Yes. Blender scale has nothing to do with it, the real world value is what says how much water is in the domain. (i.e. a cup sized amount vs. a bucket sized amount vs. a swimming pool sized amount.)
[Edit]
If your frame-rate is 25 frames per second, and ending time is 4.0 seconds, then the baking will end at frame 100.
Not exactly true. The baking will end at frame 100 if you have set Blender to make 100 frames. The simulation will always compute over however many frames Blender is set to animate for. So if you set it up with 4 frames, the entire 4 seconds would simulate over 4 frames.
i get it. two pictures would be great of a coffee cup (100mm across) being filled by resolution X and another with liquid at resolution Yā¦ (hint hint)
Cool! If I upload renders with a more subsurfed handle could you replace those?:o
As for the bake time, it is so dependent on system RAM, CPU, and the scene being simulated that it would hardly be fair to post benchmark times. It really can vary that much.
you know, fluid simulations can be much better, even with a low resolution, if u use particlesā¦ http://blenderartists.org/forum/showthread.php?t=96105
that fluid sim had a resolution that was lower than 100, i think and yet the results looked like it was baked at a high res.
Particles are one part of the fluid sim that I donāt have a handle on yet, perhaps you could fill me in on that Jeepster? I tried using them and just got a white haze that blocked my fluid from viewā¦
take a look at the .blend of the scene i had http://dominomf.com/smcfiles/Jeepster%5B%5D_fluidsimtest.blend
i find that the settings i used there usually work for all of my fluid sims
note that thereās a small plane somewhere in there for the particle settings
It is the second time i download a .blend to renderā¦ and i donāt get else than a single blank or flat image: 1black scare with Jeepsterās file this time.
Any special setting to make before trying to render?
:o How is Yafray (something like a independant renderer plugin or so? ) to getā¦ and is it OSX compatible?
:eek::o Baking the fluid firstā¦ humā¦! To wich temperature should my oven be set?
More seriously: i clicked on ābackā for a tryā¦ but iām ignorant of that processā¦ so all i got was an error messageā¦
I havenāt managed to bake a fluid that takes more than 1GB, not even with LargeAddressAware build. I guess it is a limitation of current fluid engine. Also, note that first frame may well take only a few hundred MBs but latter ones go over one GB. Thatās why your bake fails. Try to bake one frame at the middle or end at full res so youāll see if it works.
I have noted that used memory doubles when you set surface subdivision from 1 to 2. I think using āgenerate particlesā will also add memory requirements, as they increase surface area and complexity. Ordinary fluid-sim generated particles probably eat less memory.
Just to fix my outlandish&exaggerated claims, changing subdivision from 1 to 2 doesnāt really double the memory (at least in simple cases), but it does add to it.
I made a quick&simple test case, and on resolution 100 Blender said baking would take 97MB. Actual memory used by Blender executable during single frame bake was 177MB (naturally not all of this is used by fluidsim, but Blender with empty scene takes around 23MB). This was with subdivision 1 and particles at 3. With subdivision on 2 actual memory was 244MB, which is around 28% more. I presume this will go up if you use low viscosity fluids, which tend to shoots lots of particles.
Blender fluids use adaptive grid, so it may be that complex scenes with obstacles need more memory than Blender thinks. I donāt know how fluid sim code actually uses memory so I canāt say for sure.
Generally, if you are approaching 1GB in task manager you may want to tune down surface subdivision level and baking resolution.