40 you cut the video, and skipped few steps. I could not make it after that time.. Why?....
At 15:56 I dropped down a SOP network node and then I double clicked that node to go inside that node, that's why you see a new network.
Hope that helps!
First: With this way, seperate to two dop, we don't need phase 1.
Second: We can do in only one DOP. Don't need phase 3.
If you want two object iteraction reality. Only work with POP grains and using Multi Solver so far.
That's my idea. Hope anyone share the way to make two dop iteraction really work!
THat is true, I did try using POP too just to test at the time. If I remember correctly the POP does work with the Multi solver and vellum solver, so you can use that. I didn't try using POP grains though. But POP grains is slightly older and Vellum grains does have those newer fancy features like clumping effects for snow. I do find that Vellum grains is very very fast, I don't have anything to compare it to, because I don't work with POP grains that much.
You do make an interesting point, if we could make different dop networks work together that would bring a lot of new possibilities. I have no idea how to approach that.
this is interesting, but... what about the actual interactions? it have to be some way to use the mutual relationships, but i still cant figure out how exactly it works.
anyway thanx 4 the tutorial, this is a tricky approach
the issue is Vellum is a lot newer than RBD sims in Houdini and I believe they way they are implemented is different. This is going by some of the random bits I watched from SideFX's webinars. I think in the near future SideFX will make the Vellum work with the RBD using the multisolver, this is a total guess though.
The approach shown in the video is a little more advanced. Try doing just the first two phases and I think that will really help understand the workflow, before adding the 3rd phase.
if you need true interaction, try to imitate your rigid body by glued grains
Do you have vellum hair and flip interaction tutorials? Please teach me
I'm planning on making some videos soon on the new Houdini 19 Vellum Unified Solver which allows us to do vellum, flip particles, grains, more all in one solver! Soon! It's on my to do list!
@@bubblepins thanks for sharing! ❤️
Good evening bubble pins. There is a scene in a curtain with simulations of particles, liquids, terrain, etc., all cached. I want to transfer everything to 3ds max, add materials and render in Corona. How to do it better? Using the Houdini Engine? Why in 3ds max, because the same object (quite simple) in Houdini Redschift renders 4 minutes, and in Max Corona renders 1 minute. I have two minutes of animation in Houdini and this is 2880 frames, so I want to render in Max render. Thank you in advance.
WOW Max Corona is so fast! Now I gotta check the rendering engine out when I get some time. To answer your question, I used 3d max 2 years ago for 2 weeks and gave up because that software was way too expensive for me. I have exported many things to Blender. If 3d max is similar to Blender, I would try to export it as FBX, but FBX doesn't always work, especially with animations. Then if FBX doesn't work, I would try exporting as Alembic. In my experience, Alembic export aren't very efficient, they cache the entire mesh and every point and every point transform as a new point in the next frame. So you lose all the sim data like velocity, force, impulse, etc. And you need all that for rendering. One way you can try is to attach all these attribute data into the Color and use that as a medium to transfer the data back and forth.
So, does it mean that there is no true interaction between Vellum and RBD?
As of now, I don't know of any. The main thing is that the multisolver doesn't work with the vellum and rbd or maybe I just don't know how to get it to work together.
Hope that helps!
hi, i love your tuto but plz 360 p is not enougth. 720p should be the minimum ;)
It should be 1080p. Something went wrong! Thank you so much for telling me! I will correct this asap.
I didn't do anything, but it seems that UA-cam fixed itself. lol It should be 1080p now, please let me know if anything goes wrong.
@@bubblepins it now ok ! Thx
Hi, every video after uploading need some time to be available for watching in HD, it's about 1-5min I think
@@yaroslavhoncharenko106 Thanks for telling me!
just insert both solvers into a merge and turn on mutual interaction
have you tested that? I did but the cloth behave very strangely
Where, in what solver? The merge node has no option for mutual interaction.
@@chillsoft in merge/ affector relationshp part you can choose mutual there
@@omidmehdipour4912 exactly
@@omidmehdipour4912 In what node? The merge node? Is this a DOPs thing? I'm doing all in SOPs for convenience! :P
how to reveal something by vellum cloth??
Could you give me a bit more information? I'm not sure what you mean.
@@bubblepins i put cloth on my car , and created a scene in mountains, to reveal it , by dragging the cloth backward? how to do that? i am sharing my hip file too, forums.odforce.net/topic/48539-vellum-cloth/ here u can check my hip file , if u want :)
I'm watching in 180 at the moment
Thanks!
Thanks. But It's very very difficult. Is there any different way to achieve interaction?
Not that I know of. This is the way I came up with because the multi-solver didn't work for me. Usually, in order to combine different types of simulations, you would use the multi-solver, but I couldn't get that to work with vellum & rbd.
This multi-phase workflow can be logically simplified a little bit more, but with the cost of more nodes to setup.
