Nicolas Heluani

Ok...
I think I have an answer to my own question.
Leaving it here just in case someone else is wondering.
Basically "recipe as tool" is probably the best way to do it.
But sadly there is no way to automate this except on script.
There is a Recipe builder, but that will require to move all contents into the recipe builder and then export.
To import will require another script to install the recipe hda(unless you want hundreds of materials to be loaded at startup)and then call it in.
I know this sounds cumbersome but once set up you will have material library that can easily be work upon and updated on demand, add different render engines materials and such.

Maybe I am over complicating this? Basically I hope I am over complicating this.

EDIT: it seems this can also be achieve by actively removing all __activate__ and hiden primitive parameters added by the edit material network this totally then feels it can be an added function to the tool just dont add this params and function as a new material library.

Hi there!
Is it possible to import materials from a usd file not as references but as actual materials into the material library.
I know you can reference the usd and then use edit material network to recreate the network. But that doesn't serve my purpose. As it just add a lot of complications to what should be a simple task, as the edit material network is basically doing half of the job already.

I have basically rebuild my material library as usd files. Now I can easily access them individually but I will like to be able to open a material as a completely new material so I can customize a lot of changes and then export them as unique material.

TLDR: Is it possible to use usd material as "recipes" for materials?

I feel this should be possible, you basically have all the info there, right?

I didn’t mean realtime renders. But apps that put Realtime performance over procedural workflow as imperative. Soft like substance painter or zbrush excel as they focus on realtime feedback while working on destructive workflows. While Houdini puts premium on making sure all tools are procedural and talking to each other. A few years back I would discard the idea of SideFx even exploring this as the market is pretty much dominated by those two and Blender comes as a good second option. But with cops, Vulkan and the already powerful tools within Houdini. At least substance painter seems very achievable almost as a by product of the improvements they made.