Is it possible to add a checkbox for Stoke to delete particles in the stoke simulation leaving the fume grid?
I will ask around, but I am not sure if we could do it easily.
Stoke particles generally do not leave the grid since they adopt the velocity at the current sampling point and don’t accumulate Speed like PFlow does. In that edge case it results in zero velocity. You could drop a Magma that deletes close-to-zero-velocity particles, but you risk that these same particles might spring back to life if the Fume suddenly gives them a non-zero velocity at a later point in time…
Ah ok,
I had a fume sim that was colliding against the fume grid, but the stoke simulation particles left the grid and became a huge cloud around it,
i tried volume select with delete particles which works as a post process but it doesnt really help the simulation
ill try to post some screenshots next time
[attachment=0]particles_problem.mov[/attachment]So it seems i’m having the same problem but with real flow sim. I’ve tried with particle culling at the original prt loader and it’s working… but when using stoke to make all the partitions, they appear again… kind of desperate feeling to have this small issues while in production. If I can’t solve this issue fast i’ll try krakatoa partitioning. Gonna try out this volume select to kill particles inside stoke.
Meanwhile, what’s the correct procedure to kill particles with velocity=0 inside magma?
In the example attached we can see gray particles with correct motion and kill info (krakaprtloader) and in orange we see particles stopping (stoke partition).
Thanks,
PS: Already tried krakatoa partitioning and it happens the same. When particle should be “killed”, it stops. HELP! 
I’ve tried another aproach with pflow + realflowfilebirth + realflowfileupdate + realflowdeathtest just to check if motion was correct… and it’s exactly where particles should be deleted that somehow krakatoaprtloader and stoke are not interpreting this info.
My next move is to export data from real flow in krakatoa prt format and see if it solves the problem.
Cheers,
Update:
Prt export from real flow is a success. Particles are deleted when they are supposed to… Nevertheless, I remember something about the id channel export from real flow… any warnings about this Bobo?
Cheers for now,
After digging, fighting, emails and so on, this was the solution we found to delete particles that were in fact with velocity=0
Hope this can help someone with the same problem!
Cheers,
Hurray for Darcy coming to the rescue!
Glad you’ve got it working.
To give credit to Bobo, that’s pretty much what he suggested (deleting particles with close to 0 velocity).
If you ever simulate with regular Stoke (not Field Sim), you can delete 0 velocity particles by checking “delete dead particles”.
Another way of deleting particles more gently would be to add a blend node in the graph and set the output to selection and then basically set up a flow that says “blend between selection float value 1.0 and float value 0.0 with the blend value being velocities between 0.0 and 0.2”. Then add a Krakatoa Channels Delete modifier afterwards with the soft selection option checked. Or use the same flow and output to density but then you don’t see the result in the viewport. To problems like these there are generally several solutions and being aware of them can come in handy in other solutions.
Hi tobbeo,
100% agree with you. Our problem turned out to be a little more complex, since the particles going to velocity 0 were more than the one’s getting killed (real flow sim). It was stoke grid spacing problem, and even when we found a “good” value where particles follow up the original one’s, the solution was so different that we could not use at the end of the day. Back to krakatoa partitioning method.
Cheers,
I see… So from the sounds of it in this thread and others it seems like Realflow guys need to update their PRT exporter to the latest version. If I understood things right based on this thread and another .
i sent an email to their product manager - might help if some users asked…