@@bubblepins LOL!! That's cause I knew if I'd have continue watching the video, I'd have run into a loop error xDD. Needed a good sleep! I'm here, back again to face this tremendous tutorial :D
Thanks for the comment! I think SideFX is starting to shift away from DOPs. I find that all their newer updates are focusing more and more on supporting the simulations in SOPS.
Wow this was the toughest to understand out of all your previous tutorials that I've seen haha! I watched it thrice and I still feel I'm missing bits and pieces. Could you maybe make a blog post summarizing this particular tutorial?
@@bubblepins I think the logic behind and use of VOPs in the node network in geometry vop in dopnet would require some explaining. The rest I think I got it after one more view 😄
I do agree the node network for if blocks could be designed a bit better, but it is a hard task to make if conditions into nodes. VEX is definitely a lot simpler. I'm hoping to come up with a mini series on the side for VEX soon, but I have the same problem, I'm not sure how to teach VEX...
@@bubblepins yea totally agree, translating codding concepts to nodal based systems is a dificult task, let alone SOPs for loops :D. You surely will make great VEX learning videos, you have great teaching skills. I reckon a good way to introduce VEX to someone familiar with VOPs might be building your code in VOPs and then the same in VEX, and comparing them line by line, node by node? Just an idea
thanks alot for the tutorials I couldn`t find much info about RBD sims with custom forces. The Tutorials are well explained - but I find it in general very laborious to get to the final result. - particularly to handle the Glue-constrains seperatly - there should be an option to break them automatically by a given force - this would make such tasks alot easier
This exists in RBD tools from Houdini 18. There's a breaking thershold panel in the RBD Tools, which is a new feature in Houdini 18. I am currently working on some videos on RBD Tools which has a completely different workflow compared to the dopnet. Stay tuned!
thanks for nice tutorials! it was a liiiiiiiitle bit confusing (i have only one output of the voronoifracture node) because mine is 16.5 version! other than that everything was well explained and perfect! thanks again
Check out my other video demonstrating destruction without using the voronoii fracture node ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-fMft1nQLmd8.html
Great series, been trying to figure something out for a while.. do you know how you might break soft constraints? for example when using a bullet solver?
Yeah it just depends on how strong your soft constraints are set and double check in the bullet solver under the Constraints TAB double check the thresholds to make sure they will break under normal conditions. When in doubt, just make these thresholds values very small and you will start to see something break. Hope that helps!
Hey, first of all - thank you very much. Taking the time to explain simple stuff step by step is exactly the reason why your tutorials stand out. I was looking for a way to explode my rbd-object and this is pretty much what I was looking for except for one aspect. My scene is a small room with various prefractured objects and I want my "explosion" taking place in the room in slow-motion. With your method I can do exactly that, but it's affecting every object at the same time. Since there is some distance between those objects it would be realistic, if the nearest to the explosion would scatter first. And on top of that I would like to only affect the fractured parts that are facing the explosion, so that everything behind them gets pushed away from the particle, but not from the force itself ( since the force of a explosion would not be able to go through a wall ) I was wondering if you know a approach to this, or a website to look all this up. Thanks again!
I think you can add an attribute to your target geometry (the one being destroyed) and then inside the geometryvop read that attribute to figure out if the force should be applied to that geometry. I think this is a very interesting question, I'll try this out when I come home tonight and upload a potential solution to my blog. Come check back tomorrow!
Hi, I uploaded an example HIP file where I control the force field and which geometry it is applied to inside the dopnet using a target attribute. bubblepins.com/blog/controlling-force-field-for-specific-geometry-in-destruction-simulations-in-houdini
@@bubblepins Thank you so much! I'll try to implement and understand exactly what you did there as soon as i get home tomorrow. If you don't mind, i will let you know how that worked out for me. Again, thanks a lot for taking the time!
I think you could try adding a wrangle to the sopsolver to record when the impact happens, but that's only possible with a solver node, because that's the only node that can recall attributes from previous frames to remember the impact state. Then you know exactly what frame the impact happens, you could do something like Current Frame subtract Impact Frame then you use that to determine the force. The larger the difference of the two numbers, the more you diminish the force. Hope that helps!
Hello there. Just trying various methods of RBD descruction and found this really quite useful. Excellent work! Only one question though. After the initial explosion my scene goes into bullet time. Any vague idea why that might be? And thankou again.
Thanks for the feedback! What's bullet time? Is the blast pieces traveling too fast after the impact? If I didn't misunderstand, I believe the constraints are too weak on the fractured geometry and the impact might be too strong for it resulting in a larger explosion. Try tweaking the constraints' strength and making them stronger and maybe try making the impact less. Hope that helps!
@@bubblepins thankyou. It's actually doing the opposite, Goes fast then they slow down. Like in movies when they slow the footage to see closeup action in more detail. Looks cool but I can't see what;s causing it. I have no drag as far as I know in the scene.
@@bobvelocity Are you previewing this in the viewport or doing a flipbook animation? It's best not to trust the Houdini viewport, because it's always computing something in the background, flipbook will give you a more accurate feedback. If it's not that, I wonder if there's a hidden force that's being animated by accident slowing down the pieces, try checking out the force of the pieces after the dopimport node and see if the force, or v has changed significantly, then you'll know somethings up. Or if you're brave enough, you can check out the force inside the rbd solver within the dopnet. To see that you'll need to expand the rbdobject/geometry and you'll get a similar spreadsheet layout as you would in the sops world. See if that helps!
Thanks for this video but I had one issue .when I connect force with if block node inside vop they give an error ' the force has different type ' how to solve that error .
I'll need a bit more description of the setup, but I'm guessing you might have plugged something like a constant into the force plug, because judging from the error message it sounds like you're trying to plug an integer, float, or string into the force vector. You'll need a vector to vector matching type. But its hard to say from just a single error message, it could be something else. Hope this helps so far.
It might have been fast forwarded in this video, I do a lot of heavy post processing after recording. My machine specs are THreadripper 3960 - 24core, 64gig of ram, rtx 2080 and I had updated my pc to a NVME M.2 ssd drive which REALLY helps speed up the caching in Houdini. I think that was the best upgrade I ever made, because NVME's are relatively affordable when you compare it to the amount of performance boost you get. Hope this helps! I even wrote a blog post on this: bubblepins.com/blog/affordable-hardware-upgrade-for-houdini
I'm sorry I don't understand well. Are the search points of 'near point vop' search only one for each point of Sphere brought into Output1? (If I set 1000 points with the Scatter Sop, 1000 Pieces will be affected? ) And if what I said above is correct, the reason why the whole wall breaks down is because each of those points pushes each piece away, is it right? I think it is probably wrong. Please let me understand...ᅲ
I explain this in more detail with the previous video. I made an animation illustrating the closest piece and what that vop is trying to do. Here's a link, this link will bring you to the exact time in the video where the explanation starts: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-PH-GnpAaLhI.html But to answer your question briefly, this vop iterates through each piece (or each point, same thing because we packed the geo). Anyways, it find the nearest point beside each piece. So there's only 1 nearby point. Nearby node only returns one point number, it return one value, and this is the nearest piece to the force point (the sphere that was created in the previous video). So the objective is to assign directions to all points on the sphere, which is our force field. This video focuses on how to use the force field, the previous video focuses more on creating the force field. Hope that helps!
That's actually a two part simulation. Do the Fluid simulation first, then convert it into polygons and then feed it into the RBD simulation as a collision geometry. At least this is the only one I know how to do it.
