Tips and Tricks: A typical HYBRIDO scene using displacement

Hi all,

Here we have how to make a HYBRIDO scene using displacement.

1 GridFluidDomain (in surface tab put the Calculate surface to Always).

2 Put the Displacement tab in the GFD parameters to “Calculate displacement: Always”. The configuration for the displacement parameters is not important yet.

3 Simulate a couple of frames for the GridFluidDomain.

4 Add a GridMesh and put the parameter “Shader: Displacement”. Build the mesh.

5Return to the Displacement tab in the GFD and change the displacement parameters as you want. The viewport will show the changes so it is not necessary to simulate again. (Keep in mind what value you put for the Detail Threshold in the Surface tab for the GFD so you’ll need to apply the same when generating the foam, if not it may be displaced).

6 Put the Calculate Surface in the GFD to Always and change the Source to “From particles”.

7 You can modify the auto radius, smooth and auto cell size to calculate the mesh. Click on the “Build grid button” in the surface tab to check if the mesh is working.

8 Once you have a Displacement and the mesh you want, you should switch off the GridMesh.

9Next, simulate the GridFluidDomain changing again the Source to “From Field”  and keeping the Calculate Surface to “Always”. Once it is finished, put the GFD to cache mode and change the Source to “From particles”.

10 Now, switch on the GridMesh and simulate again. (note you can use the calculate mesh button too)

11 Once the GFD and the GridMesh is simulated, put the “Calculate Surface” parameter in the GFD to cache, the “Surface to” parameter to “From Field” and the GFD node to cache too. Switch off the GridMesh.

12 Add the splash (or splash_and_foam) emitter and put the “@use displacement” to YES. Simulate.

13 Once the splash is simulated, if you put a splash emitter put it to cache mode. Put a foam emitter (keep in mind that the foam can’t be affected by any daemon force or other emitters) and put the “@use displacement” to YES too. Simulate.

14 Put the foam in cache.

15 Once all elements have been simulated, switch on the GridMesh again to see the whole result in the viewport.

luisM.

91 comments

  1. Gracias por compartir tanta sabiduría Luis Miguel.
    Tengo una duda, al hacer mi particular escena de barco, la malla deja un espacio enorme hasta el casco del barco, he dado más resolución a las partículas pero aún asi la gridmesh se “aparta” mucho del barco.
    A que se debe?
    Gracias de nuevo por tu tiempo

    1. Hola Mikel

      He visto que tenias problemas con hybrido y el espacio que deja entre particulas y la malla, y me gustaría saber si encontraste la solución, si es así estaría muy agradecido si me pudieras echar una mano.

      un saludo y gracias.

      Carlos

      1. Hola Carlos,

        Eso es porque tienes una resolucion muy baja para el dominio y las particulas no se estan ajustando lo suficiente.

        This is cause you have a very low resolution for your Hybrido domain and the particles are not adjusted enough.

        luisM.

  2. Thank you Luis, for the wonderful set of tutorials. YOu’ve helped many of us out there, have a good start with Realflow, which would otherwise be hardly possible:
    I’m trying to create a similar scene with the above one. Everything seems promising on my part but there is always some unwanted shaking, or appearing/dissappearing of water bubbles, which I couldn’t figure out although tried almost every parameter.
    Ofcourse my resolution may be low. 125.000 particles with 3 selected for particle sampling. (That’s because I ‘m still in testing phase for several parameters, so I have to see the results of each parameter fast.)
    Can you have any recommendation to me by looking at the video files below:

    http://www.nezihkanbur.com/rf2.avi

    (I used a grid mesh here, as it has some artifacts I tried a standard particle mesh below, it has a more blobby look)

    http://www.nezihkanbur.com/rf.avi

    Grid mesh is a bit more like water in this case so I prefer using it but as you see water bubbles come and go erratically. Particles themselves are continuously moving while the mesh snaps and pops.
    Can you pinpoint the reasons by just looking at the video.

    I can’t find enough parameters for the mesh.

    Thank you

    1. Hey!

      I saw the videos. You’ll need a lot of of resolution to get a detailed mesh. I see you have very few particles and this is the cause which you have big ball in your mesh. In addition, you’ll need to play with the parameters into the surface tab by selecting the GridFluidDomain. Bear in mind you have a Build Grid Mesh button in this tab to make the mesh from there. If you want to get more resolution for your mesh, check the Auto Cell size “automatic” to NO, and put a value to try. To start to play with this value, take a look the cell size info in the viewport when you have the selected the GridfluidDomain. If it the mesh is too blobby, try to put low values for it. Be careful with this because of very low values will be wasting more RAM. Another parameter to check is the Calculate Surface. Make sure it is Always and not in cache. And for last indication, if you choose “from particles” instead of “from field”, the mesh will be more adjusted to the fluid.

      I hope it helps🙂

      luisM.

      1. Although I ‘ve seen some of your explanations on scene scale throughout this blog, it is still not clear in my head: does this scale setup affect the performance of the program, and which scale should be the most efficient?
        And can you say anything about what scale values should be set together with scene scale set to 0.01? (I mean those daemon scale, and particle scale…)

        I have a large theme park model, trying to create some waterfalls and streaming water. It’s modeled in its real scale and in metric system. 1meter is 1 meter. When I import any of my meshes, it’s invisible until I scale it down to 0.01 (or even more).
        in 0.001 water flows faster but needs more resolution, and in 0.01 water becomes slower and more particles released. I feel like they are the same in performance.
        On the physics side, smaller scale value literally exploded the particles. Effect is not the same.

