Use Radiance to predict solar irradiance , not illuminace

1

Hello

I have a very general question: I am trying to use 2-phase method to predict solar irradiance on indoor objects indoor. However, I think radiance is better at predicting illuminance, not irradiance. I said this because I think Radiance only accept the transmissive value for RGB (aka Tvis), not the entire solar spectrum.

I am wondering if there is any workaround if I need the result to be presented in solar irradiance?

Thanks,

Shen

2

The RGB values can mean whatever you want them to mean; you just need to be consistent. If you want results in radiometric units, your light sources, reflectances and transmittance values must all be in radiometric units. For example, you could use the red channel to represent total solar flux (radiometric), the green channel to be photometric flux, and the blue channel to relate to circadian sensitivity. You need to normalize the channels with factors such that equal-energy white light translates to an RGB value of (1,1,1), but then you should be OK. The 179 lumens/watt conversion is the standard conversion factor, but you will need to derive others.

3

Thanks @Greg_Ward for the prompt response!

While I sort of understand what you meant in your response, I still have a few remaining questions that I hope to get answer for:

  • When I use genskymtx, the resultant value would be expressed in the radiometric unit (i.e. watt), not in luminance?
  • So is it correct that if my end goal is determining solar flux, then I can just specify Tsol (as opposed to Tvis) in .mat file, and by the end of the day I don’t even need the 179 lumens/watt conversion since the calculation workflow was all handled in radiometric unit?

Here are the files that I was using. I also showed these file in a previous post:

https://drive.google.com/drive/folders/1rlytTAlTyT6WJJq7nYHJQAz4QyFk6hh7?usp=sharing

Sorry if I sound ignorant. I am trying to learn Radiance while using it.

Thanks,

Shen

4

I believe gendaymtx attempts to create RGB values (in radiometric units) adjusted so that only the visible spectrum is considered. To get the total solar flux, you need to specify the -O1 option, then modify the materials as you say.

5

Dear @Greg_Ward,

Thank you for your response. In terms of RGB values, what’s the relationship between the RGB value, (1,1,1) and the weighting factors when calculating illuminance (i.e. 179*(R .265 + G .67 + B*.065))?

Thanks,

Shen

6

The 179 lumens/watt value is the standard Radiance conversion factor for equal-energy white light over the visible spectrum. The three coefficients for R, G, and B correspond to the relative contributions to the Y (luminance) channel for each primary as they are defined in the CIE color space. An radiance value of (R,G,B)=(1,1,1) yields a luminance value of 179.

7

Dear @Greg_Ward,

Thanks. Probably the last question before I stop bothering you: what are these factors R*.265 + G* .67 + B*.065? Where do those three coefficients come from? Do you have the literature that i can reference?

Thanks,

Shen

8

As I said, they come from the Radiance CIE primary coordinates, which are defined in ray/src/color.h. These were taken from a very old paper by Gary Meyer, that predated the sRGB standard and so differs slightly from those primaries. You can read any book on computer graphics or color science to fill in the rest.

1 Like