FLUIDOS II for Daz Studio - update 2.2 [commercial]

runidjurhuus

Hi :) hoping someone could give me a tip on a shot I'm working on for a short.

I've got a woman under water in a bathtub, she's coming up into a seating position and I want the water to "slide" possibly "bounce" off her body in a realistic way.

here's a viewport render of the scene uploaded to youtube, I shaded the water gray so you can see it better.

any ideas on how I can improve on this, my settings are:

Fluidos Domain Liquid

FPS: 30

Cell size: 1.75

Sub level: 2.50

Marker Particle Scale: 2.5

Subsurface adapt: 0.050

Enable moving obstacle: ON

Every thing else is default

The woman is set as solid obstacle body force ON, intensity to 10 and range force set to 2

any help would be appreciated :)

Alberto

runidjurhuus said:

Hi :) hoping someone could give me a tip on a shot I'm working on for a short.

I've got a woman under water in a bathtub, she's coming up into a seating position and I want the water to "slide" possibly "bounce" off her body in a realistic way.

here's a viewport render of the scene uploaded to youtube, I shaded the water gray so you can see it better.

any ideas on how I can improve on this, my settings are:

Fluidos Domain Liquid

FPS: 30

Cell size: 1.75

Sub level: 2.50

Marker Particle Scale: 2.5

Subsurface adapt: 0.050

Enable moving obstacle: ON

Every thing else is default

The woman is set as solid obstacle body force ON, intensity to 10 and range force set to 2

any help would be appreciated :)

Hi!

Try this to simulate adherence:

https://youtu.be/h3m9LkFyAzM 

runidjurhuus

Thanks Alvin

Must have missed this video somehow, I thought I had seen all your uploads, will give it a try, thank you :)

Alberto

runidjurhuus said:

Thanks Alvin

Must have missed this video somehow, I thought I had seen all your uploads, will give it a try, thank you :)

You're welcome, Runidjurhuus!

If you have Fluidos II complete edition, you can use the filters also to refine the results. They're OpenVDB filters. OpenVDB was used in film production.

Here is an example of "The Croods". They applied a dilation filter, next a smooth filter (as mean, gaussian, median, laplacian or mean curvature), and finally an erosion filter. You can find all of them in Fluidos II.

Moreover, the adaptive meshing can reduce considerably the polycount with values as small as 0.001.

https://www.openvdb.org/documentation/

https://artifacts.aswf.io/io/aswf/openvdb/openvdb_production_2015/1.0.0/openvdb_production_2015-1.0.0.pdf

ramasys101_f0ccb71062

Hi,

 Can .vdb files be imported from other programs (realflow in my case)

thanks.

Alberto

ramasys101_f0ccb71062 said:

Hi,

 Can .vdb files be imported from other programs (realflow in my case)

thanks.

Fluidos II can read Level set class vdb files. It will try to import the data. But I've not tested Realflow vdb. Do you have a .vdb file I can test? 

ramasys101_f0ccb71062

Try this https://file.io/qLdfktjio9WZ

Alberto

ramasys101_f0ccb71062 said:

Try this https://file.io/qLdfktjio9WZ

Thank you!

Fluidos II cannot read the file. I happens that the vdb class file  has grids of "GRID_UNKNOWN", instead of "GRID_LEVEL_SET". 

Anyway, I'm going to update Fluidos II soon, and I will take this into account. I'll let you know.

 

Aidolatrie

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Alberto

Aidolatrie said:

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Hi!

Here is the .duf file:

https://drive.google.com/file/d/1tINsmJR24FmVKLCVXnp9ef4M9NPK0d_Y/view?usp=sharing 

Let me know if you have more trouble with the settings.

Aidolatrie

Alberto said:

Aidolatrie said:

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Hi!

Here is the .duf file:

https://drive.google.com/file/d/1tINsmJR24FmVKLCVXnp9ef4M9NPK0d_Y/view?usp=sharing 

Let me know if you have more trouble with the settings.

First of all, thanks very much for sharing the .duf file.

Second, after much painstaking studying and subsequent tweaking of settings, I was able to create an animated scene of a girl walking on snow in heels. Attached is a rendered still of it.

Although I was largely pleased of the result, the footprints, especially the deeper ones, still look noticeably blocky despite running a ~13 hour simulation with cell size of 0.5 cm and resultant file size of ~80 GB.

Is there a way or a few to smoothen the footprints so that the gradation of depth into the snow looks much smoother?

Thanks very much.

