about using copy to points emitting vellum objects , each frame im emitting a low poly vellum object , and want replace this for the higher poly object , using point deform , and its ok for the first object being emitted , but cannot find the way to keep replacing all vellums low deformations being replaced by high poly objects... i check the piece attributes , try use some for each point do this or that but no luck yet....
Hi Tim. Great tutorial and tricks as always. Thank you! Advance stuff is very appreciated. Any chance to extend this tutorial by showing how to source different geometries?
Hey, thanks for the tutorial ! I was wondering how can I emit in vellum and your video came at the perfect time PS : If you want to delete a channel parameter you can do : CTRL + SHIT + click on the parameter to delete it
Please can I change the emitting rate. That is to emit every second or more. And can I colour each balloon? I realised that I when I apply random colour to the source point in the sopsolver or the geometry wrangle, it doesn't apply to the balloon
You can change the rate by the amount of points they are coming in. For example if you put in your scatter $F%5==1 it will scatter 1 each 5 frames If you’d do ($F%5==1)*3 it would emit 3 every 5 frames. To do a random color the best way is the randomize an attribute on the incoming points, make sure it copies it to the vellum source (copy instance attribute tab) then you can asign a color from that attribute from there after you do the sim.
Hi, thanks for this, very useful. One minor thing I've had problems with is that the increase in restlength seems to be a bit hard to get consistent. The comparison with the restlengthorig seems to vary wildy - if I put
Great stuff Tim. Have been looking into Houdini. Is Houdini good to learn for motion graphics, advertisements, etc... Usually it is said that Cinema 4D is industry standard for those things and simple user interface and easier and quicker to get things done. Do you have any experience on this side? So for motion graphics and tv ads use Cinema 4d and for vfx learn Houdini?
Thanks! This really depends on the type of motion graphics you want to do. Cinema 4D is indeed the main tool for a lot of motiondesign studios, because it’s easy to use. However, any motiondesign studio that does somewhat higher end motion design will probably be using houdini (or wanting to use it) to some extend. If you’re going to be doing motion design at a studio that primarily does design work and 3D on the side (for logos and simple animations) then c4d is probably the go to there, but for anything more higher end you’re quickly going to end up with houdini, as it’s a lot more flexible. So if you want to work primarily as a designer and do some 3d, c4d is fine. If you want to do the higher end productions then houdini is going to be very in demand. c4d is extremely easy to pick up anyways if you happen to need to know it at some point so wouldn’t worry too much about that, if your houdini skills are good enough you’ll have a much easier time getting a job. edit: I used to use c4d before i got into houdini btw
I dont know why, but now using 20.0.625, if i want to scatter points inside the sop solver, the data name of the sop solver need to change to "Geometry ConstraintGeometry", or else it will give me the error "The number of points in the geometry and constraints do not match."
Hey Tim, this is really cool to see your process - particular i liked your method for removing intersections on the bubble sim. I've done a similar thing before but less streamlined. One thing i notice though is if you have the vellum substeps greater than 1, it breaks. You find that too? I feel like i've had this before where its trying to process that multiple times because of the increase in substeps. Really helpful though : )
I think its trying to source 3x per frame for 3 substeps. I think the enable solvers trick is it perhaps but i haven't used that it much, will look into it
My fault for doing something maybe a little too tricky for week 2 hahaha. Great find on the gas intermittend solve, i had not though of that. What i would normally do is set it to only enit on ($FF%2)=1, that will only evaluate to 1 every other integer frame, so you could fix it like that but the gasintermittend solve seems like best solution. thanks!
@@TimvanHelsdingen ah but i think your way evaluates a little faster so that good to know. And good point. I was thinking how to only make the sop solver evaluate each timestep but makes sense to just force the sourcing to do that. Thanks!
