Twist and shear Shader issue

[TheBadger]

Hi.

I am having some good fun with the Twist And Shear Shader (hereafter the TSS)
Its great on the macro level, no complaints or questions.

But Im having issues controlling Finer details.
I have hooked my TSS up to a Distribution Shader via the child input. But I think this may be a problem, because I don't know if I want everything before the distribution node to be effected by it... Do I? The strata I suppose, but everything?

1) The restrict by altitude is causing problems. I get crazy distortions and surface explosions. Also, my test renders go from very fast to very slow, and my entire system slows down too. Currently the TSS is affecting everything, and I wanted to limit it to higher areas and larger rock formations.

2) None of the fractal breakup node parameters appear to do anything. For example, Noise flavour, Variation Method, buoyancy From, clumping. I don't know what I expected these setting to effect, but I thought I would see something happen.

As it is I have a nice and desirable look over all. But there are some sharp spikes I would like to soften, and I would like to make the edges of the displacement more rounded over all, to appear more eroded and suggest a wetter climate. I would also like to limit that softening to the larger more prominent areas, so spiky displacements nearer the ground may remain.

Here are some image examples to clarify my intentions.

Thank you.

[Dune]

What I always do is add the TSS as child to a surface shader and blend that by a distribution shader (or whatever). Setting the fuzzy zone helps softening the angle at which it twists at certain height. I think it affects all before.

[TheBadger]

Ulco, its funny! Thats what I thought I should do. But I had never used the TSS for anything serious, only played with it a bit. So I went on the forum search and found instructions. They told me very specifically to just use the distribution S. But I thought that sounded a little wrong. I guess I am ready to start thinking for my self, a little, maybe.
Thanks for clarifying it for me, especially about using the child input.

Actually that thread I mentioned did have some useful advice too. But it does not work right with out that surface node.