Reflectance and Spectral Content

[davis.jacobj]

I have two questions for anyone who might know some light/color science...

1.) How can reflectance spectrum be calculated if both the light information and material information is known? Information would be intensity vs. wavelength.
2.) Having a photograph, can an overall spectral content be generated? Again, intensity vs. wavelength. I guess it could be almost the same as the first question, except that the material spectrum would need to be measured/calculated first.

I've spent most of the day trying to find this information here and on Google and haven't come up with anything.
In advance, I appreciate any advice, suggestions or criticisms anyone has.

Jake

 

[tirediron]

Way over my head; I'm guessing if anyone can help you, it's mostly likely Helen B.

 

[KmH]

That's well beyond Physics 101 and the math would also be fairly complex, like Fourier Transform analysis in 3 dimensions.

I doubt you will find any canned Internet web page that approache those subjects, but their might be some Intranet pages that do. Unfortunately Intranet pages are not routinely acessable from the internet.

Depending on the class of materials you want to explore, you may well have to make your own reflectance measurments before you can begin.

 

[Helen B]

1) If you have the radiant or luminous intensity vs wavelength information for the light source, and reflectivity vs wavelength for the material you can easily calculate the relative radiance or luminance across the spectrum. Do it in, say, 10 nm bands and multiply the intensity of the source in that band by the reflectivity in that band. If you want absolute values for the radiance or luminance in each band then the calculation is more complicated.

2) I'm not quite sure what you are asking. You can't derive spectral information from a trichrome photo, because different spectral compositions can result in the same RGB values (similar to metamerism in vision) so there are an infinite number of possible answers. If you knew the reflectivity vs wavelength properties of the material you still couldn't derive a single answer if the nature of the light was not also known. You could get a result if you made assumptions about the light, or knew its spectral distribution.

I hope that I have understood what you are asking, but I may not have done.

Best,
Helen

 

[davis.jacobj]

1) If you have the radiant or luminous intensity vs wavelength information for the light source, and reflectivity vs wavelength for the material you can easily calculate the relative radiance or luminance across the spectrum. Do it in, say, 10 nm bands and multiply the intensity of the source in that band by the reflectivity in that band. If you want absolute values for the radiance or luminance in each band then the calculation is more complicated.

2) I'm not quite sure what you are asking. You can't derive spectral information from a trichrome photo, because different spectral compositions can result in the same RGB values (similar to metamerism in vision) so there are an infinite number of possible answers. If you knew the reflectivity vs wavelength properties of the material you still couldn't derive a single answer if the nature of the light was not also known. You could get a result if you made assumptions about the light, or knew its spectral distribution.

I hope that I have understood what you are asking, but I may not have done.

Best,
Helen

1.) My instinct was to do just that but I wasn't sure. I don't need to know this but just for fun, what other factors go into finding absolute values? I'd imagine you'd want to divide reflectance by its max value to bring its highest value to 1 and then multiply by some kind of overall reflectance coefficient for the material.

2.) After some thought, I guess I was really asking: Does a photograph have an overall spectral response? Assume you shined a light with equal energy in each wavelength on it to get its "true" reflectance. What would you actually be measuring, if anything useful?

Thanks for everyone's response!
Jake

 

[jake337]

You could try posting your question here as well.

Imaging Technology: Theory, Alternatives, Practice and Advances. - Open Photography Forums

 

[davis.jacobj]

You could try posting your question here as well.

Imaging Technology: Theory, Alternatives, Practice and Advances. - Open Photography Forums

Thanks for the suggestion! Looks like a great forum.

 

[Garbz]

2.) After some thought, I guess I was really asking: Does a photograph have an overall spectral response? Assume you shined a light with equal energy in each wavelength on it to get its "true" reflectance. What would you actually be measuring, if anything useful?

It does. Assuming perfectly linear response of the sensor (crazy assumption ) what you are measuring is the integral of the union between the incoming light and frequency response of each of the filters on the sensor. Red, Green, and Blue. But the fun does not stop there. You'd need to take into account and characterize the entire system to derive even remotely useful information out of it. Even then you're still only getting a rough approximation given you're looking at an entire spectrum and guessing what's there based on a red green and blue sample. You can determine how it's presented (which is what the camera does) but never how the spectrum looked initially for reasons Helen said.