Forum: Poser 14


Subject: I Am So NOT Impressed!

anupaum opened this issue on Dec 08, 2025 ยท 69 posts


hborre posted Wed, 10 December 2025 at 3:14 PM

nerd posted at 1:44 PM Wed, 10 December 2025 - #4502170

I suspect the problem emanates from the anisotropic node. Cycles removed that node from the latest version of their render engine. Instead they use the Glossy node to perform both normal gloss and anisotropy.

We setup SuperFly to automatically replace the legacy anisotropic node the new implementation. Basically we just plugged the old connections into the new node. The intent was to not break all the shaders that used the Cycles anisotropy. Clearly we missed something. The test cases we were using didn't fail.

I've responded to the support tickets submitted with this as well. Now we have to fix what Cycles broke.

I took a good look at the arrangement and ran a very simple render comparing the Anisotropic and the Anisotropic BSDF nodes. They behave very differently when connected to the PoserSurface node and rendered in Superfly.  I added the Edge_Blend node and rerendered the scene; I observed a slight change in difference.  Now, I recalled Bagginsbill stating that the Edge_Blend node was a hack for simulating Fresnel before there ever was a Fresnel node.  I replaced the Edge_Blend node with the Fresnel_Edge node and rendered.  I discovered that the renders exhibited the same effects except the Anistropic BSDF node was stronger under similar conditions.  My takeaway from this little experiment is to replace Cycles node as much as possible when adhering to the legacy PoserSurface node shader setup.  Otherwise, results will vary.