Fluids in Carrara and a plugin by Alberto is coming

1246

Comments

  • ImagoImago Posts: 5,152
    Alberto said:

    Thank you to all for your interest! The plugin is almost finished.

    I'm writing the manual. 

    Yay! laugh

  • PhilWPhilW Posts: 5,145

    Hey we are Carrara people, we normally make do without an up to date manual! Seriously looking forward to giving this a try.

  • TangoAlphaTangoAlpha Posts: 4,584

    *Excited* (I have a waterfall to make . . . .) 

  • ImagoImago Posts: 5,152
    PhilW said:

    Hey we are Carrara people, we normally make do without an up to date manual! Seriously looking forward to giving this a try.

    True, just put a .TXT file along the plugin with this simple tutorial:
    1) Run Carrara.
    2) Load your content.
    3) Run the plugin.
    4) yell "why this hasn't came before!"

    I can barely stay inside my own skin in the waiting for this... laugh

  • DartanbeckDartanbeck Posts: 21,551
    Alberto said:

    Thank you to all for your interest! The plugin is almost finished.

    I'm writing the manual. 

    yes​ yes​ 

  • PhilWPhilW Posts: 5,145
    edited December 2017

    So I am doing some beta testing for Alberto and it is all working well so far. Once I have done more testing and tried all the functions, I'll probably make a tutorial video to support this and for all potential users - It's not that complex but it does have a particular way of working that people will need to be aware of. I'm enjoying it so far and just delighted to have this working in Carrara!

    Here's a video of some of my early tests.

    Post edited by PhilW on
  • DUDUDUDU Posts: 1,945

    Very promising!

    It's difficult to do better in Realflow, I can't wait to make the comparison.

  • PhilWPhilW Posts: 5,145
    edited December 2017

    I haven't been pushing the parameters too high at the moment, but it seems to be reliable and relatively fast to calculate so far. So fluids can be more detailed than those I have used so far.

    Post edited by PhilW on
  • Wow Phil - those tests look great.

    I especially like the one with the water flowing between and around the cubes - we'll finally be able to make nice beach waves using something like that. What were the calculation times like for that one?

  • PhilWPhilW Posts: 5,145

    I've not been timing the calculations but it feels relatively quick so far. Obviously the higher you push the settings the longer it will take. Most of the sims have been perhaps a minute or two, I don't think I've yet done one that has taken over five mins. You can define the size of the calculation grid, and there is a warning on this - halving the size of the grid increaes calculation times and memory by a factor of 8 (as you might expect - 2x2x2). You can of course do lower res tests to set up a scene and then do a higher res version for final images.

  • PhilWPhilW Posts: 5,145

    I spoke too soon - current sim (with higher settings) is taking around 30 mins.

  • ImagoImago Posts: 5,152
    PhilW said:

    So I am doing some beta testing for Alberto and it is all working well so far. Once I have done more testing and tried all the functions, I'll probably make a tutorial video to support this and for all potential users - It's not that complex but it does have a particular way of working that people will need to be aware of. I'm enjoying it so far and just delighted to have this working in Carrara!

    Here's a video of some of my early tests.

    Oh, boy! OH, boy!! OH, BOY!!!
    Looks... It's... *drops from the chair uncounscious*

     

     

    Can't wait to lay my hands on it! laughcheeky

  • DartanbeckDartanbeck Posts: 21,551

    yes

  • ImagoImago Posts: 5,152
    PhilW said:

    I've not been timing the calculations but it feels relatively quick so far. Obviously the higher you push the settings the longer it will take. Most of the sims have been perhaps a minute or two, I don't think I've yet done one that has taken over five mins. You can define the size of the calculation grid, and there is a warning on this - halving the size of the grid increaes calculation times and memory by a factor of 8 (as you might expect - 2x2x2). You can of course do lower res tests to set up a scene and then do a higher res version for final images.

    I was re-reading your post and got a little question (sorry if I ask here, but I'm too eager! blushcheeky)
    The smaller the "domain" the harder the calculations? Shouldn't be the opposite?

  • PhilWPhilW Posts: 5,145

    I see where your confusion lies, I could have phrased it better!  If you have the fluid domain set to 10 ft x 10 ft x 10 ft, with a grid size of 0.5 ft, then the calculation size will be 20 units by 20 units by 20 units (so 8000 blocks - would you call them voxels?).  If you halve the size of the grid (keeping the fluid domain size the same) then the calculation size is now 40 units by 40 units by 40 units - so 64,000 voxels.

    I hope that makes it clearer.

  • ImagoImago Posts: 5,152
    edited December 2017
    PhilW said:

    I see where your confusion lies, I could have phrased it better!  If you have the fluid domain set to 10 ft x 10 ft x 10 ft, with a grid size of 0.5 ft, then the calculation size will be 20 units by 20 units by 20 units (so 8000 blocks - would you call them voxels?).  If you halve the size of the grid (keeping the fluid domain size the same) then the calculation size is now 40 units by 40 units by 40 units - so 64,000 voxels.

    I hope that makes it clearer.

    Yes, now is clearer, thanks! wink

    I guess simulation times won't be longer than the ones we already have with "plain" system, anyway!

