[Fisio SB addOn released] Fisio: Physics Simulator

1111213141517»

Comments

  • AlbertoAlberto Posts: 1,382

    stem_athome said:

    Alberto said:

    The precision in Fisio RB has embedded the substeps; precision 2 uses the same method as precision 1 but with ten times substeps.

    Hi,

     

    Thanks for the info.

    How many substeps in precision 1?

    100 substeps per frame (thus, higher FPS increases the total substeps of the simulation).

    The  3 and 4 precision levels use other methods, each at 200 substeps per frame.

    All the levels of Precision automatically increase the substeps if there are links in the scene. Precision 1, by 16 times, precision 2 and 4 by 10 times, and precision 3, two times.

     

  • stem_athomestem_athome Posts: 516

    Hi,

    Thanks for the reply.

    So in a sim with links at precision 1, there will be 100 X 16 (1600) substeps. That is a lot.

    I am attaching a scene file to show issue I am having. It is a simple rig. A main body with 8 cubes (4 blue, 4 red) and a ground see pic.

    I ran a number of sims with different fixed links.

    1. Body with 4 fixed links, one to each blue cube. P1 (precision 1) and P2(Precision 2) both failed (blue cubes moved)

    2 each blue cube fixed link to body. P1 OK

    3 I added fixed links from red cubes to blue cubes (as in scene file) P1 red cubes translated along X, failed. P2 red cubes translated along X just more slowly.

    4 I changed the links of red to blue and made them blue to red, still failed.

     

    The only way I found to get the fixed links to hold was to reduced the Density of the main body down to 10 (from default 1000), which IMHO should not be needed.

    When running a sim on my main test rig, there are wheels connected to the main body via 4 links (suspension system). Reducing the main body density is not a solution, as the density of the other objects in the links are causing issue to the other links. I have attemted to balance out the density of the objects to get the rig stable. But not been able to.

    links.jpg
    500 x 400 - 18K
    duf
    duf
    link_test.duf
    74K
  • AlbertoAlberto Posts: 1,382

    stem_athome said:

    Hi,

    Thanks for the reply.

    So in a sim with links at precision 1, there will be 100 X 16 (1600) substeps. That is a lot.

    I am attaching a scene file to show issue I am having. It is a simple rig. A main body with 8 cubes (4 blue, 4 red) and a ground see pic.

    I ran a number of sims with different fixed links.

    1. Body with 4 fixed links, one to each blue cube. P1 (precision 1) and P2(Precision 2) both failed (blue cubes moved)

    2 each blue cube fixed link to body. P1 OK

    3 I added fixed links from red cubes to blue cubes (as in scene file) P1 red cubes translated along X, failed. P2 red cubes translated along X just more slowly.

    4 I changed the links of red to blue and made them blue to red, still failed.

     

    The only way I found to get the fixed links to hold was to reduced the Density of the main body down to 10 (from default 1000), which IMHO should not be needed.

    When running a sim on my main test rig, there are wheels connected to the main body via 4 links (suspension system). Reducing the main body density is not a solution, as the density of the other objects in the links are causing issue to the other links. I have attemted to balance out the density of the objects to get the rig stable. But not been able to.

    Use the Precision 4, it's the more robust when using links. I tried your scene with P4 and it worked.

  • stem_athomestem_athome Posts: 516

    Alberto said:

    Use the Precision 4, it's the more robust when using links. I tried your scene with P4 and it worked.

    Hi,

    I posted the simple rig to highlight the problem with materials. But you do not see that as an issue.

    On the more complex rigs I am building, if simply changing to precision 4 was a solution to the instability in the rigs I see, then I would use it, but it is not.

    As no one else has reported these problems, which is probably due to no one building complex types of RB rigs, I will stop taking up our time on this.

    Regards,

  • CyperImerCyperImer Posts: 1
    edited February 27

    Alberto said:

    hookflash said:

    Alberto said:

    hookflash said:

    Alberto said:

    hookflash said:

    Alberto said:

    jhandy said:

    can't load fisio engine functions

    It seems it's a problem similar to this: https://www.daz3d.com/forums/discussion/comment/7874911/#Comment_7874911 

    Check these comments:

    https://www.daz3d.com/forums/discussion/comment/7878706/#Comment_7878706

    https://www.daz3d.com/forums/discussion/comment/7879396/#Comment_7879396

    Let me know if the comments help you, please.

    I'm having this same issue. CFF Explorer reports the same missing DLLs, so I tried installing the VC++ Redistributable you linked to. The installation failed, as I already have a more recent version installed (which I reinstalled, but that didn't solve the problem). Any advice?

    Edit: One thing that might be worth noting is that, when I click "Run Simulation" a second time (after it fails the first time), Daz Studio crashes.

    If the redistributable installation fails, it usually is because it's already installed.

    Could you post a screenshot of the contents of the "/DazStudio4/libs/Fisio" folder? And a post of the report of CFF Explorer, please.

     

     

    Screenshot-2024-01-23-091100 Screenshot-2024-01-23-091243

    Also, not sure if this is relevant, but the non-softbody version of Fisio seems to run fine on my system. Only FisioSB is giving this error.

    Is Filament working in your system? Check, please, "dzfilamentapi.dll" with CFF  (it's directly inside "/DazStudio4/").

    If with Filament you see the same missing, you can update your system with any (or both) of these Windows updates;

    https://www.microsoft.com/en-us/download/details.aspx?id=53587

    https://support.microsoft.com/en-us/topic/update-for-universal-c-runtime-in-windows-c0514201-7fe6-95a3-b0a5-287930f3560c

    CFF Explorer seems to be showing the same files as missing for dzfilamentapi.dll, but, oddly, Filament runs fine. I tried updating with the files you linked to, but installation fails with the error, "Another version of this product is already installed." Very strange...

    Yes, very strange: CFF and FisioSB both miss the files, but Filament doesn't. Let me see what I can find about it.

    Have you found a solution to this? I am also having the same problem as described above and have went through the different steps as well.

    Also,  I've manually installed the api-ms-win-crt dlls such that there's no missing dependencies but the simulations are still failing in the same way as before unfortunately.

    If I try to "Run Simulation" again after the error message, then daz studio crashes. A crash report snippet is attached.

    2024-02-27 114204.png
    1271 x 759 - 155K
    2024-02-27 114352.png
    521 x 499 - 38K
    Post edited by CyperImer on
  • AlbertoAlberto Posts: 1,382

    CyperImer said:

    Have you found a solution to this? I am also having the same problem as described above and have went through the different steps as well.

    Also,  I've manually installed the api-ms-win-crt dlls such that there's no missing dependencies but the simulations are still failing in the same way as before unfortunately.

    If I try to "Run Simulation" again after the error message, then daz studio crashes. A crash report snippet is attached.

    Not yet, but your CFF report could help. 

    The "api-ms-win-crt-heap-l1-1-0.dll" in your computer is for a 32-bit machine (it says Portable Executable 32), but the requirement is a Portable Executable 64. So this file is not compatible. Could you install the 64-bit version?

    See the following image from my computer:

    CFFapi.png
    1021 x 699 - 132K
Sign In or Register to comment.