        It takes a lot of time to test a single particle. I’m trying to test with the smallest amount of models at the same scale but even that takes a lot of time😦

      2. Hi!

        I always advice to scale the geometry in your 3D platform and don’t change the scale in RealFlow. Why?. Because RealFlow will work more efficienly with scale 1 that if you scale down or scale up the scene. So, how to do that?. Keep in mind that each grid square in RealFlow is like 1 meter. Then, you can create a simple Realflow cube and export it as SD. Next, import the cube.sd into Maya (or any 3D platform) and scale down your scene with regards this cube. In this way you can export your 3D scene with a correct to work in Realflow. Why to work with real scale in Realflow? It is due to if you change the scale in RF, all parameters will need to be scaled as well to work properly. So, if you increase the scene scale per 10, for example, the gravity will need to be scale per 10 as well if you want a real behavior simulation. If you need to work with high scale in your 3D platform, import the scene simulted, group it and scale up.

        I hope it helps.

        luisM.

      3. Thank you very much indeed! After some tests with simple meshes according to your instructions, I found where my scene was in realflow.🙂
        (It is extremely large)
        Although I hate scaling things in 3dsmax, I decided to rescale how much effort it requires. Now my scene is a lot more stable and resolution values are just like starting a scene from scratch.

        There is still something I couldn’t figure out: I mentioned about some unwanted appearing or dissappearing of particles in my first message.

        Looking at the progression of correctly moving particles, (particles which are moving at a constant or expected speed) there are some group of particles on the GFD which appears to be like jumping forward.
        They suddenly appear actually out of nowhere, and continue to move like normal particles but many of these appear during a boat’s slamming on water. It also resembles particles moving at 10fps, while there are others moving at 25fps.

        I’m not sure about the adequate resolution, but very high values didn’T seem to solve this issue.

        I thought this may be due to this timing precision thing (1 -333 ) I increased it and decreased also. Nothing helped.

        Do you have any ideas?

      4. Hi!

        It is difficult to understand without watching the scene or a video. My first thinking is you have an object which is moving very fast. Usually, that behaviour is due to suddenly changes in the movement for your animation. You can imagine a fish swimming for the fluid with a smooth movement, so if the fish has new movements like a robot, they could make a bad behaviour for the fluid.

        About the substeps, I think it is a common error and maybe we had not explained very well. The substeps don’t affect for the Hybrido system, they are only for SPH fluids like splashes, foam, and normal emitters like circle, cylinder, linear, etc emitters. Hybrido has ist own internal substeps when calculating the fluid.

        Show me a video with the problem if you don’t have objects with animation for your scene.

        luisM.😉

      5. If I may, just in the morning I noticed something that I need to ask:
        When I look at the GFD from the top, I saw that some cube (er let’s say cell) of particles are like arrayed and there is an obvious spece between them. Think of a grid and one cell of the grid is full with particles and the next empty like below:

        O O O O O
        O O O O O

        You already know that, but I wanted to decrease the space between them but no parameter did that. I thought some jittering parameter would correct it but it didn’t (Not to mean that it’s a problem)
        CAn’t we just jitter all the fluid? Is it all normal and has no effect on the quality?

      6. Hi Nezih,

        You have right. There is an array of particles for the Grid Fluid Domain. Each space between particles is due to the Voxel size which is representd by a cube. By default you have 8 particles for each cube so, if you increase the resolution, the cubes will be smaller but they will have 8 particles per cube yet. How to change the number of particle for each cube? Easy. Increasing the Particle sampling parameter in the GFD. By default is 2, so how it is a cube will be 2x2x2= 8 particles as we said before.

        Of course you can apply a jitter for the distribution of particles. In order to do that, select the emitter and changing the jitter value in its parameters.

        luisM.

      7. I know I’m taking a lot of time but I’m so stuck.
        You can see the preview at the following link:

        http://www.nezihkanbur.com/64.vmw

        and the realflow file with sd objects included at:

        http://www.sendspace.com/file/eil5pi

        – This is a pool of about 50×60 in scale (meters probably) 1.000.000 in resolution and 3 as sampling.
        – The proxy boat object is not a rigid body, it’s manually animated as sliding into water and floating, in 3dsmax.
        – It slides over a water canal into the pool.

        1250-1650 frame range of the pool and boat simulation is all that matters here. There is also a particle emitter which I thought would be for flowing water in the canals but I couldn’t even create the pool😦

        I don’t see the jumping particles a lot in the video here but there is a noticable snapping of the water mesh from place to place. Just like several magnets are randomly pulling it. I understand it’s due to the location of the particles. The particles are the magnets that pull the polygons as they approach to the core fluid, but which parameter will give it a smooth motion instead of shaking?

        Furthermore, I ve just noticed that simulating the last 400 frames, drained my 1 TB harddisk completely. 1 gb of particle data for each frame total of 400 frames😦 and it’s still low in resolution. How will such a simulation be handled?

        I tried realwave as I’ve seen this impossible, but it also doesn’t allow me to work easily at this size?

        If you have a chance to check or setup the scene with acceptable parameters, I have to warn you that there is a coordinate system problem with the boat mesh in the sendspace file. Although it seems to play back correctly, it frequently rotates about 90 degrees during interrruption of the simulation by viewport operations. It may be just a visual artifact I thought at first, but it affects water behaivour,too. Below is a corrected boat mesh which no longer flipped that way.

        http://www.nezihkanbur.com/boat.sd

        I appreciate if you can help me reach a reasonable quality and file sizes by any means.

        Thank you

      8. Hi Nezih,

        Sorry I can’t load the video. Could you put it again?

