On Tue, Nov 12, 2013 at 12:33 PM, Patrik D'haeseleer <patrikd@gmail.com> wrote:

The figure I linked to earlier suggests the 50% drop-off should be around 420-520nm, which covers pretty much all of the blue range of the spectrum.

This is another reason I'd like to play around with subtracting the red signal. Depending on what the spectral response curves are, it might be possible to play around with multipliers and get close to the shape of the graph. (Granted, that it might also turn out to just including the blue signal, but I really need to *know* what that is, not make an educated guess about it.)

The photodiode you listed seems pretty much a perfect match. No need to muck around with narrow band-pass filters.

For the purpose of the intended distribution context and audience, I should **avoid** requiring other to be mucking around with photodiodes. The fact that a photodiode detector's curve matches another really doesn't change that.