Thanks for reporting @derekflood . It seems to come form the simplification. It has been added to the list of things to fix.
EDIT: this is fixed for 5.7.
Thanks for reporting @derekflood . It seems to come form the simplification. It has been added to the list of things to fix.
EDIT: this is fixed for 5.7.
That’s great news! Also I was really excited to see refraction working, that’s quite impressive
In the meantime, for other folks who may be encountering this, since this also effects the Substrate StandardSurface Translucent, a possible workaround is to go into the Material Function and bypass the coat’s Vertical Layer.
That looks so good. Ice/Crystals are going to be amazing…
Are you using translucent-colored transmission for the shading model? I’m trying to replicate the look and having some inconsistencies.
I’m having an issue in 5.6 where my clearcoat slab has reflections that shouldn’t be there when the material becomes more complex. It happens when the bottom layer’s roughness is less than Lumens reflection’s “Max Roughness to Trace” (which I need set to 0.8 or 1 for HQ automotive scenes). When F0 of the coat is set to 0 and only the simple version of the material is present there are no reflections as expected.
When the complex version is switched on reflections appear on both materials even with F0 on 0. If the complex material is removed these reflections stay present until the level is reloaded. The complexity seems to be based on number of instructions rather than any specific nodes used.
This problem does not occur in 5.5. The reflection environment is a skylight with a HDRI Specified Cubemap. The only CVars that I have found which stop this happening are r.Lumen.Reflections.Temporal 0 and r.Lumen.Reflections.Temporal.NeighborhoodClampScale 0.
Yes I am.
I am using Unreal Engine 5.6. and trying to use ray-traced translucency. With legacy material system I was having a lot of reflection issues, most of which resolved as soon as I switched to substrate, so that’s why I am reporting the remaining issues here.
ISSUE # 1:
Reflections are a little bit dull with ray-traced translucency compared to raster.
ISSUE # 2:
Raytraced Refractions are not letting the GI pass through objects inside translucent objects. but when i disabled ray-traced refraction and kept ray-traced translucency enabled, GI inside translucent objects gets better.
right now, the values that I am using are:
Max. Primary Hit Events = 64;
Max. Secondary Hit Events = 64;
GI issue only happens with ray-traced refraction. Only primary hit events are causing the problem. Anything higher than 1 starts to darken the glass as I increase the primary hit events value.
Quick question:
I am about to move from legacy material system to the substrate. so i will be converting my materials and will try to make sure each material type has proper shading.
should i build each using “Substrate Slab” or using “Substrate Shading Model“ node will create material quality and accuracy equivalent to “Substrate Slab“.
I have following material types in legacy materials which use specific shading model right now:
The conversion process is automatic, you shouldn’t need to do manual conversion to achieve parity except in rare circumstances.
Does the path-tracer produce any usable ground truth?
same issue in the lumen thread. you could cross reference. ; )
i used Path Tracer for reference, and I posted a more detailed comparison in the Lumen feedback thread. there you can check the path tracer result as well.
I am using UE 5.6. And yes, it should happen automatically, but when I look into the material, it doesn’t appear to be a substrate setup.
Based on my understanding, a substrate node needs to be connected to the front material option.
Right now, I am in a hurry and don’t have time to learn the substrate nodes from core. but i would like to in future so for time being to make things quick and accurate, I want to know if using “Substrate shading model“ node and setting a specific shading model on this node based on my legacy materials will give me accurate substrate materials compared to if I build a specific type of materials using the “Substrate Slab“ node.
Could you rephrase your question, perhaps?
If I understand you correctly, you’re wondering about how the Substrate shading models relate to the legacy shading models? I assume by ‘accurate substrate materials’, you mean ‘comparable results to legacy materials?’
So, it depends on what you’re trying to do. If I remember correctly, the Substrate slab exists for all common shading behavior that isn’t a specialty case (IE, Eye, singlelayerwater, hair, volumetric). So, as long as you’re not using any of those, you should be able to achieve comparable results.
As for the shading models question, that really depends on what you’re trying to do. For anything opaque, it is functionally identical with a handful of exceptions. For transparency, it’s more complex. The concepts of transmittance and coverage are a completely different paradigm from UE4’s ‘opacity’ setting, so if you’re trying to author transparent materials you’ll need to learn the basics of those shading models. In terms of feature sets though, Substrate supports everything legacy supports and more. Use the legacy conversion nodes if you need to create a material that has parity with legacy.
Could one of the developers please answer a small technical question for me?
In the Substrate docs, they mention that substrate is energy-conserving. Is this energy conservation absolute across all layers, or just the base? Is it like Adobe’s standard surface, where absolute energy conservation is guaranteed for the base, but for additional layers it simply won’t add additional energy?
I wanted to know if using shading model node i will be able to get more physically accurate materials. I asked because substrate result looks closer to path tracer compared to legacy materials.
Anyway, I just decided to make materials from scratch; it will be a better learning experience.
here are images for my first question:
Left is diffusion and Right one is two sided wrap.
and here is material setup for diffusion. and i also tried to multiply the two sided sign with normal map before feeding into normal but still looks dark and it start causing artifacts by darkening the front sides of some areas.
I don’t know specifically how it’s working in Unreal, but in path-trace renderers (Arnold, Renderman, etc) subsurface scattering works well on objects with thickness, and not on flat thin-walled surfaces, because there is no sub-surface, no thickness to scatter in. Hence the non-physical, approximation approach of thin-walled translucency for leaf cards and such. So given that, I’d expect the results you are seeing.
Hopefully the devs can verify if these understandings are correct here.
I am confused about the Adaptive GBuffer and Blendable. Can someone explain it?
and if i have to choose between 1 or other how can I set it and from where? and which will be the default option?
I am trying to create a glass material. and I am using “Roughness“ to control base roughness of glass and while trying to add imperfections like fingerprints using “Second Roughness“ and weight.
Shouldn’t 2nd Roughness also blur out the glass based on roughness value? right now it’s just making the glass flat and clear, only stripping down reflections. only first roughness input blur the glass and 2nd seems useless. i wonder if it’s bug.
Am I doing something wrong? I thought the second roughness is to create imperfections, such as dust particles and fingerprint imperfections. I wonder if i am doing things in substrate correctly. Right now, I am using the 2nd roughness on all my material types for imperfections like dust particles and fingerprint marks.