Alberto

Aidolatrie said:

Alberto said:

Aidolatrie said:

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Hi!

Here is the .duf file:

https://drive.google.com/file/d/1tINsmJR24FmVKLCVXnp9ef4M9NPK0d_Y/view?usp=sharing 

Let me know if you have more trouble with the settings.

First of all, thanks very much for sharing the .duf file.

Second, after much painstaking studying and subsequent tweaking of settings, I was able to create an animated scene of a girl walking on snow in heels. Attached is a rendered still of it.

Although I was largely pleased of the result, the footprints, especially the deeper ones, still look noticeably blocky despite running a ~13 hour simulation with cell size of 0.5 cm and resultant file size of ~80 GB.

Is there a way or a few to smoothen the footprints so that the gradation of depth into the snow looks much smoother?

Thanks very much.

Hi!

To smooth the footprints, you can reduce the viscosity. Or, better still, you can use the filters (increase the iterations); this is much faster because you don't have to redo the simulation, only havo to remesh it.  The gaussian filter is the stronger. The mean is similar, but milder. You can increase the footprint width by using the erode filter.

I suggest to use a geoshell of the shoes as the solid obstacle. This way, you don't need to reduce too much the cell size.

 

 

emu42

Hello,

I am trying something a bit unusual (unsupported?) with the Fluidos Mesher. After running the simulation I applied D-Formers to make some manual adjustments to the mesh. That seemed to work at first, but after reloading the scene they no longer seem to have any effect on the mesh.

Is there a way to a) make this work or b) convert the mesh into static geometry that can then be edited?

Alberto

emu42 said:

Hello,

I am trying something a bit unusual (unsupported?) with the Fluidos Mesher. After running the simulation I applied D-Formers to make some manual adjustments to the mesh. That seemed to work at first, but after reloading the scene they no longer seem to have any effect on the mesh.

Is there a way to a) make this work or b) convert the mesh into static geometry that can then be edited?

Hello, Emu42

For a) and b), the answer is yes.

1. Using D-Formers is possible. But, sometimes, after reloading the saved scene, the D-Former disconnects from the mesher. To re-connect it, you can use the scrip I attached. Apply it to the D-Former Base.
2. To convert to static geometry export the mesher as .obj and then import the .obj to get the static geometry.

emu42

Alberto said:

Hello, Emu42

For a) and b), the answer is yes.

1. Using D-Formers is possible. But, sometimes, after reloading the saved scene, the D-Former disconnects from the mesher. To re-connect it, you can use the scrip I attached. Apply it to the D-Former Base.
2. To convert to static geometry export the mesher as .obj and then import the .obj to get the static geometry.

Thanks, Alberto! I went with the second option this time. The solution seems so obvious now, I should have thought of this myself. :)

 

Mark_e593e0a5

I am currently moving my DAZ install from Mac to a Windows 10 rig. While Gescon and Parsis work as expected, I get an error with Fluidos II, even with a blamk DS install (just DIM, Daz Studio and inital compionents, plus Fluidos II)

The error I get is a DLL that cannot be loaded

I have checked, an the file seems to be there:

As stated, this is a very plain install. The only thing that was different from a 4.12 DS install on Windows is that Fludios II tries to install a Visual C++ runtime, and that installation stops beciuase a similar runtime is already installed.

What's going wrong here?

Alberto

Mark_e593e0a5 said:

I am currently moving my DAZ install from Mac to a Windows 10 rig. While Gescon and Parsis work as expected, I get an error with Fluidos II, even with a blamk DS install (just DIM, Daz Studio and inital compionents, plus Fluidos II)

The error I get is a DLL that cannot be loaded

I have checked, an the file seems to be there:

As stated, this is a very plain install. The only thing that was different from a 4.12 DS install on Windows is that Fludios II tries to install a Visual C++ runtime, and that installation stops beciuase a similar runtime is already installed.

What's going wrong here?

Do you have OpenCL installed in you system? (is the GPU driver updated?)

 

Mark_e593e0a5

Killing the GPU driver and re-insztalling fixed the issue. Thanks a lot 

 

Alberto

Mark_e593e0a5 said:

Killing the GPU driver and re-insztalling fixed the issue. Thanks a lot 

 

You're welcome. 

Aidolatrie

After additional experimentation @Alberto, I found a workaround to create smoother footprints. However, that required me to run another fluid simulation to create a new .vdb file, but this time with the "cube" set lower to not generate footprints. Afterwards, I executed a union of the two .vdb files (one with footprints and the other without) to create a another .vdb file that features the combined fluid mesh as seen in the attached image.