        I saw the scene. I would change the follows parameters.

        For the GridFluidDomain when simulating the main fluid, in the surface Tab

        surface tab

        For Splash and Foam, as the scene is huge, I advice you to put a splash emitter and a foam emitter, so you can use the trick to put the substeps to 1 when simulating foam. So the splash will be

        splash

        Let me know when you put the video again.

        luisM.

      9. sorry. I’ve spelled “wmv” wrongly🙂 above

        http://www.nezihkanbur.com/64.wmv

        should work. This preview is a bit like, mixture of several different cached ranges, but you’ll see what I mean at the shaded frames.

        Actually, after seeing that the only workaround could be increasing the resolution and my 1M res. together with 3 sampling value filled up an empty 1Tbyte HDD🙂 I decided to use the old “standard particle mesh” method instead of grid meshes, as it doesn’t have that sort of problems, however I’d still love to what’s the reason for the shaking of grid fluids.

      10. Hey,

        Have you changed the viscosity for the fluid? It seems be a viscosity issue and looks strange. If you boat effect is reduced to this simple area, I advice you to make the same with the domain. You don’t need this huge domain if the boat is affecting only to that part of the fluid.

        I’ll make this scene with a reduced area, it can be the half what you have now. I would increase the resolution and not the particle sampling yet, because the mesh aspect is due to the few resolution you have for your particles. Once you are happy with the resolution, you can try to increase the particle sampling but keep in mind you’ll need to make a good value for both parameters. Increasing only the particle sampling won’t be very useful for the simulation because of the voxel size to collide with the objects. Check the mass for the object too, because I think it has few mass or its density is very low. Last thing is to check the “Compressibility” parameter. The fluid is bouncing due to this parameter. Value close to 1, will make the fluid more unstable and values close to 0 more stable but the simulation will take more time.

        I hope it helps😉

        luisM.

      11. Since I’m trying to finish this to a deadline, I’ve tried every method one after another, and I’m literally distributed🙂 among several subjects.That splash problem was also because I was so confused. (I have mistakenly changed the k volume’s inverse parameter and expected to see splashes)
        Though I love to experiment and I have been familiar with most of the functions you mention, I couldn’t move a single step forward.

        I feel that I have a lack of resolution, but as I mentioned before 1.000.000 resolution with 3 as sampling made 1 gig frames. 3 sampling is something I can only dream of.🙂 About mesh radius parameters you pasted above, I’m a lot surprised, because I test each parameter to see how it affects calculation times. A radius like 0.001 is also something I only dreamt😀
        I’ll try those values.
        That domain is at the extents of my camera’s viewpoint, that’s why I made it that size.

        To put it altogether, splashes seem fine, only needs mesh resolution may be. Foam is also ok, (though I haven’t yet seen it rendered)

        1. I have problems with the core fluid. It’s always blobby. Viscosity is at it’s defaults. .0000001 something. Not onlythe mesh but the particles are also like jumping forward, appearing dissappearing, that’s something I couldn’t figure out.
        2. I couldn’t see a calm wave on the core fluid, too. I once had it on an experiment scene. Assinged a displacement and that was all, with defaults. (May be stream was on- but I don’t remember anything about it)
        Iswitched to normal particle mesh from grid mesh and I lost it completely.

        May be there will be no need for such waves but the water is so vicscous, as you say. That’s also something that I don’t understand.???

        If you can open the scene without problems; If it’s not a problem to you and do not require much time can you just set the appropriate parameters on the file so that I can simulate here.

        I feel really bad about spending days on just finding the appropriate parameters for a few settings, while there is a lot more to do after setting. simulate and render.

        I expect a problem with my meshes, too. That’s why I’m desperately in need of such help.

      12. Hi Nezhi,

        I think you have a blobby fluid, it is due to you have very few resolution for that scene. Problem increasing the particle sampling instead of the resolution is you’ll need to relax the fluid so it can be adjusted for the objects before simulating. Another problem increasing the particle sampling is if you have an animated emitter. Anyway, you’ll need to increase the resolution for all domain so you have the enough resolution to avoid the blobby effect.

        Check the scene scale as well. If the fluid seems viscosity, it could be the scene scale. Keep in mind if your scene is scaled up, you’ll need to increase the gravity value. For this reason it is important to create the scenes with a real scale, to avoid to increase or decrease the daemons forces.

        luisM.

      13. Luis I very much appreciate the help. I ve tried almost all threshold values for splash and sometimes used them together, creating a wide range of effects from explosion to spitting:) but not a tidy splash. Previous values (esp. that 0.4 curvature I believe) brought me almost to the end .

        Now meshing is left, but I’m stuck there again.

        I use splash&foam, instead of using them separately- but I couldn’t figure out how splash and foam elements will be seperately exported this way.

        I know you’ve already said here and at the alligator tutorial, to use them separately; If that’s the only choice – I definitely will, but the effect is literally “perfect” now, so I don’t want to restart the splash simulation.

        How will the two components of splash and foam be exported so as to have separate materials?

      14. No worries, I’m here to help.😉

        I think you don’t have splash because the boat is not hitting the fluid to make a splash. I advice you to increase its mass or density. There is a parameter in the “Grid Fluid Interaction” tab when you select the objects, and it is “Interaction factor”. Increasing this parameter, the object will push the fluid stronger.

        You can use splash&foam emitter as well, of course. Once it is simulated, you’ll have two separate binary files to load the splash and the foam. The alligator tutorial is simply a recommendation for that sort of scenes.😉

        luisM.

      15. Sorry for filling the blog with messages. After creating everything from sctratch, splash works fine now but I need some waves on the grid domain surface.

