OTOY Forums

I did some test to see what different result I get with Daylight, PMC and PT kernels.
I only did some little noisereduction in postproduction.
The setup is with hdri background and one white-emitting object from the back of the camera.
The DL does not show the material as well as the PMC and PT kernels.
The material is plexiglas (from DB) and adjusted a little to get the right pink look I was looking for.
I must say there is almost no difference in PMC and PT.
Maybe it is better to test with the same samplerate or rendertime?
(the image names show the samplerate and the rendertime aswell.

Maybe someone has a good idea how I can test better test the kernels with this scene to learn?
greetz,

Hi,

PMC and PT should converge to the same image, as they both calculate full global illumination. They use a different algorithm so depending on the scene one of them may work better than the other.

Direct lighting only calculates direct lighting, reflections and refractions, but it should render much faster than PT or PMC. if you use DL the render may look better if you increase speculardepth a bit (to 16 or so). This will make it trace more refractions inside the plexiglass before it stops tracing.

For the material, have you changed the transmission color or the absorption color under medium? The first models absorption on the surface while the latter models absorption inside the material. Usually using medium absorption looks better (since you specify what light is absorbed you should set the absorption color to bluish green).

--
Roeland

Hi Roeland,
Thanx for your help and explaining.
I did a new rendering with increasing speculardepth from 5 (I think?) to 16, and indeed.. look at the improvement within DL.
I had not tweaked the plexiglas material bij adjusting the medium node, but only by changing the RGB transmission node
and also a few changes in opacity and reflection.
The next step I tried is the medium absorption node and I set it to a darkgreenblue as you suggested.
You can see the settings and results.