Try this. Set up the Vellum sim. Then file cache it. Then set up a rigid body sim separately like how you do it normally in a different geo node. Then just do the same thing in the video, throw down a static object node inside the rigid body simulation and just link it to the separate Vellum sim file cache.
This way involves more nodes to setup, but it's easier to understand.
@@bubblepins Thanks. I will try it. Love your tutorials.
Interesting. Do you know how would i go to instance (multiple different) vellum objects? And after that, it would be good to replace each with proxy object?
hmmm.. I think you can use the merge to combine different objects, but give them a different group name and when you apply the vellum constraints you can filter it by group name so the constraints are only applied to certain objects. Say you wanted object A to be have more stiffness than object B.
As for the proxy, I think proxy is always a good way to optimize the scene and speed up the cooking performance. You can try what I did in the last Live Stream on Vellum Skin, I used a point deform to transfer from the proxy Vellum Solver object back to the original object.
Hope that helps!
@@bubblepins I haven't checked the last stream, so definitely will!
By the way, yes, in constraints, the object affected can be set by choosing with a group. But i was rather thinking on different objects as different meshes. Imagine having 2 types of flower, one Orchid flower and a Rose flower...so two different (like a box and sphere). So by this, i was thinking of instancing them on a grid so that i can simulate bunch of them falling through the air. For that, i believe 3 things are needed.
1. To be scattered through a dop net, probably with scatter to points in vellum source (not continuous),
2. To be scattered (1.) but as different objects, for example two flowers (or two basic prim. box and sphere), that are distributed randomly on the scattered points.
3. And to be able to correctly replace these simulated objects (that by now come out from a simulation DOP net, and have been scattered on a grid) with the proxy meshes (for example, you mentioned with the point deform). My guess is with a copytopoints, but can't get it to work currently.
It seems straightforward, but i had problems with the setup, mainly with step 2. and 3.
Im going to see the livestream...
@@NextroneOfficial You have an interesting idea there! It sort of reminds me of a scrap paper scene I did for my Members-Only videos. I used the Vellum Solver to model some scrap paper and have the scrap paper dumped into a garbage can using the Vellum Solver so it would end up all over the ground near the garbage can. The thing is my scrap paper scene was very light, because all the paper were just plane 2D grid at low poly count. As for the petals for flowers, I think the biggest challenge may be poly count and performance. The copy to points may not be too suitable for animation, but who knows I might be wrong I'll need to test it, but copy to points can definitely be used for the starting position. Proxy objects used during animation, I would stick with point deform. If I get time after I finish off the Vellum Clothing mini series I'll try the petal scene out!
@@bubblepins Hey! I looked through the stream. I believe the way you approach making tutorials is unique, and that one can really enjoy all the bits and pieces of the explanation that you carefully try to convey to the audience. I haven't find the particular case in it (i think i haven't missed it).
Right, but it's a bit different over here, scrap paper is a plane and flowers are closed 3D objects :-)
Also, i do not intend to add separate vellum constraints to the petals...just a vellum per whole flower is totally fine because they would merely collide while flying, and i think it wouldn't make a difference since there would be many of the instanced and different ones and the eye wouldn't pick the difference in realism i think.. so no ground plane would be included in a shot, just the fall through. Anyway, it's a bit tricky setup.
@@NextroneOfficial The scrap paper scene is in a different video. Have you considered using textures for the petals instead of a 3D objects? This will speed up the performance, and collisions should still work because you can add Vellum thickness to the Vellum Solver. But I'm interested in your idea, because I always wanted to procedural model flower petals. This could be very cool for 3D printing them out too!
How did you create a ball falling into cloth(pinned to all four sides), I am trying to achieve a similar effect?
The ball just drops from free fall (affected by the gravity force) and I pinned all four corners of the cloth so it stays in position. Do you have a particular issue you're having trouble with?
@@bubblepins I have rigid body solver setup for free fall of the ball and vellum solver for cloth and both solvers are plugged into the merge with Mutual relationship the ball kinda interacts but it weirdly passes through tearing the cloth in a weird way .........If u can provide a file of the setup it would be a great help as I am working on a similar project and I can just replace my model with yours .....Thanks for your help!
@@gameswamy1465 The thing is Vellum object doesn't really work with the rbd object in the dopnet with the mutual relationship or even with the multi-solver. Or at least I never got it working that way. I did both simulations separately even with the sphere dropping onto the cloth, using the technique I showed in this video. I know there are a lot of comments on below here saying that they want a "true" interaction, but Vellum isn't designed the same way RBD solver was designed. This may change in future Houdini versions. For now, the technique in this video was what I came up with as a work around for anyone that needed this.
Hope this helps!
@@bubblepins I found a workaround which worked for me .........I solved the sphere as a VELLUM SOFT BODY and not RBD and changed the settings for the sphere that it wont deform much.... And increased mass etc.......so this works if we don't want fracture ......I think for fracture RBD we have to use your way .....but if just and interaction in this simple free fall this was the workaround I came up with .........Thank for you help....👍👍👍
confusing af
this is not "interaction". just wasted my time :(