        My grid mesh which used to cause troubles, created beautiful wave patterns without any object interactions. Standard mesh is like concrete. ALthough there is a noise pattern on the surface it doesn’t move like the previous one. Ithought it was due to displacement, but I couldn’t turn on any displacement action on the standard mesh.

        How do we add a calm wave pattern on grid domain with standard mesh? Do we have to re-simulate all?

      16. Hi,

        Usually, before simulating the scene it is normal to relax the fluid. In order that, you’ll need to lock the timeline with the lock button, next simulate the scene and once you have the relaxed fluid, make an initial state for it. Unlock the timeline and simulate the scene. Keep in mind you have a reset to initial state if you made the initial state for your fluid in the initial state tab for the node selected.

        luisM.

      17. May I ask you a single question:

        As you’ve already said before, by splash effects are necessary only around the boat object but this water size is necessary because it’s the borders of camera’s viewpoint. There were lands at the back so I’ve built a mesh similar to the shape of the pool, imported them together with the boat object, and assigned it to the domain emitter to reduce particle amount, increased the resolution to 3 million, it’s getting better but it requires more. There’s serious amount of wasted particles away from the boat.

        I wonder if there is a simple way to increase the resolution inside a box for example. Of course the wave effects should be seamless at the areas of different resolutions.

        I’ve had some questions about density and viscosity above. if you can say a few words about it, too, I’ll be ver glad.

      18. Hi once again,

        Not sure if I’m right but after watching a grid fluid while an inclined box moves from the bottom of the domain to the top, (like an island coming out of water) I ve seen stair like particle groups jumping down the stairs.🙂
        Those are reflected to the grid mesh as sudden come and go effects.

        Increasing the resolution more than 1million doesn’t help me here.
        I either need something to smooth that jumping particles or a way to keep the high resolution areas at the edges of the boat.

        Second one should be my option here because problem is really at the edges of the boat. However there is also a serious viscosity problem.

        1. Is there a way to increase the resolution in a certain area? It could work if I could just increase the res. inside a bounding box linked to the boat.

        2. What should be the actual viscosity and density values. V is 0.000001 for grids but 3 for normal emitters? and density is 1000 kg/m³ I decreased it to 1- it’s nore fluent, but I’m a bit scared. to try it at the highest resolution.

        3. As a last question, you might have noticed that my water is a bit slow. In the documents I’ve seen some frame rate recommendation if water moves slow. Does this mean I have to let’s say decrease fps to 15, and when 25 consecutive realflow frames are taken into 3dsmax water will move a bit faster?

        thank you for the patience🙂

      19. Hi again:)

        It is probably you’ll need more than 1 million of resolution to avoid the blobby effect depending on the scale you have in the scene.

        1- Unfortunatelly, it is not possible. You’ll need a kind of multiresolution but it is not possible to make it for now.

        2- You know that the fluid water density is 1000, so any object with less density will float on the fluid surface. If the object density is greater than fluid density, the object will sink.

        3- If the fluid is slow is because you have a high scale for the scene. Check the measure for your boat keeping mind that each grid square is like a 1 meter. I don’t advice you to modify the frame rate for now to get the speed you want for the fluid. You could increase the gravity force instead.

        luisM.

  3. Only the splash component of the Splash& foam composite particles turn to mesh but, foam stays as it is. When I import foam particle bins to 3dsmax ,I see nothing in the viewport but triangles on render. Render lasts very long.

    In order to convert foam to mesh, I’ve been trying to add separate splash and foam particles for the GFD but no matter what I did, I couldn’t get any particles.
    I have A Grid Fluid, A splash&foam as stated below, they all work fine they’re all cached.

    [+] Grid Fluid Domain
    Boat (custom object)
    Gravity
    Grid_Fluid Emitter
    * Grid_Fluid_Splash_Foam
    Slide (custom object)
    K-Volume

    *[+] Grid_Fluid_Splash_Foam
    Gravity
    K-Volume

    I just replaced 2 splash&foam lines with (*) signs, with the new GFD splash. Nothing happens. Do I have to reset something? I hate to destroy simulation data without intention.

    1. Hey😉

      You have the scene working fine as I can see in the connections. About you are loading the particles in 3dsmax and you can see the particles, it is weird. Check the padding size you have for the files, it means, if you have the first frame as “Splash_00100”, for example, the particles will be visible in the frame 100. How are you loading the particles in 3dsmax to render them? Particle Flow, Real Flow plugin, Renderkit, etc? I paste a screenshots with a Standard mesh for splash and foam emitter. It would be the same for the Splash&Foam emitter.

      Splash&Foam parameters

      luisM.

  4. Luis,

    I’m pretty sure that I couldn’t create a correctly scaled scene.
    ALmost on realflow docs I’ve read the rf grid size as 1 m and my scene was using 1cm as smallest unit. That time, I wasn’t even able to see my objects in real flow until a scaled down to 1/100 of the original. I used scene scale parameters at first and compared 3dsmax’s scaling with simple objects.

    I thought I understood what’s going on, and having a hard time scaling all the objects and animation in 3dsmax I could hardly get my scene in realflow in real sizes. Now real flow scene scale params are at 1. However I though having real world scale would be ok. This pool is around 40 meters, from what I understand that I have to adjust my scene scale to 1 grid=1 meters, my 40 m pool should fit in 40 grids- but I started to believe that’s not the case.
    On almost all tutorials I see every scene -no matter how large- fit in grid sizes of 10 grids large at most. And very provingly, by scaling my scene to %20 I see more reasonable splashes even at lower resolutions.

