Foggy morning - Fog Tests
Posted: Wed Mar 20, 2013 6:00 pm
Hi, I have done several tests with the camera inside a sphere, following Enricocerica's advices.
I placed a 1 meter diameter sphere with normals pointing inside, at the camera's location, by copying the Pos value of the camera into a Placement node. I used a second placement node just between the sphere mesh/obj node and the placement node used for location, just for the sphere scale.
I noticed several interesting things :
- The fog is outside the sphere. So, by adjusting the sphere scale in Octane, you can adjust the distance where the fog starts to be visible.
- Adjusting the opacity of the sphere as well as the roughness allow interesting controls on fog density.
- Direct lighting seems to give better rays than Pathtracing, through trees for example.
- For my own, I used AO mode 3 with Diffuse Depth 2-3.
- Reflection 0.0.
- Transmission 1.0.
- Scattering direction is -0.5 on every image.
You may notice some unexpected artifacts on right and left sides of image RangerStation_136.jpg, for example. I tried with Smooth disabled on the sphere, and these artifacts disappeared ! See image RangerStation_137.jpg.
So, Smooth mustn't be used on the sphere !
I also noticed that Pathtracing and PMC kernels almost remove the blue color when the fog is strong.
High grain with Pathtracing and PMC is due to fast render (around 100 samples only).
Don't forget to disconnect the Sphere node to make the point with the focus pointer, otherwize you will be blocked by the sphere surface if your focus point is outside the sphere !
I placed a 1 meter diameter sphere with normals pointing inside, at the camera's location, by copying the Pos value of the camera into a Placement node. I used a second placement node just between the sphere mesh/obj node and the placement node used for location, just for the sphere scale.
I noticed several interesting things :
- The fog is outside the sphere. So, by adjusting the sphere scale in Octane, you can adjust the distance where the fog starts to be visible.
- Adjusting the opacity of the sphere as well as the roughness allow interesting controls on fog density.
- Direct lighting seems to give better rays than Pathtracing, through trees for example.
- For my own, I used AO mode 3 with Diffuse Depth 2-3.
- Reflection 0.0.
- Transmission 1.0.
- Scattering direction is -0.5 on every image.
You may notice some unexpected artifacts on right and left sides of image RangerStation_136.jpg, for example. I tried with Smooth disabled on the sphere, and these artifacts disappeared ! See image RangerStation_137.jpg.
So, Smooth mustn't be used on the sphere !
I also noticed that Pathtracing and PMC kernels almost remove the blue color when the fog is strong.
High grain with Pathtracing and PMC is due to fast render (around 100 samples only).
Don't forget to disconnect the Sphere node to make the point with the focus pointer, otherwize you will be blocked by the sphere surface if your focus point is outside the sphere !