Although I am pleased with the end result, the time, effort, and resouces are too expensive for me to do this on a regular basis. Nevertheless. I did learn a lot (mostly the hard way), and thanks for all the help.

Alberto said:

Aidolatrie said:

Alberto said:

Aidolatrie said:

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Hi!

Here is the .duf file:

https://drive.google.com/file/d/1tINsmJR24FmVKLCVXnp9ef4M9NPK0d_Y/view?usp=sharing 

Let me know if you have more trouble with the settings.

First of all, thanks very much for sharing the .duf file.

Second, after much painstaking studying and subsequent tweaking of settings, I was able to create an animated scene of a girl walking on snow in heels. Attached is a rendered still of it.

Although I was largely pleased of the result, the footprints, especially the deeper ones, still look noticeably blocky despite running a ~13 hour simulation with cell size of 0.5 cm and resultant file size of ~80 GB.

Is there a way or a few to smoothen the footprints so that the gradation of depth into the snow looks much smoother?

Thanks very much.

Hi!

To smooth the footprints, you can reduce the viscosity. Or, better still, you can use the filters (increase the iterations); this is much faster because you don't have to redo the simulation, only havo to remesh it.  The gaussian filter is the stronger. The mean is similar, but milder. You can increase the footprint width by using the erode filter.

I suggest to use a geoshell of the shoes as the solid obstacle. This way, you don't need to reduce too much the cell size.

 

 

Alberto

Aidolatrie said:

After additional experimentation @Alberto, I found a workaround to create smoother footprints. However, that required me to run another fluid simulation to create a new .vdb file, but this time with the "cube" set lower to not generate footprints. Afterwards, I executed a union of the two .vdb files (one with footprints and the other without) to create a another .vdb file that features the combined fluid mesh as seen in the attached image.

Although I am pleased with the end result, the time, effort, and resouces are too expensive for me to do this on a regular basis. Nevertheless. I did learn a lot (mostly the hard way), and thanks for all the help.

Alberto said:

Aidolatrie said:

Alberto said:

Aidolatrie said:

Hi @Alberto,

I'm trying to replicate the scene depicting the Genesis figure walking on snow in Fluidos II (by using the file you provided that depicts the genesis figure walking on water instead as a starting point, see here). Even after turning on the viscosity and setting it to 500,000, I still cannot produce the same look and feel of the footprints as depicted in the Fluidos I promo image and video.

Can you tell me what additional settings you used for the Domain and Cube? Can you also provide the .duf file of the scene? I tried looking through the provided tutorials and recipies, but no progress.

Thanks for your time.

Hi!

Here is the .duf file:

https://drive.google.com/file/d/1tINsmJR24FmVKLCVXnp9ef4M9NPK0d_Y/view?usp=sharing 

Let me know if you have more trouble with the settings.

First of all, thanks very much for sharing the .duf file.

Second, after much painstaking studying and subsequent tweaking of settings, I was able to create an animated scene of a girl walking on snow in heels. Attached is a rendered still of it.

Although I was largely pleased of the result, the footprints, especially the deeper ones, still look noticeably blocky despite running a ~13 hour simulation with cell size of 0.5 cm and resultant file size of ~80 GB.

Is there a way or a few to smoothen the footprints so that the gradation of depth into the snow looks much smoother?

Thanks very much.

Hi!

To smooth the footprints, you can reduce the viscosity. Or, better still, you can use the filters (increase the iterations); this is much faster because you don't have to redo the simulation, only havo to remesh it.  The gaussian filter is the stronger. The mean is similar, but milder. You can increase the footprint width by using the erode filter.

I suggest to use a geoshell of the shoes as the solid obstacle. This way, you don't need to reduce too much the cell size.

 

 

You're welcome.

I'm working now in an update that wiil speed up Fluidos II. 

mikethe3dguy

For rising smoke animation (from a cigarette), would the Lite Edition product work, or do I need the full Fluidos II product?

Alberto

mikethe3dguy said:

For rising smoke animation (from a cigarette), would the Lite Edition product work, or do I need the full Fluidos II product?

The Lite Editon will work. 

mikethe3dguy

Alberto said:

mikethe3dguy said:

For rising smoke animation (from a cigarette), would the Lite Edition product work, or do I need the full Fluidos II product?

The Lite Editon will work. 

