Hi all -- it's so great to see so much activity on the list on spectrometry. There didn't used to be as many spectrometry folks on here -- I guess we can thank Kickstarter.
Re: blu-ray -- i'm very glad to see some MATH to back up what we basically determined empirically by just testing lots of different discs :-P. I'm not as good at the theory, and like to just try lots of things out -- but it's so great that we now know *why*...
I think we're going to have to do our own radiometric calibration using a known light source-- maybe a 'black body' source, i.e. just a very hot glowing object. Or, a light bulb we mail around :-). Those have very easy to model spectra, and it seems like we should be able to figure out what our cameras are seeing.
If data is very reliable between cameras of the same model (in the Kickstarter we'll be buying a batch of thousands) then we'll only really have to do this once per model. Or if light bulbs are very consistent, we can all just use one light bulb at a given distance or something. That would be nice -- keep in mind that ALL data on the SpectralWorkbench website is currently not radiometrically calibrated-- so if a given webcam just sees reds more brightly, we don't really know... yet!
that's for a "back-illuminated CMOS sensor" -- the kind in the iPhone 4 and many webcams. I think this means that if we use that type of sensor, and remove the glass lenses (maybe use a pinhole) -- we could get quite a bit of UV. Maybe even down to 200nm! But it's just an idea... we'll have to try it.
Finally -- as some of these ideas become more concrete, we should try to move them into research notes on the website because then we're creating more permanent documentation. It's hard for people to dig through the mail archive to find this stuff!
Jeff