Jump to content


  • Content count

  • Donations

    10.00 CAD 
  • Joined

  • Last visited

  • Days Won


Andrea last won the day on June 20

Andrea had the most liked content!

Community Reputation

72 Excellent

1 Follower

About Andrea

  • Rank
  • Birthday 12/31/1993

Contact Methods

  • Website URL

Personal Information

  • Name
    Andrea Sbabo
  • Location
    Lost in the Alps

Recent Profile Visitors

955 profile views
  1. That's because of the node "Gas Resize Fluid Dynamic". If there is no smoke at all after the first frame or the amount of density is below the value of "Field Cutoff" is going to resize the bounding box to the origin. If you are sourcing is SOPs you can use the "Tracking Object" options so the bounding box is going to be at least as big as your source. Alternatively you can use "Delay Frames", but this is going to calculate the full bounding box for the amount of frames specified here
  2. Flip sim goes crazy when I add collision

    The issue here is not the velocity of flip but the velocity of the collision geometry which is very high ( too much). It's spinning really really fast, I can see values between -1100 and 600. You can see immediately this problem also in DOPs, enabling "Collision Velocity" under Guides. If you don't want to change the sphere animation, you can reduce the velocities by using a wrangle before OUT_COLLISION_POLY or in the collision tab reducing the velocity scale, which is basically the same thing. Here is the hipfile. Does this help? Issue with FLIP simulation.hiplc
  3. DOPNet or not ?

    If you look at the grain's constraints in the second output of vellum configure grains SOP you can see that they are just points with a position information, and makes sense since it's based on a point method. So if you have setup everything correctly previously you don't need to do anything more complex than that.
  4. DOPNet or not ?

    Your first post is not really clear. DOPs and POPNets are the same thing, just a POPNet describes a DOP with a POP solver inside. As we have seen we can initialize vellum constraints in SOPs or we can create the vellum constraints during the simulation for specific behaviours. Yes you can create constraints just in the DOP Network but if you mean emitting particles using a POP solver, assign vellum constraints and expect them to be grains it will not work. I have never tried but I think you need to assign to the particles certain attributes like @isgrain to make them work.
  5. Smoke - flat leading edge

    Hi! Can you post the hipfile? The volume looks low resolution but shouldn't be that visible the flat edge
  6. What I mean is using it for knowing how far is the point inside the VDB and move the point accordingly.Of course you can also using booleans i you want to cut it
  7. For sure you can setup the main bending to be procedural and introduce more subtle details in the same spots with noises. Do not use just noises. In my previous experience with some "procedural destruction" (not with cars) I noticed that I got the best results by instancing a couple of objects with different scale on the surface (spheres/cubes with a low frequency noise), transforming them to very low resolution VDBs and using volumegradient to control the displacement of the surface. This way I was getting consistent and smooth deformations (no intersections). Later on, with a couple more nodes I've added a variation in the noise to the various main parts (in your case for example to the doors, body of the car ...). This last steps introduced some cracks that was very helpful in my case. Never tried with metaballs, maybe they are better If you have a lot of "extra" objects and details in your car, you should consider to create some sort of proxy geometry, deform this one and transfer back the deformation to the original car. It's going to be faster to deform and probably cleaner. Thank you! I would like to make it again someday, I had no idea what I was doing at the time
  8. Hello! Maybe you can try with manual sculpting of the damaged parts and then add on top the noise (I think I would go with worley and antialiased). You can also get extra details with the shader and painting cracks. Doing a real simulation would be cool but between geometry preparation and setting up the simulation, you risk to lose a lot of time.
  9. Doing what I mentioned before gives you a good collision geometry without extra stuff but of course the spheres are not perfectly close one to the other. I have always used this for getting correct collisions and ti was enough. To be honest I don't know if the spheres have ever been perfectly packed with no gaps
  10. For each sphere I was looking at the distance to the surface of the geometry. If the pscale was bigger than the distance to the surface, then I set pscale to be this distance. Of course later I had to instance back again spheres to these points.
  11. For sure this was also happening in Houdini 17 because of this I had to correct with a wrangle the pscale of the spheres after "VDB to spheres"
  12. Same here in Houdini 17.5.293. I created the scene also from scratch and I get the same result, so it's not related just to your hip. Changing a little the parameters or decreasing the VDB resolution, helps. But seems random, I haven't noticed any correlation
  13. Hide parameters through python

    Thanks for letting me know ahead! I haven't encountered this issue yet, so I haven't spent much time solving it. Luckily in my case I can also just create a folder for all the parameters created by python but definitely would have been much cleaner to hide the parameters in a standard way, without workarounds
  14. How to make whitewater size smaller

    To make them smaller you can reduce the pscale attribute to a smaller number. From your image seems like you don't have a lot of white water particles, so probably this is a low res simulation. When you are going to increase the resolution of the simulation, the particle scale is going to be reduced automatically by the white water solver. Of course you can always tweak it later
  15. Hide parameters through python

    Hi Luca! Thank you so much! It works perfectly Curious to see that you are calling the folders by their label and not by their name. I'm very new to this kind of stuff but I was so sure that functions need to reference to parameters and folders by their label only if we specifically look for them. Good to know I was wrong