procedural population variation

[dandelO]

The reddish and black Bots are rendered with a Get position node inside the Parts-shader.

How did you achieve that, Volker?
I can make it work if the 'get' node is contained within another node in the parts shader but not if it is just directly inside it.

EDIT:
Ah, I just read above, I guess it's down to the Mac/Windows thing, then.

Hetzen: Try placing all the functions inside another container node, like a colour adjust shader. I think it should work...

[Volker Harun]

Checked ... On my WinSys the render does not work ... this is stupid ...
Well, maybe we can get TG2 soon for iPad and Steve Jobbs is buying Planetside ,-)

[Hetzen]

Well, just deleting the left hand blues, and placing the right hand side PFs in a colour adjust shader seems to be doing a GI pass. So far so good.

[Hetzen]

I think that works Martin. I'm stopping this render and try it on some teapots. As you do.

[dandelO]

It does, pop the left hand nodes in a colour adjust, then assign the square scalar to the input of the colour adjust shader. That then goes to the colour function fine.

Disabling the other unused nodes won't work, you'll have to delete them or, place them inside their own container. I think a group node should work like the colour adjust shader, it's a good node to use as a container, though, as it only has one input.

[Volker Harun]

I am sitting here eagerly awaiting the first images ,-) ... I stopped working on my WinSys, as it alway shuts down after 10 minutes due to overheating.

[dandelO]

I reduced the render size for quickness but just making a new container does the trick.

*Edit: Using a group node works just fine as well. It seems you just can't have the functions on the same level of network as the desired shader you're inputting to.

[Hetzen]

Hrm. The red PFs on the right hand side do not work. I'll try the blues now.

[dandelO]

Wahey! The famous teapot!

[Hetzen]

Oh Yeah. It's not a test without a teapot.

The Blues are working, without RayTrace errors  . Shame about the PFs not.

[j meyer]

Very interesting!
Has anybody tried with RTO unchecked? Just as a test to see if it's
related to the other PF problems with objects.

[dandelO]

Aye, bummer.

I can still get some degree of fractal variation by setting the variation functions as colour function of a surface layer then, using a power fractal blend shader on that. The main function still works correctly, supplying the world-space variation but, the fractal is still only applied to the object-space. It works but it isn't quite 'there', yet...

[Volker Harun]

@jmeyer ... interesting idea ...
@all ... what happens when combining the Reds and the Blues with an Add-Colour or Multiply-Colour?

[dandelO]

Good thinking, J Meyer. It's still the same, though. The fractals are still only applied per-object, not per-population.

[FrankB]

I'm using a fantastic tree model that I made for simple testing purposes...

[attachimg=#]

.obj

pure eye candy, Martin