Hi guys,
I'm fooling around with the shadow catcher. The setup is simple. I've got some geometry which:
I can do this easily with a rayswitch node, shadowmatte for primary rays and a surfaceshader for the other rays. Now, one thing isn't right yet. Shadows are cast from the objects, but they aren't there in the reflections.
So - is it possible to use the output of the shadowmatte in a shading network, so I can combine it with the backplate? I imagine this outputs a closure which can't be used?
I can get around this if I render out my shadow pass first and pipe it back into the shading network of the shadow catcher, but ideally wouldn't go this route. How is everyone else handling this?
Solved! Go to Solution.
Solved by zenop. Go to Solution.
Answering my own question here to help everyone else with this.
I can pipe both the AiShadowMatte and the projection into the AiComposite shader. Then create an AiRayswitch and pipe the AiShadowMatte into the camera ray input, and the AiComposite node into all the other ray type inputs.
Practically what this does is combine the flat projection with the shadows, and then we replace the primary rays with the shadowcatcher again.
@Zeno Pelgrims
This workflow still leaves a lot to be desired though. I've made some further tests with it, adding some nodes to your setup but still not getting the results one would hope for:
Beloved Solidangle, what is the correct workflow for getting physically plausible results?
Cheers,
Alex
I have been trying to achieve what Alex tried here but without any luck. I have a different lighting setup but I can't get the reflections to look right. Does anyone have att better solution for the reflections?