Simple scene with a fairly low poly count, but made more render engine intensive with 60 bounces and a 5x5 pattern. Everything in the advanced settings were active other than motion blur. Live renderer used, approximate 20 minute render time.
Martin
nice.
out of interest, is there ever any situations where you can have too many ray bounces (ignoring time issues)? Can it create unwanted noise or artefacts in some cases?
edwardo
To be honest Edwardo I have no idea, as far as I know the bounces are used by KeyShot purely for getting through what should be transparent geometry, I can feel an experiment coming on ... :P
That'll teach ya
Martin
Results are in - 2 Minutes approximately per render, bounce count is in the image name ...
Martin
Quoteout of interest, is there ever any situations where you can have too many ray bounces (ignoring time issues)?
To determine how many bounces just count your surfaces. I usually add 2 more just for grins. I did not count your panes, but if there are 30, then it would be 60 + 2. Be careful if using offset surfaces, like a liquid in a glass jar, as it's easy to miscount. "Over-bouncing" just wastes render time.
Bill G
Hi Bill - There are 60 panes there so optimum should have been 122, but there is no real gain in this scene and angle beyond 84. Maybe there would have been if left rendering longer
Martin
Hi Martin
Nice tests,
For me there is no discernible difference between 9,10,11 and 12
8 bounces would be OK for that particular scene, any more is just a waste of CPU cycles
On the Vases scene I would say that 60 bounces is overkill In Extremis...
J
How strange, the minimum bounces i can use to get through the glass is 18, the higher the bounces in my case at least, gives better caustics :/
Here's the model if anyone want's to try it ...
Martin
Martin
I hate to say this, but I don't think there are any caustics in your image, all I can see is the reflection of the HDR you are using... I could be wrong though
J
Sorry J - my statement was ambiguous, in the vase image the caustics are not showing as expected, it's as if they get lost in the fat base of the model, but in 'general' the higher the bounce the better the quality of the caustics, i might be wrong though as i go by how it looks more than trying to be physically correct.
Martin
Not sure how the quality of caustics is related to reflective/refractive ray bounces...
However, more ray bounces can increase the 'depth' of caustics in some scenes.
Dries
Hey Martin... no need for the 'sorry', it's still a decent render
Here's one I did with your model... my first ever render with caustics active actually, it's not a feature I really use.
J
Got more sparkle than i did J - I shall have to revisit this and get my physics head on I think - I've got to ask though, did the profanities help ?
Martin
Feck yeah !! ;)
J
Here's an example of how ray bounces can impact focused caustics.
Dries
Cool result Dries :)
Martin
Yeah well it shows that 'bounces' do exactly what they say on the tin...
They bounce things around... ;)
Nice example Dries...
J
Here's another one.
Dries
Excellent examples there folks.
I'm having a current issue with a render that unfortunately I can not post as it is part of an advanced vehicle lighting scheme but sufficed to say, simply counting the refractive surfaces to a reflective surface, multiplying by two, plus two is not the answer.
As shown above, with no regard to caustics, I find I am not getting all the optical effect of multiple inner lenses when just using the math. There is a threshold, to ray bounces (with accuracy)/render time, but I haven't found it yet.