WIP gradient based soft physics
Added 2023-01-22 13:39:28 +0000 UTCJust a quick update on what I'm doing and planning in the short term.
Currently, the area that soft physics affects in is divided into the main, outer, areola and nipple groups in TittyMagic 5. In BootyMagic, there's just one soft physics area for the entire butt.
For the past few days I've been looking into replacing this with a system where each individual soft joint can have a different physics parameter value. I have a working proof of concept, but there's still a lot of stuff to implement, ideas to experiment with, and a lot of existing code to restructure before I can post footage to look at or a playable new version.
With individual physics values for each soft physics joint, the way that breast fat behaves can be fine tuned even more. For instance, instead of the entire "main" area having the same Fat Spring, it's possible for Fat Spring to smoothly transition from a high value in the furthest outer joints to a lower value in the middle of the breast, and a higher value again towards the areola and nipple. Or at least as smoothly as possible given the very limited number of soft joints in VAM. In other words, Fat Spring can be defined along a math curve - or a gradient. The same will work for any other soft physics parameter except for the Back Force ones (due to VAM's internal limitations).
In addition, the system allows for individual soft physics joints to be updated dynamically - this would be the local skin deformation physics that I've talked about before. I can't wait to get to try that out and see if there's actual potential there!
Doing this now allows me to implement the same systems in BootyMagic from the get go, this saves some work since I don't have to calibrate the current physics system for different butt sizes and softness values, I can just do that with the gradient based system.
Apart from this, there's a bug in TittyMagic where loading an appearance preset can cause calibration to fail silently and make the breasts non-colliding. I will try to get this fixed soon, but in the meantime, if it happens you can go to Control & Physics 1, toggle person collision off and on, and then reload the plugin.
Cheers
-everlaster
Comments
The vast majority of the performance impact of soft physics comes from two things: that PhysX which VAM uses internally for physics collision calculations runs on a single CPU thread, and that calculating collisions between the soft joint colliders is itself a pretty costly operation. Those are still true with the upcoming changes, so I'm not expecting there to be much of a change in performance one way or the other :)
everlaster
2023-01-23 11:43:14 +0000 UTCVery interesting. I'm not qualified to properly understand all the 'innerworkings' of the majority of all this. But at the moment, considering how generally-speaking taxing VAM is (and _especially_ how the default Glute soft physics drops my system to its knees); I do wonder - how any of this will affect performance, if at all? For the better? Or will it be a neutral change in that regard?
BlackstarGuard
2023-01-23 03:45:22 +0000 UTCScience!
Ganzt
2023-01-22 14:17:09 +0000 UTC
