I’ve tried fiddling around with light cache settings for ages and help files but I can never get it accurate. Not even 2000 subdivs.
The tabletop is 50mm thick and penetrates through a thick solid wall, yet I’m still getting some insane light leak:
You can see it under the table.
1000 Subdivs
DMC .85 Adaptive Amount
Adaptive Threshold .01
Sample Size .01 (default value is terrible)
I also have it set to Light Cache for primary and secondary to illustrate the problem.. it looks terrible with irradiance mapping in the primary as well.
There is also a skylight behind the window.
Are there any more tricks with V-Ray I can pull?
Apart from Brute Force
Edit: Retrace threshold and check sample visibility do not solve this light leak either.
Least Squares Fit exaggerated the problem from the light Cache.
But the main problem is the mesh itself. I’ve imported the mesh from SketchUp to Maya.
Now instead of the wall consisting of 4 planes being a single object, in Maya it exists as 4 single planes grouped/parented together.
This means that V-Ray is reading the object as four single planes instead of being a solid object so thus the light leak occurs.
I was wondering if if there was a way V-Ray could identify this a little better or does it have to be a single node for V-Ray to consider the wall to be a solid object?
I’m not sure this matters actually (in fact, I’m pretty sure it doesn’t :)) so it’s a bit odd that it has any effect on the GI. Would it be possible to get me just that part of the scene?
Okay, that makes sense. I’ve tested and you’re right (of course).
Here’s the part of the scene. I have a feeling it’s my settings (which I’ve made faster to render and make the leak more obvious), but 99% of the scene is okay just these two junctions I’ve highlighted.
It’s worse with full materials (which are all VRayMtls)
Thanks Svetlozar, was hoping that was not the case as it leads to higher render times. However this is the best way to do it as you say.
One last thing, what is the “technical” difference between “Detail Enhancement” and “Retrace Threshold”? Ae they both brute force methods used to tighten up the indirect GI but one is used for Irradiance Map and the other is used for the Light Cache?
Just curious for when to use one or the other or both.
Yes - “Detail Enhancement” and “Retrace Threshold” are Brutte Force method of calculating difficult areas for IM and LC.
“Use Retrace Threshold” won’t increase your rendertime too much since it only handles a very small part of the entire render.