Help please! Running the first simulation on GPU results in windows freezing Houdini. It runs fine on CPU, but can't get it to work on GPU. Happens on both 18.5.596 and 19.5.368
Are you running out of vram perhaps? If not, are you on recent drivers? Also you may want to upgrade to a newer houdini 19.5 build, there were some opencl issues in earlier builds that made vellum crash
@@TimvanHelsdingen Vram doesn't even get close to full, but by looking around online so far, I found that OpenCL had problems with pressure constraints so it might be just that
Hi ^^ This is maybe not the right video or context to ask a very specific question but there is something I am still not able to figure out by myself and can only copy and repeat from other tutorials. It is the type of inputs of a geometry wrangle. I still cannot find a "trick" to tell to myself to understand what means "Myself" and "Myself (No reads of outputs)". If someone could give me a simple explanation to differenciate both type of inputs and help me understand what they do exactly, that would be awesome ^^
If you put a geowrangle in dops to “myself” it means it will run on the dop geometry itself (so geometry or constraint geometry) alternatively you can point any inputs to other things. You can just reger to them with 0,1,2,3 as inputs in the wrangle. It’s not that much different form a wangle in sops, but sometimes you might need to reference something from somewhere else, so that’s what it works the way it does.
![Houdini Vellum Cloth (Beginners) [Tutorial] | Emitting Vellum Cloth Bands + Shrinking/Shrink Wrap](/img/n.gif)
Great lesson! So many solid tips! Thanks so much for sharing your knowledge 🙏
Omg your video intro was so use. Good to hear what will be going on
about using copy to points emitting vellum objects , each frame im emitting a low poly vellum object , and want replace this for the higher poly object , using point deform , and its ok for the first object being emitted , but cannot find the way to keep replacing all vellums low deformations being replaced by high poly objects... i check the piece attributes , try use some for each point do this or that but no luck yet....
Hi Tim. Great tutorial and tricks as always. Thank you! Advance stuff is very appreciated. Any chance to extend this tutorial by showing how to source different geometries?
Good idea! that might be interesting
Thank You very much, Tim!
Hey, thanks for the tutorial ! I was wondering how can I emit in vellum and your video came at the perfect time
PS : If you want to delete a channel parameter you can do : CTRL + SHIT + click on the parameter to delete it
perfect timing!
That’s a good tip for deleting channel parameters, thanks!
Thanks man. Awesome work!
Man, great stuff! Thanks for that!
It won't allow me to access the groups I outputted in sops in the attribute wrangle in the sop solver.
Please can I change the emitting rate. That is to emit every second or more. And can I colour each balloon? I realised that I when I apply random colour to the source point in the sopsolver or the geometry wrangle, it doesn't apply to the balloon
You can change the rate by the amount of points they are coming in.
For example if you put in your scatter $F%5==1 it will scatter 1 each 5 frames
If you’d do ($F%5==1)*3
it would emit 3 every 5 frames.
To do a random color the best way is the randomize an attribute on the incoming points, make sure it copies it to the vellum source (copy instance attribute tab)
then you can asign a color from that attribute from there after you do the sim.
Thank sir 😁
i liked the beard much! so much wisdom
Hi, thanks for this, very useful. One minor thing I've had problems with is that the increase in restlength seems to be a bit hard to get consistent. The comparison with the restlengthorig seems to vary wildy - if I put
wow thx a lot for the tips!
Everyone should learn vex. It's not that hard, it's very very powerful in Houdini
a lot of people are intimidated by it, but it’s pretty easy if you just start using it with simple snippets like this
most people/studios are intimidated by houdini @@TimvanHelsdingen great tut by the way, thanks for sharing
Great stuff Tim. Have been looking into Houdini. Is Houdini good to learn for motion graphics, advertisements, etc... Usually it is said that Cinema 4D is industry standard for those things and simple user interface and easier and quicker to get things done. Do you have any experience on this side? So for motion graphics and tv ads use Cinema 4d and for vfx learn Houdini?
