After the 5.6 update, there was a change to the material function that interprets OpenPBR translucency materials that introduces thin film to all materials, regardless if Thin Film weight is at a value higher than 0. This is due to thin film thickness being piped into another part of the graph before thin film weight can influence it. Below are images detailing the problem, as well as my fix in the proper material function. I’m hoping this helps people out who are using OpenPBR MaterialX.
Below you can see what should be a glossy clear plastic material that seems to be a bit rough on import.
Within MF_Substrate_OpenPBR_Translucent material function, we can see the problem. The image below is from the inside of the Thin Film section within the function. thin film weight and thin film thickness get input, and then piped into reroute nodes.
This in itself is fine, but thin film thickness gets immediately piped into the specular portion of the network after the reroute node.
So you can see that thin film thickness is applied to the material regardless if thin film weight is above 0.
By introducing a simple multiply node before thin film thickness gets rerouted, we fix the issue, allowing thin film to work as intended.
Which results in the material now looking correct.