    Post edited by Imago on
  • DartanbeckDartanbeck Posts: 21,551
    Imago said:
    PhilW said:

    I see where your confusion lies, I could have phrased it better!  If you have the fluid domain set to 10 ft x 10 ft x 10 ft, with a grid size of 0.5 ft, then the calculation size will be 20 units by 20 units by 20 units (so 8000 blocks - would you call them voxels?).  If you halve the size of the grid (keeping the fluid domain size the same) then the calculation size is now 40 units by 40 units by 40 units - so 64,000 voxels.

    I hope that makes it clearer.

    Yes, now is clearer, thanks! wink

    I guess simulation times won't be longer than the ones we already have with "plain" system, anyway!

    Never underestimate a user's power to bring their own computer to its knees!

  • StezzaStezza Posts: 8,051

    Our very own manuel ... how cool smiley

  • PhilWPhilW Posts: 5,145
    edited December 2017

    In case people haven't seen the video done by Ryan L. Guy to promote this plugin - this is not Carrara specific and I don't think we have the "lego blocks" render option, but it gives an idea of the scope of what you can do with this. I'll have more Carrara examples in the next couple of days.

    Post edited by PhilW on
  • I wantz nows heart

  • ImagoImago Posts: 5,152

    In the video I see meshes becoming fluids... Looks interesting, REALLY interesting!
    Little question: Can some kind of force be applied to those meshes? For example a char that hurls a tomato against a wall.

    If we can do such a thing, added to the ability to have adehesive fluids (heart!!!) many of us will...*Drops on the floor drooling*

  • PhilWPhilW Posts: 5,145

    Just a quick overview.

    All the simulation takes place in a "fluid domain" which the user can define. You can also define the scale of grid that is used - the finer the grid, the more precise the simulation will be but at the expense of time and memory.

    All objects to be included in the sim must be parented to the Fluid Domain. This activates an option in the Effects tab to turn any mesh into a liquid. So the simplest simulation would be to define a domain, put one object in there (a primitive or something more elaborate), turn that into a fluid and then run the simulation.  Any object not turned into a liquid will be a barrier to the flow.

    There are also sources and sink shapes that you can define. A source could be something like a tap or a hose and you can keyframe the flow. A sink will absorb the liquid and so act as a plughole or drain.

    In addition to the main fluid simulation, you can activate Diffuse Particles which is a post-processing simulation to add the appearance of spray, bubbles and foam. This can add a lot of realism to the simulation.

    You can define a number of forces in addition to the main gravity force (which itself can be adjusted). The added forces can be directional, point, damping and rotational. I've not really played with these yet but they seem to add a lot of scope.

    You can add a force to static objects to either attract or repel the fluid.  Adding the attraction force can make the fluid stick to the surface.

    I think that covers the main features to give you an idea of the scope of the plugin.

  • ImagoImago Posts: 5,152
    PhilW said:
     

    You can add a force to static objects to either attract or repel the fluid.  Adding the attraction force can make the fluid stick to the surface.

    Only on static objects?
    No moving ones or animated chars?

  • DUDUDUDU Posts: 1,945
    PhilW said:
    ...

    In addition to the main fluid simulation, you can activate Diffuse Particles which is a post-processing simulation to add the appearance of spray, bubbles and foam. This can add a lot of realism to the simulation.

    "is a post processing...": is that mean the particles don't inter-act with the fluid?

  • AlbertoAlberto Posts: 1,436
    Imago said:
    PhilW said:
     

    You can add a force to static objects to either attract or repel the fluid.  Adding the attraction force can make the fluid stick to the surface.

    Only on static objects?
    No moving ones or animated chars?

    Yes, you can add a force to animated objects, even chars. The Ryan L. Guy simulator itself doesn''t move the solids objects, for its point of view, are static; but for the Carrara user they are animatables. The surface force is intended to get more realistic simulations.

     

    DUDU said:
    PhilW said:
    ...

    In addition to the main fluid simulation, you can activate Diffuse Particles which is a post-processing simulation to add the appearance of spray, bubbles and foam. This can add a lot of realism to the simulation.

    "is a post processing...": is that mean the particles don't inter-act with the fluid?

    The post-processing is an internal feature of the simulator, as a second simulation. Therefore, the diffuse particles are totally dependent on the first simulation results, so to speak.

     

  • McGuiverMcGuiver Posts: 219

    I can hardly wait....lots of ideas are running through my mind.smiley

  • DUDUDUDU Posts: 1,945

    Hi Mc Guiver, I am sure that when you enter this feature, you will be fascinating to this system!

    I use Realflow for two years and to have this option in Carrara really exite me!

  • PhilWPhilW Posts: 5,145

    Sorry for any confusion from my description but Alberto has addressed the issues raised very well.

    The diffuse particles to add spray, bubbles and foam creates 3D objects fitting with the simulation of the main mass of fluid. I don't think you will be disappointed!

    I used the term static objects to differentiate from the fluid, but yes they can be animated too.

  • ImagoImago Posts: 5,152
    Alberto said:

    Yes, you can add a force to animated objects, even chars. The Ryan L. Guy simulator itself doesn''t move the solids objects, for its point of view, are static; but for the Carrara user they are animatables. The surface force is intended to get more realistic simulations.

    Yay! This makes me even more eager to try your plugin! laugh
    Your plugin opens us to numberless options!

  • PhilWPhilW Posts: 5,145

    Here are a second batch of rendered examples that I have done in beta testing Alberto's wonderful plugin. There is something very satisfying about doing these. Most of these include the extra spray, bubbles and foam effects that are also included.

Sign In or Register to comment.