Thanks. I ended up getting the full version since I realized I may be trying more difficult simulations later. I'm trying to work out how to create an effective smoke simulation for the application I described. In the manual there seems to be a missing step where you step the user through their first smoke sim on page 5: step 6 mentions "the sphere" so it looks like a sphere primitive is needed somewhere. Should that be placed above the Source, or if not: where? Also, any tips you could provide to help me create smoke similar to a burning cigarette would be appreciated.

Fun product! Thanks!

Alberto

mikethe3dguy said:

Alberto said:

mikethe3dguy said:

For rising smoke animation (from a cigarette), would the Lite Edition product work, or do I need the full Fluidos II product?

The Lite Editon will work. 

Thanks. I ended up getting the full version since I realized I may be trying more difficult simulations later. I'm trying to work out how to create an effective smoke simulation for the application I described. In the manual there seems to be a missing step where you step the user through their first smoke sim on page 5: step 6 mentions "the sphere" so it looks like a sphere primitive is needed somewhere. Should that be placed above the Source, or if not: where? Also, any tips you could provide to help me create smoke similar to a burning cigarette would be appreciated.

Fun product! Thanks!

Sorry, the step 6 is an error, ignore it, please.

Attached is the .duf of a smoke scene similar to the one in the Promo image (the burning cigar). There are two meshers, you can enable one of them (or both, if you prefer). One will show smoke as particles, the other one as fluid mesh. Se the same Baked files folder to the Fluidos Domain and the meshers.

If you have any doubts, let me see please.

Thanks to you!

mikethe3dguy

 

Sorry, the step 6 is an error, ignore it, please.

Attached is the .duf of a smoke scene similar to the one in the Promo image (the burning cigar). There are two meshers, you can enable one of them (or both, if you prefer). One will show smoke as particles, the other one as fluid mesh. Se the same Baked files folder to the Fluidos Domain and the meshers.

If you have any doubts, let me see please.

Thanks to you!

This helps a lot Alvin, thanks. I need to:

1) Extend the animation to 2 seconds (done)

2) Reduce the velocity (played with Velocity (y) but it's not having any noticeable effect - whether it's 50, 100 or 1 I can't see any difference, will try reducing temperature next)

3) Reduce the size of the source so I get a amaller point of emission

4) Last, figure out how to integrate this into my scene. I'm assuming I can merge the Fluidos scene into my existing scene, combine the Fluidos objects into one group, move the group to the needed x,y,z position, run simulation, and render?

 

Thanks

mikethe3dguy

Alvin,

Reducing the vertical rise velocity is proving a challenge. Temperature seems to affect it but not in always predictable ways. Smoke rising much too rapidly still.

I can reduce the emission source size using Radius, but if I adjust much below 3 (cm?) there is no smoke emitted at all.

Alberto

mikethe3dguy said:

4) Last, figure out how to integrate this into my scene. I'm assuming I can merge the Fluidos scene into my existing scene, combine the Fluidos objects into one group, move the group to the needed x,y,z position, run simulation, and render?

Yes, it's correct.

mikethe3dguy said:

Reducing the vertical rise velocity is proving a challenge. Temperature seems to affect it but not in always predictable ways. Smoke rising much too rapidly still.

I can reduce the emission source size using Radius, but if I adjust much below 3 (cm?) there is no smoke emitted at all.

Using the Beta bouyancy factor (At Smoke and fire properties of the Fluidos Domain) is an easier form to control the raising velocity. The default value is 1.0, if you reduce it, the raising velocity will decrease too.

 

mikethe3dguy said:

I can reduce the emission source size using Radius, but if I adjust much below 3 (cm?) there is no smoke emitted at all.

You have to reduce the cell size a little below the Radius. However, the lower the cell size, the slower the simulation.

Another option is to change the Geometry type of the Source to Cuboid, this type allows smaller sizes of the source without reducing the celll size of the Domain

mikethe3dguy

Okay thanks! Is there anything else that will affect the rise speed as well - is there any relationship with Velocity (y) for example? I haven't fully explored it yet but so far reducing Beta Buoyancy reaches a point where much of the smoke falls below the source, similar to reducing temperature. Having a hard time reaching any kind of sweet spot.

mikethe3dguy

Hmm, never occurred to me before, but I just took the timeline of the simulation and increased it from 60 frames to 120. Then I dragged the "100% completion" key at frame 60 to new frame 120. Instant half-speed animation. So it seems that all I really need to do to adjust the simulation speed is fiddle with the completion key at the end of the timeline; either what frame it is placed at, or the percentage value, or both. Simple.