Page 3 of 5

Re: Light tracing and pool

Posted: Thu Aug 08, 2019 4:23 pm
by Fox
@ lacilaci
The last render looks very good.

Re: Light tracing and pool

Posted: Thu Aug 08, 2019 4:27 pm
by lacilaci
Thanks guys, it's better than nothing, but I have to be careful about camera angles since I won't get any refraction on those caustics :?

Re: Light tracing and pool

Posted: Thu Aug 08, 2019 5:46 pm
by provisory
You could use some wavy textured emission (with disabled DLS) on the bottom of the pool.

(Of course, this is a crappy, fake solution that even Cycles able to do, but it looks, there isn't any better solution here either at the moment.)

Re: Light tracing and pool

Posted: Thu Aug 08, 2019 6:10 pm
by lacilaci
provisory wrote: Thu Aug 08, 2019 5:46 pm You could use some wavy textured emission (with disabled DLS) on the bottom of the pool.

(Of course, this is a crappy, fake solution that even Cycles able to do, but it looks, there isn't any better solution here either at the moment.)
The problem with this "solution" is that the texture would emit light even in shadows unless I would actually made the texture according to lighting... and still it wouldn't be properly distorted nearing the water level.

The way I would fake it in cycles would be creating invisible plane at water level with a mask in shape of caustics and that plane would be visible only for shadow rays and then adjust water body shader so that it transmits enough lihght. It is still a bit weak cause it's just a shadow trick...

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 11:25 am
by Fox
I tested the BiDirVM. The caustics didn't got sharp at 4000 spp. Maybe better luck with dispersion and scattering off, or smaller start radius.
vm startradius 0.002.jpg

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 11:39 am
by lacilaci
Fox wrote: Mon Sep 02, 2019 11:25 am I tested the BiDirVM. The caustics didn't got sharp at 4000 spp. Maybe better luck with dispersion and scattering off, or smaller start radius.
vm startradius 0.002.jpg
Close... I'm hoping that the new soluiton will be faster and better still :D

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 1:57 pm
by Sharlybg
I tested the BiDirVM. The caustics didn't got sharp at 4000 spp. Maybe better luck with dispersion and scattering off, or smaller start radius.
Nice test but it is not the kind of thing we should guess. it should work right out of the box. it is too computational intensive for that.

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 2:48 pm
by Fox
Sharlybg wrote: Mon Sep 02, 2019 1:57 pm Nice test but it is not the kind of thing we should guess. it should work right out of the box. it is too computational intensive for that.
It's pretty intensive indeed.
------

On the side note, i remade the water surface almost mirror smooth, highest wave is 1cm (for the 20m pool it's very little). All the cylinder parts under the water are visible with glass material (no rough or null mix). It was visible before, but was hard to make out, because it's flat as pancake in the far end.

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 6:01 pm
by FarbigeWelt
Fox wrote: Mon Sep 02, 2019 11:25 am I tested the BiDirVM. The caustics didn't got sharp at 4000 spp. Maybe better luck with dispersion and scattering off, or smaller start radius.
vm startradius 0.002.jpg
Wow! What a relaxing view! Hope, water as warm as comfortable to dive in slowly.

How much time did this render take on what cpu?

Re: Light tracing and pool

Posted: Mon Sep 02, 2019 6:55 pm
by Fox
FarbigeWelt wrote: Mon Sep 02, 2019 6:01 pm Wow! What a relaxing view! Hope, water as warm as comfortable to dive in slowly.

How much time did this render take on what cpu?
It must have been 8-10 hours overnight on e5 2620 v4 cpu.
I play with the scene more, see if something better comes out of it.