Down in Chalmette, Louisiana, folks have been using DIY spectrometers to try to detect light from refinery flares, and wanted to help refine the setup used to detect spectra of far-away sources like the flare they're looking at. Both Scott Eustis from Gulf Restoration Network and folks from NASA's DEVELOP program are involved. You can read some more about the recent work done on this technique by looking for notes tagged "flare".

I did a few things to my spectrometer, and there were some advantages and some disadvantages:

First, I put a bit of clear, but slightly whitish plastic bag over the front. This reduces the light entering the spectrometer, but it also means that you don't have to point it nearly as precisely -- it's much easier to align with a telescope or tube. And since the flares are really bright (read Scott's post, linked to above), I didn't think it'd be a problem.

Second, I added a long tube, as suggested on the plots-spectrometry discussion list, which allows me to point it at just one light source. You can see this above, as I'm testing it by pointing at a lamp. I believe that this only really works because it doesn't have to be aligned perfectly with the spectrometer -- because of the diffuser (plastic film) we added in step 1.

Third, I stabilized the camera by putting a foam block and a knot inside the chamber to ensure that no tugging (which could easily happen in the field) moves the camera. This should be resolved in the new kits we'll be releasing soon, but all old kits ought to have such protection. I know Dan Beavers puts a zip tie around the base of his kit to restrain the cable.

I believe that this setup would work just as well with a telescope or binoculars, and probably better since it'd concentrate more light on the diffuser, and still wouldn't have to be lined up with the spectrometer -- just with the target.