    I remember creating scenes more simply before, no problems with resolution, they were probably because I never dealt with real sizes but just fitted my scenes insize a few grids.

    Now it’s almost impossible for me to rescale the scene one more time. I can’t stand another core fluid simulation. It seems good enough looking at the videos below.

    Although particles seem very beautiful I couldn’t yet create a reasonable grid mesh. There are 3.6 million particles for the core fluid I can’t find any parameter combination for grid mesh that will make a simple surface. Core fluid always has an annoying low res effect.

    Hav a look at the file at:
    http://www.nezihkanbur.com/grid.wmv

    Mesh resolution is high enough but mesh surfaces blend to from one particle group to the other. displacement effect is strage, highly viscous effects at the back. Some radius is at a large value but for the grid mesh I can’t determine. I’m checking almost any value.

    http://www.nezihkanbur.com/test.jpg

    here are the parameters of the domain. Not to mention that I have tried almost any value from automatic to infinity, This was some value that I found smooth on single frame. Actually at no value I could have it work.
    In fact I can say I couldn’t yet understood which parameters equals to the polygon size in real mesh. Displacement quality seems to increase it but right now it seems my surface is already waving very displacingly?!!🙂

    http://www.nezihkanbur.com/finalparticles.wmv
    here is the final particle effects why can’t I create a simple mesh. I decided to create splashes with thinkbox’s frost, however I need a core fluid. Frost dies calculating 3.5 million particles.

    Any life savings?

    1. Hi Nezih,

      Now, in RealFlow2012, the GridMesher is controlled by the surface tab by selecting the GridFluidDomain. It is due to, in RealFlow5, you had several parameters to control the mesh for splashes, foam and gridmesh. But they were the same parameters for all splashes, foam and gridmesh so if you change any thing for the splashes, the foam will be affected too. For this reason we thought it would be better to control the common parameter is the same tab, and it is the surface tab inside the GridFluidDomain. In your scene, I think you have a very high value for autocell size so the polygons are very big. Try to decrease the auto cell size value so you have a more detailed mesh for your particles. Keep in mind too, once you are meshing with a GridMesh, to change the parameters “Calculate…” in the surface tab to Always, because of “On cache” should be used to calculate the splashes and foam.

      I hope it helps.😉

      luisM.

  5. Hi once again,
    Meshing can use multiprocessors or multicores I believe,but on my pc when I look at task manager during meshing I see CPU is at 12% which means it’s not using the other 7 cores? Is meshing, single threaded? IS there a setting for this? The only setting I know, 8 cores is active, and they work during simulation.

    1. Hi Nezih,

      Only the StandardMesh is singlethread, however the Renderkit and GridMesh are parallelized. There is not a parameter to indicate this. Sorry.

      luisM.

  6. Hmm…
    Never mind, now it’s working.
    All i did was to mesh the grid domain and also, changed a bit the Radius Threshold and now i generate Foam properly.
    Strange…
    Depending on particle count and scene scale, one could say that sometimes it works, other it doesn’t…
    Anyways, cheers…

      1. Thanks for the reply, Luis…
        I’m simulating the foam right now, though half way i run out of disk space; this single project occupied already a 2TB disk and i haven’t even meshed anything yet, except the base ocean mesh.

        Anyways, one problem i have right now is that in three frames out of 600, the foam just explodes. For example, it’s okay on frame 140, it’s exploded on frame 141 and it’s fine again on frame 142. I tried to resim that frame with and without the displacement affecting the foam, tried also different settings for the foam, but no matter what i do, the foam explodes in that frame.

        So, i’m thinking the only way to save those two frames, is to interpolate the particles from frames 140 and 142, to re-create the frame 141. Though i’m not sure how to do that, atmo. Any hints/tips about this maybe?
        Thanks again…
        Gregg

      2. Hi Gregg,

        I saw that behavior before. Could you check if you have any animated object which their axis were rotating for those frames like 180 degrees for the frame 140 and -180 degrees for the frame 141?. If it is so, you’ll need to run a little batch script for these objects to fix that wrong axis behavior. The script is like

        object.getParameterCurve("Rotation.X").eulerFilter()

        “Rotation.X” axis is depending what axis you want to apply the euler filter.

        luisM.

      3. Hello, Luis…
        The only animated object in the scene is the boat, which is animated using the SD file from Max and it’s only a simple forward movement with no fancy rotations. There are no gimbal lock jumps in rotation channels for that matter…

        For fixing those frames, i’m thinking of taking the two frames before and after the broken ones in a separate Particle Loader and interpolate them to create a new middle frame. Or maybe try the Re Timing option instead.
        Would these methods work, or there’s a better/easier/faster way to do this?

        Thanks…
        Gregg

      4. Hi Gregg,

        Do you have a video to see the problem?. If you have particles disappearing for some frames is an odd thing and maybe a bug. Let me know.

        Interpolating particles it would be a solution but I prefer, maybe, the retimer option.

        luisM.

      5. Hey, Luis…
        Not sure what happened but my most recent posts, where placed some posts above the older ones…
        Three times in a row; hope this wont happen with this one as well…
        Just wanted to make sure you saw them and not lost them in the mix…
        May 3rd is their date posted…

      6. I’ll re-post it here, in case you missed it..
        ———-
        Not disappearing; exploding somewhat.

        Here’s a direct link for a video that shows the problem:
        [video src="http://www.sub-pixel.com/tmp/Errors.mp4" /]

        It’s a 17.2MB file; view it with Quicktime, where you can check it frame by frame to see the problematic frames.

