Announcement

Collapse
No announcement yet.

Under ocean close to surface issue

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    in the most cases the rendering require ram less than the one used in the simulation, because of this we didn't provide ram predictor.
    however, there are few cases when you can overload the memory
    - the geometry volumetric mode produces so-called "fragments", this is renderer's internal information that is used in the deep compositing and in the antialising, each transparent layer generates one fragment and because phoenix generates multiple transparent layers you can easily reach many GB RAM usage even with low res simulations.
    - foam bubble/splash/cellular mode with multiple overlapped bubbles. by default the option "optimize bad systems" should prevent this situation
    - too big skyline subdivisions of the ocean.
    - convert to fog option of the foam shader used with too small cell size.
    ______________________________________________
    VRScans developer

    Comment


    • #32
      Thanks Ivaylo, I'll just have to keep an eye on my memory management.

      I'm now creating a sim of a dishwasher. But the only way I can effect the preassure that the water is ejected with is "discharge"
      I was assuming that "velocity" in the source was supposed to do that?
      I'm using this nightly build "phoenixFD_adv_21010_max2014_vray_24_x64_24113 " with Max 2014 #64
      Can you try it when you have the time and also I mailed you (support) a question regarding the texture option for temperature and discharge.
      Probably just me

      Comment


      • #33
        the velocity value of the source have another function, it is related to the velocity of the object that produces the liquid, actually it is multiplier, not velocity directly.
        the discharge parameter controls how many liquid is produced, but the liquid amount is directly related to the liquid velocity. it's like in the home water tap, you can't increase the water amount without to increase the velocity.
        about the texture - it is used as a multiplier to the corresponding value. the white means 1 the black means 0
        ______________________________________________
        VRScans developer

        Comment


        • #34
          Ok so I'll just go on as before with a higher discharge value.
          The answer I got from support was that I have to use Planar from Object XYZ to make the checker work. Well it does but what if I have a mapped object (often the case) and for me that is always Explicit mapping channel. What then? It does not wor at all for me unless I change to Planar from Object XYZ. But that is obviously not correctly mapped.

          My brain is to congested!

          Comment


          • #35
            well, the mapping should work with any mapping. if the checker refuses to work with explicit mapping this may be a bug. i will talk with the support to clear the case.
            ______________________________________________
            VRScans developer

            Comment


            • #36
              I'm still running some test for the washing machine.
              There is a surpricingly big difference in the sim solution when having a static grid versus having it linked and rotating with the emitter.
              Does the turning of the arm that emitts fluid create so much turbulence with a static grid that is totally destroys the ejection "ray". It all becomes a cloud of liquid.
              I guess that when the grid is linked to the turning arm emitting the liquid the sim does not register the arm as something that disturbs the volume.

              But what if I have two arms running at different speeds/directions. Guess I have two options. 1. exlude from sim. 2. have to simulations.

              Comment


              • #37
                can you post two images for both the cases?
                ______________________________________________
                VRScans developer

                Comment


                • #38
                  Sure, I'll post you the scenes too. I'll send them to support.

                  Comment


                  • #39
                    Hello! Yes, the explicit mapping seems to be broken, we received a report simultaneously from another user as well. I'll post here when we have progress on this
                    Svetlin Nikolov, Ex Phoenix team lead

                    Comment


                    • #40
                      good, those scenes I sent you. if you have the time. Reduce the emitter radius to half and test the sim. I got a very strange result with a lower radius. No emission or only emmision during the first say 10 frames.

                      Comment


                      • #41
                        The issue is fixed, you will find the fix in tomorrow's nightly build
                        Svetlin Nikolov, Ex Phoenix team lead

                        Comment


                        • #42
                          You are the best. Do you want us to use the nightly builds? If so is it here in the forum that you want me to report issues or by support mail? Support kindly hinted that they (you) don't offer support the nightly builds (fully understandable).
                          What ever works best for you.

                          -Lander

                          Comment


                          • #43
                            Oh, well since the nightly builds are going to become the next official release one day, it's good to have them used so that we can catch and clear the bugs
                            Svetlin Nikolov, Ex Phoenix team lead

                            Comment

                            Working...
                            X