playing with caustics

[epilectrolytics]

Have you experimented with PGI caustic cache and volumes? I mean the light rays in the fog/smoke type? I can't get good results.

Thanks!

I have not yet tested caustic cache with scattering volumes.
OIDN does not work well with scattering yet, so I do this kind of work still in Cycles.

[provisory]

I have not yet tested caustic cache with scattering volumes.
OIDN does not work well with scattering yet, so I do this kind of work still in Cycles.

OIDN works perfectly for me with scattering volumes and BiDir.

[epilectrolytics]

I think I have not tested this combination either
My experience, in architectural settings with PathOCL is like what @FarbigeWelt describes here.

Once the current caustics craze is over, I will investigate the OIDN-with-volumes topic further

[epilectrolytics]

First attempt at the iconic cubo prism with hybrid path.
I didn't want to render longer than a day and chose 500 samples/frame which was probably no good idea because there is some visible noise from fireflies, also because OIDN was at 70%.
300 frames in 24h on Ryzen 2700X + RTX 2070.

Glass material is 80% dispersion + 20% roughness, a big and a small point light at the same location for lighting, no background.
In order to make looping possible after a 180° turn I had to mirror the ground plane material and even to remodel the prism which was not 100% symmetric causing a shift in caustic patterns between 0° and 180°

[provisory]

Nice!

[B.Y.O.B.]

Awesome

[Racleborg]

Looks really nice!

[epilectrolytics]

Thanks!

[CodeHD]

I watched it a couple of times now, and if there is any residual firefly noise, then I can tell you it is not noticable to someone who doesn't know where to look for it

[epilectrolytics]

You're right, the noise is visible in the image files and the Apple ProRes 422HQ master clip, but with the h.265 compression for the uploaded file it is mostly smoothed out.