Hi all,

I have a source of light described by a radiance/wavelength graph (spectral power distribution) and was wondering how to import it in Radiance.

Thank you in advance for your help!

Hi all,

I have a source of light described by a radiance/wavelength graph (spectral power distribution) and was wondering how to import it in Radiance.

Thank you in advance for your help!

This is quite a can of worms, and I’m not sure you want to open it. You can read some detailed advice in the following 2002 paper if you really want to know:

Ward, Greg, Elena Eydelberg-Vileshin,

“Picture Perfect RGB Rendering Using Spectral Prefiltering and Sharp Color Primaries,”

Thirteenth Eurographics Workshop on Rendering (2002),

P. Debevec and S. Gibson (Editors), June 2002.

Hi Greg,

Thanks for your reply. I would have some questions:

- So, in very general terms, one should proceed to those conversions (?):

Spectral power distribution—> xyz values --> rgb values.

and after that create a simple light using the formal definition :

void light myspectral_light

0

0

3 red green blue

entering the rgb values there. Would that be right?

Would section 8.2.5 Converting Spectral Data into Light at pag 435 be useful for this purpose?

- Would a polar diagram of intensities give more reliable results in terms of simulation of light source compared to spectral power distribution?

Thank you for your help!

Your procedure is correct for the crude color approximation method. The spectral power distribution tells you only the color of the light, which is of extremely limited use. The polar plot tells you the spatial distribution, which is far more important. The two do not contain the same information at all.

Yes I know that polar diagrams and spectral distribution relate to different scopes, but was wondering how comparable could be results related to the same source (or two very similar sources) simulated through those two procedures. I assume that a light source generated from spectral values would not offer physically accurate results that the spatial distribution would give, but would it be completely useless?

Unless your needs for exact color are critical, I would not bother with the light source spectrum. And if they are, then I would use one of the more advanced techniques described in the aforementioned article.

I think you could do both if you thought both were important. Couldn’t you use the polar diagrams to model the intensity by angle, and then apply a color to the light source also based on the spectrum if that was important?

Yes, of course Chris is correct, you can apply both.

They are not representing perfectly identical sources but it is surely worthy, thank you both for your precious help.

I will try and see how it goes.