Testing Breast Physics (4k Video)

[Karna5]

Today I did some tests to see if I could make the ESP control the bone data scaling with breast physics. I tried setting the BSMS values large, and then I tried setting them to zero, but there was no difference that I could see in behavior.

 

To properly compare I set up three mannequins side by side and dressed them with the following outfits, all using the same bodyslide preset.

 

Left: Outfit which tries to override the bone weight scaling

Middle:  Outfit which does not try to change anything with bone weights

Right:  Control, no outfit at all, thus using the regular nude female body.

 

I also used bullet time to slow time during much of the test so I could see the movements more easily.

 

Please note the video is not intended to be salacious, entertaining or even peaceful. It's just a video of my test

 

[Kaz Aanh]

Reported & Flagged for sexual content.

[Mern]

Awesome stuff man!

[bluedog]

Thanks for the comparison.  These physics are similar to what I am getting and I am hoping that we can make it better.

 

As I said in the other post, for me the issue is dampening.  The breast behave like cloth, not like human flesh at .9 g.  The question I posed was whether we could create new physics bones on the skeleton, replacing some of the existing scaling bones so that we stay within the 4 bones / vertex limitation.

 

Of course I am very new at this.  My approach (albeit a thought exercise at this point) is to create a jiggle deformer and substitute it in for a scaling bone.  I suppose that editing the hkx in the jiggle deformer may be just as difficult as editing the hkx on the cloth bone.  I am very unclear about the topology about the topology of the FO4 joints rig, but perhaps we all are a little in the dark about it.

[formanfred]

Very well done I love it

[Mistle]

Noticing the breast crush on the right model, the middle doesn't look to bad and the left seemed to be the better of the three least from what I have seen in this test video.  Awesome work on this .