I'm coming from a different package and starting to light some scenes in Houdini. I'm running into a couple issues and was hoping to help to figure them out.

I'm lighting an exterior scene with an environment light and a distant light(to replicate the sun). The distant light has razor sharp shadows and i want to simulate some haze in the atmosphere and soften the shadows a bit. I can't seem to do it even though there are settings such as orthographic width.

Secondly what rendering engin should I be using for large exterior scenes? I'm used to v-ray. With PBR it seems to be having trouble getting a decent GI bounce in between large objects. Should i use raytracing with a indirect light? And finally what pass would should me the equivalent of vray GI render element?

Thanks for your help, look forward to being part of the Houdini community!

Use the “sun” light mode instead of distant. Then you have an angle control to set how soft the shadows are.

https://www.sidefx.com/docs/houdini/shelf/skylight.html [www.sidefx.com]

You can use this Sky Light shelf tool (Lights and Cameras tab) as a starting point. It creates a light with this sun mode by default (Type: Sun) and also a skylight which is based on a Siggraph paper, few years back. However Houdini expects that you are building a scene on the Mercury or even closer to the Sun, because the visible angle of it on the Earth is much smaller, just slightly more than half degree. So you should set the Angular Size in the skylight node (Sky Env.Map/Sun tab) to 1/4 (or 0.25, but Houdini calculates the /) - it seems ok for me based on my measurements.

The blurry penumbra of the direct shadows on a sunny day is caused by the visible angular size of the Sun, not by the haze. If there is haze, you can see the effects of it as a secondary “light source”, a glow-like gradient along the edge of the direct shadows, metallic-like secondary reflections around the crispy highlights, etc. You can set it in the skylight node with the Haziness (Sky Env.Map/Sky tab).
Clouds make the scenario more compex …

xilofoton
it seems ok for me based on my measurements.

Can you show your measurements? the angular size should be set to 0.5, not 0.25.

As an earth scientist, I'm bit of a geek in terms of natural phenomena. So I always check renderers how accurate they are

Did a double check earlier, but now it's triple:


Slit width: 0.5 degrees (cylinder with 720 sides, one blasted)


Eclipse with a poly sphere:
Distance: 150 000 (rounded)
Radius: 695.7
Scaled down the real scales by one million.
We can see the dotted fringe effect because of the low-poly sphere.


The previous two just show the direct appearance of the Sun (Render Light Geometry ON), but this is a diffuse poly plane in the center of a sphere with 1 unit radius. I beat a very tiny hole on the surface with a cylinder+boolean. The green emissive cylinder at the left has the expected diameter: 0.00925 (…but no GI)

I used Angular Size=1/4 (0.25) for all 3 checks. So it seems this is the ANGULAR RADIUS of the light source.

These are empiric checks, as much as it can be in a 3D software.

Didnt mention this in my atmospherics tutorial in 3D Artist mag #116 issue…

If I get few thumbsups/likes, I'll share other fancy scientific stuff like this

fair enough, I re-checked my analytic geometry setup and a closer look at the penumbra confirms your view. Thx!