slpash rate one example.

[victorwol]

Does anyone know what number was aprox the birth rate of these splashes on this example?

https://docs.chaosgroup.com/display/P220N/Splash#

Thanks!

[victorwol]

does the numbers of particles used on the splash increase automatically when you give more resolution to the grid?

Sorry just noticed I posted this on the problems part of the forum.....

[victorwol]

Well... now I do have a problem, the splash renders and is perfectly visible when I calculate in a low cell count, but as soon as I up the cell count to over 2 million, the splash does not render anymore....

[Ivaylo Katev]

the splash birth rate is not increased automatically, but it does not directly specify any particle count. when you increase the resolution there are more areas in the simulation where the birth conditions are satisfied and therefore more particles are born.
about the rendering - you mean the particles are visible in the viewport, but not in the rendering?

[victorwol]

That is exactly what I mean. I see them on the viewport but they are not on the render. And it is like are not even calculated because the render is super fast. It seems to be like a trigger or a limit that makes it stop working suddenly.

[Ivaylo Katev]

strange, the demo builds were limited in particle count rendering until version 2.2, but now they are not. can you send us the log file? you can find it in c:\\phoenixfd folder

[victorwol]

Sure. I'll do as soon as I get to the office. FYI I'm not using a demo version, I have a licensed one. And the liquid renders fine, it is the splash that disappear when I sort of cross the threshold of 2 million cells and an interesting amount of particles for the splash. And it is like as I increase the cells, I get more splash particles and they become less visible as the more they are until they just vanish. Do they change the size when the cell size decrease?

[victorwol]

Here is the filephoenix_maxver17.zip

And here you can see the particles.

if I change the shader from splash mode to point mode, then it does shows up on the render, but as splash, probably only 90 to 100 particles render.



if I calculate this with a low count of cells, they render fine.

[Ivaylo Katev]

well, the log shows that 93408 particles are loaded in the shader but really no content is rendered.
my hope was to see a wrong parameter in the log, but everything seems to be ok, can you send us the scene with single cache file .
btw, for the liquid-like behavior of the splash is very important the particle size and count. the liquid-like parameter just keeps in contact particles that are near enough, this means particles that are touching each other. when the particle size (or count) is low, the probability to have near enough particles is low and the option produces no effect. as i see in the image, you have liquid like parameter set to 0.5, but the splash behaves like if it is zero, too many free droplets and no chains of particles. you can see in the documentation how it must look like. i suppose the particles are too small, you can try simulation with bigger particles size, don't worry about the rendering, they can be shrunk back in the shader.

[victorwol]

Here is, https://www.dropbox.com/s/fskjqzenlm...hTest.zip?dl=0
Let me know when you get it so I can delete the file.

What is weird that without changing any parameter, if I solve the simulation with a low cont, then it renders, that is why I was suspicious about the particles becoming smalled are you increase the voxels count, why would they dissapear then?

[victorwol]

by the way here is what I rendered last night, but using the point render for the spash which is SUPER slow.

PLEASE.. I beg, make a setting to change the friction for the wet parameter on each object, it is something that adds realism to the simulation, but as it is now all the objects have exatly the same friction. That will add a lot more realism to the simulations in my opinion.

Hope is not something super complicated to implement and that you can get is sooner than later

[Ivaylo Katev]

of course we will add per object wetting and friction, but we still have a lot of work to make the flip solver ready for release.

[victorwol]

so did you had a chance to download my test scene? I'm really curious why the splashes dissapear when I add resolution to the scene.

Thanks

[victorwol]

Well... something happened that now I can't simulate anymore on this scene without crashing.... and I'm not on the FLIP mode. it calculates 1 frame and when tries to do the next one MAX just crashes... I see nothing of importance on your logs... A new from scratch scene works fine...

P start sim
SetSolverParams
Grid set to 441 328 232, cells 33558336, memory usage 2684MB
SetSolverParams
SetSolverParams
sim started
sim thread 0 started
output: E
sim thread 1 started
sim thread 2 started
sim thread 3 started
sim thread 4 started
sim thread 5 started
sim thread 6 started
sim thread 7 started
sim thread 8 started
sim thread 9 started
sim thread 10 started
sim thread 11 started
sim thread 12 started
sim thread 13 started
sim thread 14 started
sim thread 15 started
sim thread 16 started
sim thread 17 started
sim thread 18 started
sim thread 19 started
sim thread 20 started
sim thread 21 started
sim thread 22 started
sim thread 23 started
sim cycle 0 --------t:0.000000
s(0),s(1),frame loaded E:\2015\Victor\PhoenixFD-Tuts\ColdHTests\SplashInvestigation.max_Phoenix2_f rames\ND27_PhoenixFD001_####.aur, 0
SetSolverParams
s(2),s(3),s(4),s(5),s(6), expand 0.051327
displ scale:1.953125
march cubes build smap:0.085743
march cubes vertices:0.128643
march cubes interest:0.006034
march cubes conn time: 0.016061 result vertices:38590 triangles:77176 :0.000000
mesher: extend time 0.006701 vertices:38590 triangles:77176
mesher: normals 0.006539
meshing: vertices:38590 triangles:77176 time:0.264880
mesh normals:0.025845
~buildmesh ^^^^^^^^^^^^^
s(7),s(9),s(10),s(11),s(12),save frame 0 <<<<<<<<<<<<<<<<<<<<<<<<<<<< frame <<<<<<<<<<<<<<<<<<
frame loaded E:\2015\Victor\PhoenixFD-Tuts\ColdHTests\SplashInvestigation.max_Phoenix2_f rames\ND27_PhoenixFD001_####.aur, 0
expand 0.351506
displ scale:5.960463
march cubes build smap:0.099347
march cubes vertices:0.089649
march cubes interest:0.070487
march cubes conn time: 0.039256 result vertices:360044 triangles:720084 :0.000000
mesher: extend time 0.043772 vertices:360044 triangles:720084
mesher: normals 0.020981
meshing: vertices:360044 triangles:720084 time:0.367198
mesh normals:0.063773
~buildmesh ^^^^^^^^^^^^^
s(13),dt=0.016667, active:1
s(14),s(15),s(16),

[victorwol]

something really weird is going on.... on this scene and some others that I worked a few days ago, I can't run sumulations again without crasing. I have uninstalled it completelly, and then re installed it, and it does not work, it keeps crashing after simulating the first frame. Even if I delete all the FD and start from scratch on this scene, with a new FD with all by default, it just crashes after the first frame.