Weird result on light passing through glass with low glossiness.

[ogneed]

I've made one test and get unexpected result.
Here you can see 4 glass bowls.
1 and 3 have low glossy stripes at outside.
2 and 4 bowls have low glossy stripes inside.
Inside each bowls I've placed 1 cm spherical Coronalight, All those lights are instances, and have the same brightness.
Material on 1 and 2 bowls have slight Absorbtion
Material on 2 and 3 bows don't have Absorbtion

Strange why is it light spot on 1 and 3 bowls much darker then on 2 and 4 bowls.

[Ondra]

My guess is that in cases 1&3 the setup would create excessive fireflies, so the light energy gets clamped away with max sample intensity. Try rendering with MSI = 0 or bidirectional path tracing to confirm.

[ogneed]

Ok, I already do both variants. When is will done - I'll update this message.

Update: Confirmed! Both variant return correct light spot.