Surface level oddity

[GrahamMac]

I've noticed something curious about simulating liquid at lower time scales (eg: 0.01). Everything is set up to produce pouring water and it all works fine apart from bumps and animated artifacting on the surface of the liquid, which can be seen especially well on individual droplets of water. This artifacting looks as if there is vibrating noise displacement (even though there is certainly none in the settings) on the liquid surface and can be seen most clearly at macro style close up distance.

I have been consistently using surface level 0.5 (as recommended in the documentation). However, by accident I realised that by simulating with a surface level different to 0.5 and then setting it back to 0.5 for rendering the result is different than if the surface level was set to 0.5 during simulation. I assumed this parameter would only effect the rendering of the surface but it seems that it does alter the result of a simulation.

With some further testing I noticed that when I pushed up the surface level to 5, ran the simulation, then returned the surface level to 0.5 (otherwise nothing renders) the resulting render is less bumpy but unfortunately the noisy vibrating displacement problem is still there.

Can anyone explain this unexpected behaviour of the surface level parameter?

[Ivaylo Katev]

I assumed this parameter would only effect the rendering of the surface but it seems that it does alter the result of a simulation.

can't believe that
there is only one point where the surface level is used during the simulation, this is when the option "cut above surface level" is used, then the result is modified before writing to file, in way to avoid some artifacts. if the chackbox is not switched on (see the output panel) then this is not the case. another option is to have two overlapped simulators, then they will interact like solids, using the surface level to determine the geometry.
about the vibrating noise - this is a known issue, i can't suggest an universal solution, but can tell you which parameters are the most important ones against it - the conservation and the SPF value (use fixed spf mode). actually the artifact appears when the liquid is too slow and the lower time scales are predisposed to it.

[GrahamMac]

Thanks for the info! I did have "Cut above surface level" switched on! Mystery resolved!

Can you say when the next update to Phoenix FD will be roughly and if it will be addressing the issue of slow motion surface instability?
Also is the next update going to improve the look of splashing one can get, since at the moment I find it very hard to get the simulation to create nice sheets of fluid. Example: http://vimeo.com/15370828 (see 14 seconds in especially)
At the moment Phoenix FD (using v1.2) tends to create tendrils or fingers of fluid that reach out from a splash.

[Ivaylo Katev]

the slow motion surface instability is still unresolved, but about the splashes the news are better