[sudo-discuss] Spectral Response Curves

(Also posted to Noisebridge. Skip if you are experiencing deja-vu.) Hello, Right now, I'm doing research for a specialized app to let consumers validate products that can be used to treat insomnia. Specifically, I'm working on a smartphone app that can act like a light detector for light in a narrow band around 470 nanometers. This is the frequency that has been shown to suppress melatonin secretion by the pineal gland. There are specialty products for avoiding late night 470 nanometer light exposure (light bulbs, screen overlays, glasses) but these are often very overpriced, and there is no convenient way to validate them. There are also "ordinary" products that serve the same purpose, but there is no good way for people to accurately test them. Currently, people can use a CD or DVD as an ad-hoc diffraction grating and look at the resulting spectrum, but even this isn't quite good enough. I've bought a narrow bandpass filter for 470nm light, and even products that have a spectrum that "looks good" (very little blue) can have hot-spots that leak large amounts of 470nm light, and it doesn't take much to suppress melatonin. (As low as 0.5 lux for prolonged exposures.) Looking generally at spectral response curves for digital cameras, it should be possible to "synthesize" a narrow band detector by taking the blue channel response and subtracting the red channel value. http://www.maxmax.com/images/Cameras/Technical/NikonD200_SpectralResponse.j… This should effectively produce a "synthetic" instrument that has a spectral response curve that would be the camera's blue response, minus its red response. Even more helpful, the user should be able to view a synthesized narrowband image of the product, to be able to spot leaks and hot spots. What I need: I would like help in scientifically measuring the spectral response curve of the "synthetic instrument." I already have a (tiny) 470nm filter from Thorlabs, and I'm already aware of Public Lab's DIY spectroscope. I would like to use more accurate equipment, however. It is important that I can provide accurate information about the performance of the app and use rigorous procedures for measurement so that users know they can rely on the instrument. Does anyone have the expertise and access to equipment to help me out? --SCZ -- There's neither heaven nor hell, save what we grant ourselves. There's neither fairness nor justice, save what we grant each other.