Q1 - Is there a reason why we need to plug in the output into all the texture UV’s? In theory will I notice any performance gains (no matter how small) just hooking it up to my base color and using regular UVs for the others? (Early tests show no difference, but I’m going off of FPS which ofc varies)
Q2 - Lerping between Bump Offset and regular UV’s based on pixel/camera depth seems to give me big performance increase, is this just a case of shoddy profiling or is it actually doing something useful?
Sorry for the late reply, but is there a reason a lot of the materials provided by epic make use of distance blend to flat normals? Is it because the shadowing looks off in the distance?
I guess reducing obvious texture repetition and lowering overall noise in the picture is the reason for that. Lerp node does not affect affect performance.
What you are asking about in this thread is dynamic flow control. It is roughly described in this thread and this post in particular.
Bit out my league unfortunately, only exposure I’ve had to shaders is working on this terrain material which has been an on and off affair for the last year, basically me ripping code from various resources and fumbling onto the solution without knowing exactly what I did, finding it a bit difficult to be able to visualise in my head what’s actually going on, previewing individual nodes kinda helps, but not in all cases, slowly getting there; but not nearly as quickly as when I began learning blueprint scripting (my first exposure to programming).
Will be good to come back to those two threads in the future when I’m a bit more knowledgeable, cheers!
And I’ll definitely upvote those tickets, hopefully by the time I get a grasp on things it would be implemented
Another reason they may do that is so they can use a cheap diffuse-only in the distance, on a LOD (but to avoid the pop, they fade it out before the LOD kicks in).
