Sub-Surface Scattering
Forum rules
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Sub-Surface Scattering
Hi,
I've rendered a scene using LuxCore and Cycles, and although I like LuxCore's shaders, lighting and shadows better, to me Cycles wins in the SSS area.
I used Random Walk SSS in Cycles, which is the same algorithm as LuxCore uses, unless I'm mistaking.
LuxCoreRender version: Cycles version: I find it hard to achieve the same visible SSS light falloff in LuxCore as is present in the Cycles rendering (the gradual, glowy light transitions in the yellow, soft parts of the pumpkin).
Which parameters of the Glossy Translucent shader and/or Homogeneous Volume node should I adjust? Here's my SSS shader node setup for the soft yellow parts (scene is modeled in the correct scale for a large pumpkin): Any advice will be much appreciated, thanks in advance.
I've rendered a scene using LuxCore and Cycles, and although I like LuxCore's shaders, lighting and shadows better, to me Cycles wins in the SSS area.
I used Random Walk SSS in Cycles, which is the same algorithm as LuxCore uses, unless I'm mistaking.
LuxCoreRender version: Cycles version: I find it hard to achieve the same visible SSS light falloff in LuxCore as is present in the Cycles rendering (the gradual, glowy light transitions in the yellow, soft parts of the pumpkin).
Which parameters of the Glossy Translucent shader and/or Homogeneous Volume node should I adjust? Here's my SSS shader node setup for the soft yellow parts (scene is modeled in the correct scale for a large pumpkin): Any advice will be much appreciated, thanks in advance.
Last edited by MetinSeven on Tue Nov 03, 2020 8:28 am, edited 3 times in total.
visualizer • illustrator • animator • 3D designer — metinseven.nl
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
Experimented some more.
I managed to get a bit more visible SSS light falloff inside the pumpkin, but still not the glowy liquid SSS of Cycles.
New Luxcore rendering: New node setup for the soft yellow parts:
I managed to get a bit more visible SSS light falloff inside the pumpkin, but still not the glowy liquid SSS of Cycles.
New Luxcore rendering: New node setup for the soft yellow parts:
visualizer • illustrator • animator • 3D designer — metinseven.nl
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
Another slight improvement. Turned on Multiscattering in the new node setup.
New rendering: Node setup for the soft yellow material:
New rendering: Node setup for the soft yellow material:
visualizer • illustrator • animator • 3D designer — metinseven.nl
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
Slightly closer again, using a Heterogeneous Volume now, and turned down Glossy Translucent material transparency a bit further...
Nodes:
visualizer • illustrator • animator • 3D designer — metinseven.nl
Re: Sub-Surface Scattering
I think we need random walk and the improved principled shader to avoid these problematic
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
But I seem to remember one of the Luxcore devs confirming my question if Luxcore uses Random Walk SSS.
visualizer • illustrator • animator • 3D designer — metinseven.nl
Re: Sub-Surface Scattering
Your scenes seem to have quite a different lighting. The cycles version looks muc more dominated by the distant (sun?) source coming from the top left wiht a weak diffuse background, while the LuxCore version is more dominated by the latter. I don't think this explains all of it, but at least on the drops its part of that constant gradient you describe. Maybe you could use an identical HDRI for both scenes?
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
I'm using the same HDRI with the same orientation. Luxcore's diffuse shader gives a very different result on the background, hence the visible difference. But if you focus on the pumpkin you see similar brightness.
Nonetheless, I'll change some light settings to see if that really changes the SSS effect.
Nonetheless, I'll change some light settings to see if that really changes the SSS effect.
visualizer • illustrator • animator • 3D designer — metinseven.nl
Re: Sub-Surface Scattering
Have you read this? https://wiki.luxcorerender.org/Subsurface_Scattering
It explains some of the fundamental principles of using SSS in LuxCore.
It would be good to have a simple testscene that looks absolutely identical between Cycles and LuxCore when using a diffuse shader, and only looks different when using the SSS shader. Then we can iterate on it until they look identical with SSS.
You could use the LuxBall scene from here as a starting point, it already shows the same result between Cycles and Lux when using a diffuse/matte shader:
It explains some of the fundamental principles of using SSS in LuxCore.
This is correct.MetinSeven wrote: ↑Tue Nov 03, 2020 6:12 am I used Random Walk SSS in Cycles, which is the same algorithm as LuxCore uses, unless I'm mistaking.
Some notes about your last setup:MetinSeven wrote: ↑Tue Nov 03, 2020 9:15 am Slightly closer again, using a Heterogeneous Volume now, and turned down Glossy Translucent material transparency a bit further...
- You might want to leave the transmission color of the glossy translucent material white. Then you can control the color inside the volume solely via the absorption color.
- On the same topic, you might want to leave the scattering color of the volume white, so only the absorption color influences the color.
- Only use a heterogeneous volume if you need to texture volumetric properties with a spatially varying texture. If the properties are constant over the whole volume, it is better to use a homogeneous volume since it has much better performance.
- Because the glossytranslucent material does not refract light rays, the volume IOR does not matter here. It only really matters in combination with a glass material.
It would be good to have a simple testscene that looks absolutely identical between Cycles and LuxCore when using a diffuse shader, and only looks different when using the SSS shader. Then we can iterate on it until they look identical with SSS.
You could use the LuxBall scene from here as a starting point, it already shows the same result between Cycles and Lux when using a diffuse/matte shader:
- MetinSeven
- Posts: 137
- Joined: Sun Aug 18, 2019 10:19 am
- Location: Netherlands
- Contact:
Re: Sub-Surface Scattering
OK, adjusted the light strength of both the HDRI environment as well as the direct lights to get more light energy going and match the Cycles scene more.
It looks a bit brighter now, but still not the nice light falloff inside the volumes.
It looks a bit brighter now, but still not the nice light falloff inside the volumes.
visualizer • illustrator • animator • 3D designer — metinseven.nl