WIP: Part 2 of Naturalis glute soft physics driven morphing

Video credit: HPVL by PluginIdea (CC BY-ND)

I tested what the morphing looks like if instead of calculating angles relative to the glute joint, I ignore the glute joint entirely and use the average depth of soft physics joints to drive the morphs.

Soft joint depths illustrated below. (For definitions of technical terms see the end of the post.)

It works pretty well and looks volumetrically quite realistic. Compared to the previous method, there's some advantages:

A major advantage is that the rotation of the glute joint itself has no effect on the morphs. The effect is much more isolated to just collision since that's the cause of major indentations in the mesh. As a result there's much less intereference from this system to how the g-force morphs are animated by Shapeshift.

Also, it's easier to apply the morphs at high values without introducing issues due to things like the angle of the leg distorting the glute. In the previous demo, the amount of morphing was deliberately reduced to prevent excessive deformation when the leg was rotated. That's much less of an issue now. In addition, the mass of hard colliders can be lowered, thereby reducing the impact of VaM's built in bone weights on the mesh when colliding with the glute. I don't want to just entirely remove that effect though, the bone weights will still be necessarily applied in animations as the glute joint rotates, so I think it's appropriate some of that effect is present in collisions as well.

There are however some differences in this demo compared to the previous one that makes it not quite an "apples to apples" comparison:

• Mesh pressure is enabled in this demo. The morphs are driving the change in shape, while the mesh pressure is compensating for and resisting volume loss, and both systems depend on the exact same depth values of soft joints. They complement each other nicely.

• I'm not having to heavily dampen the morphs in order to prevent jittering (tiny vibrations, if you will) that made it a challenge in getting the previous version to look smooth. As a result, the morphs now apply more immediately and responsively. The jittering occurred in the previous version due to how it anchored things around the current amount of g-force (which can change rapidly). However I will still need to introduce g-forces as a factor in order to allow different sets of morphs to target different g-forces, so I might need to dampen the calculations anyway, but I'm hoping it will be a lesser issue this time around and not result in lesser quality than what you see in this demo.

Next, I'm going to replace the previous system entirely with this and make it compatible with different morph sets that target different g-forces. I do also need to actually introduce custom morphs specific to this system at some point, instead of just offsetting the g-force morphs. The g-force morphs affect a very large area around the glute (lower back, upper thigh etc), and the effect of collision on morphs should be much more localized than that.

I hope you enjoyed the demo :)

-everlaster

Definitions of a few technical terms:

• Soft physics joint (or just soft joint): an individual physics simulated object that VAM attaches to each skin vertex in the soft physics simulated area (in this case the glute soft physics mesh). The joint moves the vertex when it detects forces or collision.

• Depth of a soft joint: the distance of the joint (and therefore the skin vertex it's attached to) from the average position of all soft physics joints (center of the mesh).

• Mesh pressure: outward force from the center of the mesh outward towards the joints, expanding the mesh physically. The more joints are pushed in, the more pressure there is to expand the mesh and preserve its volume.

• G-force: acceleration of a joint (in this case the glute joint) measured in g's (one g = 9.81 m/s^2)