Liquids Test Thread

[Polinalkrimizei]

Ok here we go.
I used whaat's second OJ mat, cause it seems to be more "fleshy". Not sure if anybody gets what I mean...

I tried to make a more dynamic testscene, and then it happened: violet artifacts in 2.0.9!!! Where does this pink come from? Ono I thougt that was just an issue with 1.1.18...

I suspect cauchy b, so the next render is baking without it. And with a white background. Black is more elegant, but white is ... more common. I would need a lot more computer power now to test all the liquids

[Whaat]

YAOJR:

Rendered to about 15K spp. Using the material settings I posted earlier (not the material from the MDB). I think this material is a bit more accurate than the one I uploaded to the MDB

Also, a PHOTO of OJ for reference.

[Polinalkrimizei]

Whaat, I definitely like this material. Looks pretty realistic to me!

[OnoSendai]

Whaat:
I think in your last render, the glass medium has too much absorption, at least compared to the reference image.

[Whaat]

Whaat:
I think in your last render, the glass medium has too much absorption, at least compared to the reference image.

Actually, my absorption for the glass medium is zero.. I suppose there could be normal smoothing artifacts from not having a smooth enough mesh. I have been meaning to render this scene with an older version of Indigo to see if some of this is due to the self-intersection code that was changed.

[OnoSendai]

Oh dear
Can you send the scene to me (email or p.m. it to me), so I can check it out in the current build?
cheers.

[Polinalkrimizei]

OJ is still not clear, looking dirty... I am @18600ssp at the moment.

I also tried pixies merlot from the matDb on the laptop. Looks ok to me, but perhaps a little bit more red component? It is also not finished yet, but hey, this is supposed to be a "test"-thread!
Gonna go to the beach now...

[Whaat]

@ Polinalkrimizei: That's a really cool liquid scene! I think with some tweaking it would make an awesome liquid preview scene. Would you mind sharing the .blend?

All the liquids look a bit strange as they splash inside the glass. I wonder if your meshes are intersecting the glass properly and if the precedence settings are correct?

@ono: you have a PM

[Polinalkrimizei]

@ Whaat: Thanks! I'll share the .blend tomorrow, I'm at another computer right now. The liquid is intersecting the glass properly, but the precedence of the orange juices was wrong indeed!!
And I wondered why it looks so bad...
So I'll share the .blend tomorrow with the liquid as a mesh, and in case anybody wants, the liquid as a already-set-up-fluidsim. Whaat, do you have the power of the 64 bits? I can't calculate the fluid more accurately cause of mem restrictions. Higher resolution = cleaner fluid mesh! Anybody interested?

[Doug Armand]

Whaat, do you have the power of the 64 bits? I can't calculate the fluid more accurately cause of mem restrictions. Higher resolution = cleaner fluid mesh! Anybody interested?

I have a x64 system and 8MB RAM - not uber but I can give it a try if you want?.

[Godzilla]

Whaat, do you have the power of the 64 bits? I can't calculate the fluid more accurately cause of mem restrictions. Higher resolution = cleaner fluid mesh! Anybody interested?

I have a x64 system and 8MB RAM - not uber but I can give it a try if you want?.

Going to have 64Bit Vista and 8GB of ram in about 3 days (when it arrives)... There will be awesome fluid sims that ensue.

I challenge you to a duel, sir..... of.. fluid.. simulation?

[Doug Armand]

I challenge you to a duel, sir..... of.. fluid.. simulation?

challenge accepted!

[PureSpider]

I challenge you to a duel, sir..... of.. fluid.. simulation?

Ewwwwwww nasty...

[Polinalkrimizei]

Whoaaa, this is gonna be exciting! Thanks guys!
I had to be at my job this morning at 6am, so competition starts when I'm back home tonight.
@Doug: I can't wait for what you are gonna produce out of 8megs if RAM...
@Godzilla: Nice, a new system, if this means even less restrictions to your creativity, we are gonna see A LOT of great output!

My scene is actually quite simple to set up, so if you can't wait, just place an inflow object on top of the glass, and blender will do the rest...

[CTZn]

Higher resolution = cleaner fluid mesh!

Yes, and if you can't run the simulation at an higher resolution, at least Indigo can subdivise the mesh further, this should be a reflex every time you need more triangles to render !

Curvature threshold should be a key parameter in this case.

edit: and the glass top edge profile curve shoule be made of 6 points at least to avoid bad modeling artifacts