thanks for the quick response:
-remove the falloff map from opacity - it makes no sense to use it if you are using refraction
It is hack, yes.But I have forgotten to mention that we are using Corona in an unusual environment: it is about cartoon characters. So it doesn't have to be physical correct.
In this case the prop is rendered against a bended grey background waiting for approval. The refraction gives it a black 'outline' which is killed by the falloff.
I removed it- rendertime decreases only 2 seconds. Noise seems to be comparable.
-use path tracing as both secondary and primary solvers
ok...For other reasons not an option(but only very little difference)
-why did you increase light samples multiplier to such crazy value? is there any reason for this? leave it at default 2 or even reduce it to 1
-if you want better time/quality ratio for glossy reflections - increase "gi vs aa balance" (not too much, check values like 16, 24, 32...)
this was due to try&error:I increased GI-AA up to 64 with little difference. cranking the multiplier up to 8 however reduced the noise. rendertime increased from somewhat 5 to 8
-why are you using both time and pass limit? again, is there any reason? I think you should use one or another, preferably pass limit only, to have the same image quality across all of the frames
I'm sorry, that this was misunderstandable: I use pass limit only. But because of the huge amounts of frames that have to be rendered, rendertime per frame is crucial.
--reset render settings to defaults
this was the key! :) it rendered much better. I wondered why because basically I only changed the multiplier. It seems that the reason for the noise is the 'use' Corona env 180,180,180 RGB setting.
Can you shed some lights on this?
---it seems that is has nothing do with corona or max env: it is only the bright color which is causing the noise.
many thanks Martin