        Here are some stats about the scene:
        I’ts a 600 frames simed at 25fps, Grid domain have 24 million particles, there are two splashes with total 2.3 million particles and one foam with 3 million particles. The foam is the one that causes the problems.

        I’ll try the retimer option; haven’t touched it so far…
        Thanks…
        Gregg

      7. Hey,

        I watched the video and you have right. It is very strange because if the particles are exploding, for the next frame the particles should go away and they don’t. Have you connected the foam only with the GridFluidDomain? or is it connected to the splash as well?. The foam must be connected only with the GridFluidDomain, it means, no links with splashes or daemons (except kill daemons). Show me a screenshot of your Exclusive links. Another question, have you resumed the simulation from these frames where the particles are exploding?

        luisM.

      8. Thank you, Luis…

        I find out early in the process that stopping and re-starting the simulation from a frame other than 0, give unexpected results, which according to the latest fixes was indeed a bug. So, all the simulation passes were made seamlessly from start to finish (0 to 600) without interruptions; total simulation time for all passes (grid domain, surface data, displacement, splashes foam, ocean grid mesh) was 125 hours, or 5.2 days non stop…🙂

        So, here’s a screen grab of the foam settings and exclusive links; only “exotic” link to the Foam is a noise field, but i’d be very surprised if that was responsible for the weird exploding frames:

        Anyway, i can afford another sim of the foam without the noise field and see what that does, though i’ll wait from your feedback before i do anything…

        Thanks again…
        Gregg

      9. Hey!

        I’m glad to hear it worked. It is due because the foam is placed over the GridFluidDomain surface, and it is moving with the GFD surface movement as well, but if you put any force to avoid that, it could do crashes in the scene.

        By the way, the simulation is amazing!

        luisM.

  7. Hi Gregg,

    Have you connected the Foam to the GridFluidDomain? Remember that the splash and foam mustn’t be linked between them. Another thing is, take a look if you have a force daemon attached to the foam as well and unlinked it.

    Do you have the foam with “bounded” parameter to “YES”?

    luisM.

  8. Hello, Luis…
    The only animated object in the scene is the boat, which is animated using the SD file from Max and it’s only a simple forward movement with no funcy rotations. There are no gimbal lock jumps in rotation channels for that matter…

    For fixing those frames, i’m thinking of taking the two frames before and after the broken ones in a seperate Particle Loader and interpolate them to create a new middle frame. Or maybe try the ReTiming option instead.
    Would these methods work, or there’s a better/easier/faster way to do this?

    Thanks…
    Gregg

  9. Not disappearing; exploding a bit. I’ll make up a short video and post a link, asap. I’ll check the retimer option then….

  10. Not disappearing; exploding somewhat.

    Here’s a direct link for a video that shows the problem:
    [video src="http://www.sub-pixel.com/tmp/Errors.mp4" /]

    It’s a 17.2MB file; view it with Quicktime, where you can check it frame by frame to see the problematic frames.

    Here are some stats about the scene:
    I’ts a 600 frames simed at 25fps, Grid domain have 24 million particles, there are two splashes with total 2.3 million particles and one foam with 3 million particles. The foam is the one that causes the problems.

    I’ll try the retimer option; haven’t touched it so far…
    Thanks…
    Gregg

  11. Hey, Luis…
    Not sure what happened, but my most recent posts here went a couple of posts above older ones…
    Strange, but just wanted to make sure you saw them and not lost them in the mix….

    1. Well…
      That ship project is finished…
      You can watch it here:

      Thanks for all the help!
      Cheers,
      Gregg

  12. DOes anyone know why I can’t see the interaction panel for my SD? I’ve used SD before and don’t remember not seeing it. I have it selected, all particles for hybrido are set up, and I don’t get one shot at adjusting interaction or collision distance anywhere. Help?

  13. first of all thanks for this, it has been very helpful!
    If you have time for a question or 2?
    1)when you say turn off the mesh, you mean turn build to no, correct?
    2)when you say to simulate a couple frames, and build the mesh that is just to see and change the settings, but is not nescassary, if you aree happy with the setup?
    4)what happens when something is in cache, my expectation was that the simulation data for that item would not need to be resimulated, but that doesnt appear to be the case?
    5)process wise it makes me nervous to continue to overwrite my simulation files, are we just refining the same data so their is no reason to have the old ones, or are they hidden somewhere, in the data files.
    6)is there any clear literacture i can read that explains the workflow with hybrido, caching, active, from particle/from domain, surface vs displacement. It doesnt seem intuitiv to me and can be frustrating.

    Thanks

    1. Hello botlove!

      1) Yes, you have right, I mean turn build to NO.

      2) You have right again🙂. Simulating a couple of frames is just for seeing whether the mesh setup is working.

      4) Mnnnn. Once you put any node in cache mode, if you simulate again there won’t be new simulated files for it. Maybe it is taking long time to load the cache files for each frame.

      5) Do you mean when the objects are in cache mode?. The files shouldn’t be overwrote if they are in cache mode. Take a look to the hour for files when resimulating.

      6) Take a look at this video https://luismma.wordpress.com/2011/11/04/tutorial-how-to-work-with-realflow-nodes/. It is a basic workflow to work with hybrido.

      luisM.

  14. Que tal Luis,
    Estoy tratando de usa displacement sobre hybrido, pero el displacement no se mueve, parece estatico, no tengo idea que atributo modificar para que se mueva mas que wind speed, y he puesto varios valores y ni asi. Saludos.