Thanks! This really depends on the type of motion graphics you want to do.
Cinema 4D is indeed the main tool for a lot of motiondesign studios, because it’s easy to use. However, any motiondesign studio that does somewhat higher end motion design will probably be using houdini (or wanting to use it) to some extend.
If you’re going to be doing motion design at a studio that primarily does design work and 3D on the side (for logos and simple animations) then c4d is probably the go to there, but for anything more higher end you’re quickly going to end up with houdini, as it’s a lot more flexible.
So if you want to work primarily as a designer and do some 3d, c4d is fine. If you want to do the higher end productions then houdini is going to be very in demand. c4d is extremely easy to pick up anyways if you happen to need to know it at some point so wouldn’t worry too much about that, if your houdini skills are good enough you’ll have a much easier time getting a job.
edit:
I used to use c4d before i got into houdini btw
Thank you Tim, appreciate it.@@TimvanHelsdingen
Thanks
Hi there, for some reason i can't find DOP network when i search, i think it is a bug.
Do you know any way to fix this?
Are you using Houdini Core instead of FX/Indie by any chance?… DOPs are only available in FX, core only has the SOP level wrappers for them
I dont know why, but now using 20.0.625, if i want to scatter points inside the sop solver, the data name of the sop solver need to change to "Geometry ConstraintGeometry", or else it will give me the error "The number of points in the geometry and constraints do not match."
Hey Tim, this is really cool to see your process - particular i liked your method for removing intersections on the bubble sim. I've done a similar thing before but less streamlined. One thing i notice though is if you have the vellum substeps greater than 1, it breaks. You find that too? I feel like i've had this before where its trying to process that multiple times because of the increase in substeps. Really helpful though : )
I think its trying to source 3x per frame for 3 substeps. I think the enable solvers trick is it perhaps but i haven't used that it much, will look into it
ahhh gas intermittent solve allows once per timestep. Sorry making things complex already in only week 2!
My fault for doing something maybe a little too tricky for week 2 hahaha.
Great find on the gas intermittend solve, i had not though of that. What i would normally do is set it to only enit on ($FF%2)=1, that will only evaluate to 1 every other integer frame, so you could fix it like that
but the gasintermittend solve seems like best solution. thanks!
@@TimvanHelsdingen ah but i think your way evaluates a little faster so that good to know. And good point. I was thinking how to only make the sop solver evaluate each timestep but makes sense to just force the sourcing to do that. Thanks!
Help please! Running the first simulation on GPU results in windows freezing Houdini. It runs fine on CPU, but can't get it to work on GPU. Happens on both 18.5.596 and 19.5.368
Are you running out of vram perhaps? If not, are you on recent drivers? Also you may want to upgrade to a newer houdini 19.5 build, there were some opencl issues in earlier builds that made vellum crash
@@TimvanHelsdingen Vram doesn't even get close to full, but by looking around online so far, I found that OpenCL had problems with pressure constraints so it might be just that
@@TimvanHelsdingen Btw learned so much from this "custom" vellum solver setup. It looks like I won't be returning to the default one.
Hi ^^ This is maybe not the right video or context to ask a very specific question but there is something I am still not able to figure out by myself and can only copy and repeat from other tutorials.
It is the type of inputs of a geometry wrangle. I still cannot find a "trick" to tell to myself to understand what means "Myself" and "Myself (No reads of outputs)".
If someone could give me a simple explanation to differenciate both type of inputs and help me understand what they do exactly, that would be awesome ^^
If you put a geowrangle in dops to “myself” it means it will run on the dop geometry itself (so geometry or constraint geometry)
alternatively you can point any inputs to other things. You can just reger to them with 0,1,2,3 as inputs in the wrangle.
It’s not that much different form a wangle in sops, but sometimes you might need to reference something from somewhere else, so that’s what it works the way it does.
@@TimvanHelsdingen thanks for the reply, I appreciate ^^