wip Real-Time Body Physics (DirectX 11)

[PixieDragon]

22 hours ago, PixieDragon said:

Physics reverted back to not working again, but this time I figured out how I can get it to work, I'm just not enthused about the solution. If I turn off Memory Integrity in Windows Security > Device Security > Core Isolation then the mod works properly.

False alarm, it doesn't work properly with Memory Integrity turned off.

[joezreezal]

does this mod have belly jiggle physics yet? im ready to support

[CPVC]

On 9/26/2026 at 5:56 AM, joezreezal said:

does this mod have belly jiggle physics yet? im ready to support

 

Not yet, but it is planned.

[Pankmasdred]

hey cpvc would you also add hair physics? i think that would have a lot of popularity and make your mod be everywhere

[jeno3]

Hi cpvc,

does the official version contain anything new, or is it the same one you provided here for testing? My latest version is from August 16th.

 

Did you take another look at the breast physics—specifically what I sent you regarding offsets, or is that no longer something you're working on? I’d really appreciate it if you could somehow get that sorted out.

[CPVC]

23 hours ago, Pankmasdred said:

hey cpvc would you also add hair physics? i think that would have a lot of popularity and make your mod be everywhere

 

Maybe someday. There are some limitations I need to overcome first and special animatable hair meshes would need to be created for it to deform well.

 

14 hours ago, jeno3 said:

Hi cpvc,

does the official version contain anything new, or is it the same one you provided here for testing? My latest version is from August 16th.

 

Did you take another look at the breast physics—specifically what I sent you regarding offsets, or is that no longer something you're working on? I’d really appreciate it if you could somehow get that sorted out.

 

The latest experimental version has better breast physics. I'm using a more robust method of extracting the euler angles from rotation matrices and the results look better. I haven't done anything specifically for your issue yet, but I have some ideas in mind for how the physics might be tweaked to accommodate meshes like yours. I'll do some experimenting, but I think that maybe introducing a configurable radius variable to the angular physics simulation could be the answer.

[jeno3]

11 hours ago, CPVC said:

The latest experimental version has better breast physics. I'm using a more robust method of extracting the euler angles from rotation matrices and the results look better. I haven't done anything specifically for your issue yet, but I have some ideas in mind for how the physics might be tweaked to accommodate meshes like yours. I'll do some experimenting, but I think that maybe introducing a configurable radius variable to the angular physics simulation could be the answer.

I’ve tested the new version. Depending on the body posture, the breasts rotate sideways or inward toward the body, and sometimes they get pulled upward. This relates to the angular physics. It happens with standard breasts, too; my CC breasts just exaggerate the effect, but the general alignment is the real issue here. I tested it with various sizes, and the result remains the same. I’m not sure if specific sliders play a role, but I ran a test ranging from standard values down to negative ones, and the behavior persists. Everything is fine during normal, upright walking, but as soon as the Sims perform other poses or animations, tilting their upper bodies or shifting their posture, the breasts distort due to the angular physics. A variable radius sounds great in theory, but I suspect you’d need to determine specific offsets for different upper-body postures and Sim orientations. It would be easier if the actual physics behavior could be transferred to Blender; then you could experiment with the bones and pinpoint the exact coordinates for the offsets. As it stands, it’s all just guesswork that requires constant manual input and testing.

[brewers1]

Everything works fine and then I crash, this was in my log file......any ideas?

 

[RTBP.ERRO 00:43:20.84 18932] Uncaught exception: 
    EXCEPTION_RECORD: EXCEPTION_ACCESS_VIOLATION (0xc0000005) at 00000000052D7BC0
    #1 ???
    #1 ???: 0x7ff8418fe490 AmdDxGsaFreeCompiledShader
    #2 ???: 0x7ff8418fe490 AmdDxGsaFreeCompiledShader
    #3 ???: 0x7ff8422c41d0 AmdLiquidVrD3D11WrapDeviceContext
    #4 ???: 0x7ff84eafcd70 BaseThreadInitThunk
    #5 ???: 0x7ff84faacac0 RtlUserThreadStart
 

[CPVC]

20 hours ago, jeno3 said:

I’ve tested the new version. Depending on the body posture, the breasts rotate sideways or inward toward the body, and sometimes they get pulled upward. This relates to the angular physics. It happens with standard breasts, too; my CC breasts just exaggerate the effect, but the general alignment is the real issue here. I tested it with various sizes, and the result remains the same. I’m not sure if specific sliders play a role, but I ran a test ranging from standard values down to negative ones, and the behavior persists. Everything is fine during normal, upright walking, but as soon as the Sims perform other poses or animations, tilting their upper bodies or shifting their posture, the breasts distort due to the angular physics. A variable radius sounds great in theory, but I suspect you’d need to determine specific offsets for different upper-body postures and Sim orientations. It would be easier if the actual physics behavior could be transferred to Blender; then you could experiment with the bones and pinpoint the exact coordinates for the offsets. As it stands, it’s all just guesswork that requires constant manual input and testing.

 

Yeah, the new experimental release has improved things somewhat over what we had before, but the problem does still persist. I was hoping to get this fixed completely before the new release, but I wanted feedback on the new named pipe IPC method and other improvements, so I released.

 

I was hoping I could continue using Euler angles for simplicity sake, but I might need to switch to quaternions if I can't find a method to combine Euler angles without Gimble lock, like what I believe you are describing. I'll get it taken care of, though.