Comparison of `gendaylit` multi-channel sky with `genssky`

Dear all,

I need to compare the 9-channel spectral sky that Lark uses and a spectral 18-channel sky from Radiance v6.0.

I need help understanding how I can compare these two skies. How can the sky and sun in Radiance v6.0 be separated, like in Lark? Is that even possible? Or the other way around: is it possible to make a visualisation of sky-sun-ground in Lark/gendaylit, similar to out_sky.hsr from Radiance v6.0 — for example, for pixel-by-pixel comparison?

Also, I notice a difference in the directional emitter between the two sky definitions for the same timestep, and I cannot explain it.

Can you help me with this? I’ve added example data for the same time step from both tools below.

Any suggestions are very welcome — many thanks!


Example data:

With Lark/gendaylit, I can access the 9-channel radiance using a light modifier for the sun and a glow sky material. An example for a single timestep is as follows:

Sun and sky SPDs are read separately by Lark:

Which is simplified to:

With Radiance v6.0, there are two key parts: a light source with spectrum modifier (sunrad), and out_sky.hsr.

Directional emitters:

  • Lark:
solar source sun
0
0
4 0.369629 -0.833813 0.410038 0.533000 
  • Radiance:
solar source sun
0
0
4 0.333397 -0.796601 0.504256 0.533

Hi Nima,

I think the simplest way is to directly sample the skies.

echo 0 0 0 0 0 1 | rtrace -co+ -cs 20 '!oconv sky.rad'

You can prepare a set of sample rays to be used across different skies.