    1. He intentando en varias escenas, unas grandes y otras no tanto, pero no puedo hacer que el displacement se mueve, parece estatico…Saludos

      1. Hola Jesus,

        Echa un vistazo a ver si tienes en “Evaluation uvw mapping” a “Top projection (Average Velocity)” dentro de los parametros del Displacement. Tiene que estar en esa opcion para que el displacemente se mueva.

        Take a look at the “Evaluation uvw mapping” parameter. It must be put to “Top projection (Average Velocity)” inside the Displacement parameters. You must use that option so the displacement can move.

        luisM.😉

    2. Hola Jesus,

      Echa un vistazo a ver si tienes en “Evaluation uvw mapping” a “Top projection (Average Velocity)” dentro de los parametros del Displacement. Tiene que estar en esa opcion para que el displacemente se mueva.

      Take a look at the “Evaluation uvw mapping” parameter. It must be put to “Top projection (Average Velocity)” inside the Displacement parameters. You must use that option so the displacement can move.

      luisM.😉

  15. thanks for the reply, some of the questions were abit ovararching in their reach, so I apologize. I finally realized the realflow rfrk renderer in rf will meh splash and foa nicely, which was a major hurdle obviously:)
    Do you have some tips on emission setups, for example if I want to limit my particle emission to object contact what settings should i look at?

    1. Hi Andrew,

      No worries🙂. I’m glad you finished your work with rfrk and realflow. Do you have a link to see it?.

      About the emission setup, depending if you are talking about objects colliding with a GFD fluid, increasing the resolution for the GFd will give you a better contact with the objects but the simulation will be slower. It is good bear in mind this depending on the camera view.
      If the object is colliding with sph particles, I advice you to check the “Collision Distance”. I prefer to put a value of 0.01 here. If this value is high the simulation will be slower and the fluid could be unstable as well. If you put a value of 0.01, the “Distance tolerance” could be 0 so you don’t want some particles crossing the mesh.

      luisM.😉

  16. Hello Luis,

    Thanks for your answer about the displacement not moving. I have one more question, this is about how to animate a boat up and down, imaging a storming day, please look this video as reference http://www.psyop.tv/twinings-sea/

    Because so far the boat that I using for a personal project just moving in one axis?. I’m using Hybrido for this animnation.

    Thanks in advance.

    1. Hey jesus,

      do you have a xml scene to check?.

      About how to animate a boat up and down, you can use a box, which will be the center of mass for the boat. It must be below at the boat. Once you have the box and the boat, you’ll need to create a Multijoint, where the object A will be the boat, and the object B nothing, it means empty. Then, you choose the “Creation mode” as ” At locators bbox centers” and select the box you created as “Locator”. The force will be Unlimited so the multijoint can’t break. Depending on the position of the box as center of mass for the boat, the boat will more stable or not.

      I hope it helps.

      luisM.

  17. Hi there, I am having a problem with hybrido displacement maps in 3ds Max. I am trying to import my whitecap bitmap sequence that I wrote out in RF. In the tutorial they use Maya, placed texture2d /3d and then a RemapValue node to grab the alpha and pipe it in to the IMAGE input of the projection node. then it is scaled and rotated until it matched up with the scene. This works great, and it looks better that using just particle alone I think, but I cannot use Maya at work. I have to stay with Max. I have searched and searched but cannot find a similar workflow in 3ds Max, but can’t find anything. I am sure it is something simple I am missing but I am new to materials in Max. If there was a solution in Vray that would be even better. Thanks for any advice.

      1. Hey Thanks a lot! I will definitely try that out. On another note (I hope you don’t mind) . I am working on several RF production shots for a commercial, and another FX house generated the ocean meshes which I inherited as point caches within Max. I have the meshes loaded up in RF along with the collision geometry (a boat) . The mesh is animating exactly as it does in max, but I need to turn it into a real wave with the same motion (as the original mesh), but at the same time have it generate splash and foam from hitting the edges of the boat. I was successful in using the original mesh as a custom realwave, and it took the shape of the original, but it does not move and crashes if I try to simulate.

        Is this even possible? or should I be asking to get the original files?

        Thanks in advance

      2. Hey,

        I’m going to put a new post to know how to do that, so you can have any animated mesh you want created in 3dsmax, or any 3d platform and import it to RealFlow and use it as a realwave mesh.

        luisM.

  18. Hello,
    Somewhere along the way, I must have made an error since my splash particles are born say 10cm above the surface, this is clearly visible since the camera angle is quite low.

    I think it has something to do with my displacement.
    What I did was simulate the GFD, put it to cache, wrote out the displacement set that to cache, calculated the displacement fast and precise, put those to cache.

    Is this a do-over, or do you have any advice where to attack?

    Thanks,
    D

    1. Hi David,

      Yes, you have right. It seems you have a problem with the displacement but, one question. Keep in mind when applying Use Displacement for Splashes, you will see the result when you make the mesh, so, does it means that your particles are 10 cm above the mesh?

      luisM.😉

  19. Is there a reason my splash particles will not simulate after I simulate the GridMesh. If I simulate the GFD then the splash, then its works. But everytime I simulate the GridMesh then the splash, its stops working. It will run through the scene as if it is simulating, it even writes out the splash particle files, but no particles are displayed and when I load the particle bin files into the third party software, nothing shows.
    Here are screen shots of my settings after I simulate the gridmesh and try to simulate the splash particles:
    GFD

    GridMesh

    Splash Emitter

    Thanks in Advance.
    -Jason

    1. Hey,

      It seems to be a problem with the parameters. I’m seeing you have a very hard condition to simulate the grid mesh. Maybe you can take a look at these videos to understand how to work with gridmesh and splashes.

