Flickering Surface and other liquid-related problems

[Ivaylo Katev]

i got the reason, apparently an old problem that we had already fixed is rising again.
in the old versions of phoenix, a small piece of liquid was able to act as infinite source, this is because the classic advection is not preserving the liquid amount. it seems like this is happening again and the droplets stock to the wall are producing liquid, that falls and hits the surface, that is too fast and you have to observe the movement very carefully to see it.
beside this - decrease the liquid sharpness, this closeup scene makes the grid stamp very visible. you can even try to use strong surface mode, it was created especially to remove the grid stamp.
some surface tension also may be added, it plays important role in the small scale liquid behavior.

[MANUEL_MOUSIOL]

i got the reason, apparently an old problem that we had already fixed is rising again.
in the old versions of phoenix, a small piece of liquid was able to act as infinite source, this is because the classic advection is not preserving the liquid amount. it seems like this is happening again and the droplets stock to the wall are producing liquid, that falls and hits the surface, that is too fast and you have to observe the movement very carefully to see it.
beside this - decrease the liquid sharpness, this closeup scene makes the grid stamp very visible. you can even try to use strong surface mode, it was created especially to remove the grid stamp.
some surface tension also may be added, it plays important role in the small scale liquid behavior.

alright, thanks for the advice and the insight. I will play around with those values. For me it is hard to know which is a bug, a coding problem or a user problem. I love that you guys are always kindly there to help! very much appreciated (I wouldnt be able to do it without this support).

But I guess its good we are having this communication, seeing that you can find some problems that need to be fixed - a little like beta testing

[MANUEL_MOUSIOL]

I am sorry, but I am still having trouble with the "spikes" at the wall of the glass, and the gizmo just wont cut them off, not matter how thick I make the glass and push it into the liquid.
any ideas? i tried setting the surface tension down, set it once to zero. I tried the solid surface mode...not sure how I should resolve this..

[Ivaylo Katev]

these spikes are produced by the mesh smoothing. yes it sounds like a paradox, but this is the reason. the mesh smoothing has edge detection and does not smooth the edges. when you have an edge that is near to the detection threshold, some parts are not smoothed and there are the spikes. just set smoothing zero or use implicit rendering. i recommend the second option, because you will have a bonus beside - working gizmo. currently the mesher does not support the gizmo option, you can use a boolean object to have the same effect, but the implicit surface is better.

[MANUEL_MOUSIOL]

Ok, thank you.
Is there any more information about the implicit mode? What benefits are you talking about?
Am I just supposed to change it in the rendering options and thats it? beacause it gives me weird results.
take a look:
mesh mode

implicit mode wit gizmo

implicit without gizmo

[Ivaylo Katev]

don't forget that the gizmo by default cuts the content outside, not the internal one. the glass case is very misleading, because the common sense says that the liquid is inside the glass but it is not.
topologically the liquid is outside (it is NOT in the glass mesh), and you have to use the "invert gizmo" option

[MANUEL_MOUSIOL]

yeah, i tried that but gave me the same result as without gizmo. do you need a screenshot or so?

[Ivaylo Katev]

yes, would be good. attach both the results with the render settings visible in the image

[MANUEL_MOUSIOL]

It is the same scene as I posted in the thread about water is disconnecting with finer grid.
Here are some screens:

[Ivaylo Katev]

seems like your liquid does not penetrate into the glass, this is an essential condition to have working gizmo.
why i think this? in the first image you are rendering in mesh mode,and the glass walls are visible. see this section of the vray help, there you can find an explanation why the liquid must overlap the glass. your image looks like this example , but it should look like this if the liquid overlaps the walls
btw the gizmo option does not affect the mesh rendering mode, it is only for the implicit surface.