On Sep 22, 2015, at 7:43 AM, Roger Breton <graxx@videotron.ca> wrote:
I must say I am impressed by the claimed 1 nm spectral resolution.
Arbitrary spectral resolution is no big deal; you just use narrower slits. But that cuts down on the transmission, either increasing the noise or requiring longer integration times or that sort of thing. In practice, 10nm is plenty "good enough" for graphic arts work, 5nm all you need, 3nm as much overkill as isn't ludicrous, and 1nm near pointless but makes for some impressive-looking charts and the like. Of course, for astrophotography, 1nm just gets you recognizable Fraunhofer lines and, as such, is barely adequate for introductory amateur educational purposes...I think serious astronomy these days is done with spectral resolutions of small fractions of an angstrom.
I don't know what it's worth but it breaks my heart that they stopped at 380nm in the UV range. Why couldn't they go down to 350-360?
I'm not sure that typical affordable commercial CMOS / CCD sensors are all that sensitive below 380 nm. IR cut filters are common in digital cameras, but I don't think UV cut filters are. I could be mistraken.
Keep in mind that, in metrology, cheap is not common.
Well...it's one of those "pick n of m" cases. I've got a spectroscope with a _very_ large viewing area that will trivially do 1 nm resolution with a ~$20 bill of materials, but it's large and unwieldily. You can stick a camera in the one end and use it for spectrometry, but you'll need to calibrate it against some other instrument (such as an i1) -- and you'll obviously be limited to the intersection of the two spectral ranges unless you want to get into the realm of educated guessing. Spectral resolution, though, is limited only by how small you can carve the slit plus the camera sensitivity. I've got a last few refinements I need to make before I'll write up full descriptions, plans, etc., etc., etc. b&