- edited description
Ultimate Skin Moisture - Node group improvements
Issue #543
resolved
The Ultimate Skin Moisture materials for DAZ look weird when imported. They look like wet ink instead of perspiration. In this example, I am using the Sweat material.
Here is what it looks like in DAZ:
Here it is in Cycles:
And here it is in Eevee:
You guys have done an impressive job converting the skin shaders from DAZ to Blender, so I was wondering if anyone has any suggestions to make this sweat look more realistic in Cycles. I have attached a .7z with the .blend file. I resized all the textures smaller to make the file size smaller, but that shouldn’t interfere with tweaking the shaders.
Additional info: the Ultimate Skin Moisture product in DAZ is a material applied to a shell that can be selected separately from the body mesh.
Comments (14)
-
reporter
-
Will look into this and let you know what I get.
https://www.renderotica.com/store/cat/149_checkQS/sku/62395_Ultimate-Skin-Moisture-v1-2-CORE-G8-8-1F
-
daz studio 4.15, blender 2.91, commit 18bfb2a
Thomas, it seems there are a number of bugs to be fixed here. Test scene included moisture.duf that uses the product above.
In this case the bump map is not connected to the shell, and the bump distance is set to zero. Thus the shell doesn’t get the sweat effect. If I fix this by hand then the shell works fine. Below an example with the torso material first iray then cycles with the fixed bump. This is the main bug.
Another minor bug is that the shell refraction is 100%. In this case the diffuse channel is not connected to refraction, and this may be good since a 100% refractive material doesn’t take into account the diffuse channel in iray. But then the diffuse channel should be pruned and this doesn’t happen, it is simply not connected. This doesn’t affect rendering though so it’s a minor bug.
Then there’s another minor bug that the viewport material is set as transparent. But in this case it’s the shell that’s transparent, not the figure, so the viewport material should be solid.
-
- attached moisture.duf
test scene for ultimate moisture
-
Thomas, I managed to reduce the test scene to a simple cube with a similar material that you can test without the ultimate moisture. The issue and solution are the same. Then if you can test with ultimate moisture it’s better of course.
-
- attached shell-bump.duf
test scene similar to ultimate moisture, with a simple cube
-
- attached bump.jpg
and texture
-
repo owner
I think it should work now. I don’t have the assets, so the textures are missing when I test, but the skin looks shiny anyway. What I haven’t done is correcting the bump distance for image size, so the distance is 0.2 mm if there is no other info in the daz file.
-
repo owner
I’m not sure how the cube is supposed to look, but it doesn’t look terrible.
-
As for commit 18c6540. The viewport transparency is fixed. The diffuse pruning is not fixed.
The bump distance is not good, it has to be computed the same as reported in
#433For the cube I’m not sure to understand your question. Just render in daz then in blender to compare the look. Here it is anyway.
edit. The bump distance for the cube example should be as below.
area = 1500000 mm2 density = 512*512 / 1500000 = 0.1747 pixel/mm2 bump = 3 / sqrt(0.1747) = 7.177 mm
-
repo owner
Should work now. The geometry factor disappeared because the shell material is associated with the shell, which is an empty mesh. So I had to do some work to dig up the real mesh. If you turn off Merge Shells, the shell material Default-3 is not corrected for this reason.
As for pruning, you have probably turned off the prune nodetree option.
-
Thomas commit d3a4c52 works great you got it. The prune issue is gone too.
It doesn’t work fine with the principled option not sure why. Will look at it and let you know what I find out.
-
This is fixed with
#558we may mark as resolved. -
repo owner
- changed status to resolved
- Log in to comment
