Cor, Birger,
I think it will take a very long time to get an exposure the way you
suggest. So I offer the following.
Many years ago, when working as a research physicist, I had access to an
incredible assortment of tools. One was a scanning spectrophotometer.
Perhaps someone out there has use of one. Here is what I suggest.
Calibrate the instrument carefully. (It is very difficult to get good
absolute accuracy when measuring light intensity and spectrums. But, it
is easy to calibrate a carefull relative comparison.) Scan the spectral
output of a light boxe (different lamps may be tried) used to expose
prints. The sun could be used, if atmosferic conditions don't cause too
much fluctuation of light intensity. Also, get a hold of an assortment
of blue end cut-off filters such as Kodak Wratten HF-3, 2E, and 4. Scan
with each filter. The result will be four scans for each light source
of increasing blue-end (UV) light. Even though there is a cumulative
effect of adding more light, the added light will be in blocks of
increasingly higher frequency photons. I believe that this will be
enough to demonstrate the effect of various bands of blue/UV light.
(The data provides the wavelength distribution and flux for each
exposure.)
Of course, prints will be exposed in each of the situations. Exposure
times should be reasonable for these light intensities. If any
differences are noted in the prints, then the information of what
spectrum they were exposed to will be at hand. Also, one should notice
the spectral sensitivity of the material. More filters may be
neccessary to refine this. As more of the blue end of the spectrum is
added the exposure should increase, until the point is reached at which
the wavelength is too short. Even though there is a cumulative effect,
A frequency dependent curve should be capable of extrapolation from the
data.
Does the look of a print varry with exposure to differing wavelengths?
I don't know. Will this information be valuable? I don't know. I do
know that this will provide reasonable exposures and give the frequency
dependance of the process materials. If someone has the time to do
this, please report the results.
Jeff