We’re wondering what are the plans for material layers. We’ve been using them non-stop for our first shipped product; it was really handy to remove alot of shader permutations by wrapping the logic of not always used features in MLs and MLBs (blend assets). We’ve even expended on this feature by adding few custom tools around it for our production. That being said, it’s been known that this feature was somewhat about to be dropped by Epic, as we start some new experimentations, are we to expect to see these disappearing in future versions to be released? If so, any plans to provide an equivalent replacement? Substrate is not a replacement for it at all as it does not provide the same use.
That being said, it’s been known that this feature was somewhat about to be dropped by Epic, as we start some new experimentations, are we to expect to see these disappearing in future versions to be released?
We don’t plan on removing material layers and do intend to improve the system or take it in a new direction if beneficial, but don’t have any additional details on that currently. The layer system can be used alongside Substrate.
Any feedback you have regarding the existing system is welcome!
Thank you Alex for the answer. That being said, a general feedback from the team was related to manipulating the layer assets. They are inconsistent vs. the regular parameters from the material instances. When you are copying or moving around the layers, some settings get discarded. We added the custom ability to save settings as child layers, maybe our implementation is flawed, but we wish the parameters handling was more consistent with the rest of the material instance UI / UX. Thanks alot!
Hello - just adding some more thoughts to this discussion.
We’re also using the material layers system extensively in our project. Artists love it: it’s extremely flexible, great for rapid setup and iteration, shared base material libraries, etc.
However we are concerned about the tradeoff between this flexibility and runtime cost - the generated shaders are typically much more complex than with a non-layer setup.
While it is possible bake out materials per SM, this is currently a destructive process and requires manual setup per asset. A nondestructive cook-time baking solution would be ideal:
It would allow us to preserve the material layers for artist convenience in editor.
Global bake settings could be applied (e.g. to modify the bake setup per device profile).
Easy generation of cheap versions of the materials for raytracing.
This is something that we may look at implementing ourselves in the near future, however it would be good to know if there’s any overlap with Epic’s roadmap.
I just wanted to post a reminder that Epic will be on holiday break starting next week (Dec 22, 2025) and ending Jan 5, 2026 and there will be no responses from Epic on public or private tickets, though you may receive replies from the community on public tickets.