      https://luismma.wordpress.com/2011/10/13/tips-and-tricks-when-we-use-from-field-and-from-particles/

      https://luismma.wordpress.com/2010/08/05/tutorial-how-to-create-splash-foam-or-gridmesh-on-the-basis-of-a-gridfluiddomain/

      Basically, when generating the main fluid, the “Calculate…” parameters in the Surface tab must be put to “Always”. In this way, you can use the depth of field to simulate the splashes faster. The same thing for the GridMesh. These parameters must be to “Always” as well. I advice you to leave the detail threshold to 0.2 and from particles intstead of from Field cause, the last one, has a less approximation to the particles than the from Particles which is more accuracy. Once you have the grid mesh, the “Calculate….” parameters for the main fluid in the Surface tab should be put to “On cache”. Now, the splash simulation will be faster.Don’t forget to put the main fluid to cache and deactivate the simulation for the mesh.

      luisM.

  20. Hello Luis,

    I try to follow step by step this great tutorial. I’m actually simulating my huge 1935 Normandie ship crossing my domain during 1200 frames. All the parameters are on Always and Surface “From Field”. The simulation looks very cool for the moment.

    Now, I would like to know exactly for the next step, what parameters in the grid fluid domain I have to put in cache or not for calculating my Grid Mesh ?

    Node ?
    Calculate Surface ?
    Calculate extended velocity ?
    Calculate displacement fast ?
    Calcultate displacement precise ?
    Displacement ?

    Coz you say to put “From particles” instead of “From Field”… does it means the domain needs to calculate everything again ?

    Sorry for asking you this kind of details… but I dont want to make an other mistake on this simulation, spent too much time on it. And by the way, thank you very much for your advice concerning the Scene Scale in Realflow, Finally, I scaled all my scene in 3DS Max instead in Realflow, and now the simulation looks far away better and fast.

    Thank you so much Luis !
    I will show you the result at the end🙂

    1. Hey,

      In order to calculate the grid mesh, you should put the

      Calculate Surface = Always
      Calculate extended velocity = Always
      Calculate displacement fast = On cache
      Calcultate displacement precise = On cache
      Displacement = Never

      And selecting the “From particles” instead “from field” due to the “from particles” will adjust better for your mesh.

      No problem🙂 You know you can ask all times you want.

      luisM.

  21. Thank u luism your tutorials and tricks helped me since i am with Realflow. But i hv a question a bit out off this tutorial. Once i have simulated everything in hybrido (grid mesh, splash, wet, foam, mist) how can i use the grid mesh inside my 3d app (maya/max; i use maya) to make a large ocean where the ship moving. Hybrido simulation happens in a small spot (like a shooting spot). Thank u in advance.

      1. Thanx luism. but my question was not like that. the mesh i get from the hybrido mesh is a short piece of ocean. how to integrate it with a big ocean surface?

  22. Hi Luis,

    I was able to mesh my hybrido RPC using “RFRK3 HybridoMesh”, I am able to tweak it to my preference. I find this as time saver than meshing a high resolution build overnight in realflow. But one thing that troubles me is using my displacement to add further details to my mesh, nothing is showing up though I already inserted the tiff files and checked displacement box. I looked at the tiff files and it has all the info needed to displace. How do you really use “RFRK3 HybridoMesh” with displacement?

    Thanks in advance.🙂

    1. Hi sandra,

      What 3d platfor are you using? If it is 3dsmax, there is a bug so it is not working. I’m going to publish a tutorial to apply the displacement directly on the mesh from RealFlow.

      luisM.

      1. Wouldn’t that create extremely dense objects per frame, that would be hard to deal with in 3dsMax?

      2. Hi Greg,

        it was for RealFlow 2012, and now, for RealFlow 2013 the workflow has changed. I don’t undertands what you mean creating extremly dense objects per frame. Do you mean polygons?😉

        luisM.

  23. Hi luis, hugely informative blog u have here🙂
    could u please also put up some info about a hybrido scene with displacement in Rf2013..there is this video “08 _ RealFlow 2013 _ HyFLIP, Ocean Statistical Spectrum” which is great but i reckon there a couple of ways we can go about adding/simming displacement to the mesh..i still have workflow doubts about using displacement on meshes and generating maps to use in a platform..
    Thanks🙂

    1. Hi,

      Thanks for your words. I’m making some tutorial to explain the way to get displacement in meshes and rendering those meshes faster. It is only a workflow but it could be helpful for other people.😉

      luisM.

  24. Hello luis! Great tuto!

    I have a question please =)

    In this simulation of boat, Bhavya said he used plugin/technic “Snap Foam Particles To Mesh”

    With your tuto, it’s not usefull right?

    Thank you!

    1. Hey Francois,

      It is a scene which you can find in the demo scenes inluded with RealFlow. Demo scenes>Graphs>snap_foam_to_displaced_mesh

      luisM.

  25. Hello Sir Luis

    Ive just started learning Realflow 2012, (i dont have the latest)

    Anyways, I got the Grid Fluid Domain working – see linked here https://vimeo.com/167562657

    Now, my issue is when I tried simulating Splash and Foam, it crashes – see linked here https://vimeo.com/167562658

    I already set the GFD to cache, i also did the Exclusive Link.

    It would be a awesome if you can help me with this.

    Thank you in advance,
    Owen

    1. Hi Owen,

      Sorry for delayed answer. I advice you to use RealFlow 2015. The demo is fully functional but you can not use it for profesional works. Realflow 2012 is old and the solvers have changed. With RealFlow 2015 you will have better simulacions and it is probably you don’t have those kind of bugs😉

      luisM.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s