towje Posted October 24, 2011 Share Posted October 24, 2011 Hi all, me new here and trying to learn houdini to make use of it as a freelancer. so, thats enough of me, and no, i do not really talk like that .. ^^ ok my question/problem.. as in the topic, i have a billowySmoke simulation set up. First pyrosolver drives some particles. Then these particles are driving another pyrosolver (well, not the solver, but .. you get it ) of the "real" smoke. The particle driven one is the low res smoke simulation. A ground plane is set up as a collision object. All fine till this point. Now, if i activate the uprespyro, to upres the particle driven low res simulation, the ground plane is ignored by the uprespyrosolver, it just goes through .. I'm sure, it's an easy mistake, or at least i hope so, but i can't figure it out. ^^ any clues? would be much appreciated! .. greetings, towje Quote Link to comment Share on other sites More sharing options...
CiaranM Posted August 23, 2012 Share Posted August 23, 2012 I'm hitting the same problem in H12....any solutions? Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 23, 2012 Share Posted August 23, 2012 Upres solver doesn't handle collisions, are you sure your collisions are working in the low res? The vel fields should indicate whether collisions are working in your low res. 1 Quote Link to comment Share on other sites More sharing options...
efx Posted August 23, 2012 Share Posted August 23, 2012 Pyro in H12 uses non divergent multigrid as default which does not support collision. if you using it, change it to PCG ( non divergent that is same with H11 ) might help. But I remember that I had same problem in H11, I think collision's not gonna be perfect in upres sim. 1 Quote Link to comment Share on other sites More sharing options...
CiaranM Posted August 23, 2012 Share Posted August 23, 2012 Hi, yes the collisions are working in low res. I was not aware that the upres solver does not deal with collisions. Interpenetration becomes apparent when adding noise - the up res smoke clearly passes through thin walled colliders, while the low res sim does not. Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 23, 2012 Share Posted August 23, 2012 You could always just build a mask volume from the collision sdfs and zero out the density inside. 1 1 Quote Link to comment Share on other sites More sharing options...
CiaranM Posted August 24, 2012 Share Posted August 24, 2012 Thanks for the tips! I ended up using a gasfield VOP to sample the collision field and mask the output density appropriately. Not sure if that's the most direct way? It would be nice to move the density out along the gradient so that density is not lost. I had some issues...volume sample from file returns the density in the local space of the sampled object and does not account for transforms on its container geo. I solved this by constructing a transform matrix in the VOP by referencing the transform channels of the sampled container (see attachment). It all seems a bit hacky...Is there a better way to just drop an object's matrix into my VOP graph? VOPVolumeSample_01.hipnc Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 24, 2012 Share Posted August 24, 2012 wouldn't it be easier to object merge that field into your container? removes the transform problem. Quote Link to comment Share on other sites More sharing options...
CiaranM Posted August 28, 2012 Share Posted August 28, 2012 wouldn't it be easier to object merge that field into your container? removes the transform problem. That does it, cheers! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.