Digital response curves.

Hi,

You know with films you can just go to the manufacturer website and download the documentation listing the response curves of the film?

Anyone happen to know where you can get similar data for canon 300D/350D? A trawl of the canon website didnt turn up much.

The reason I'm asking is that I've started messing about with spectra. Here is Vegas'; blue is mine; purple is a reference image of a A0 type star.



I know the camera is only sensitive around 400-700 nm, and particularly sensitive to green; mimicing the human eye as much as possible. However it would be nice to know the magnitudes of these sensitivities so I can compensate for them in spectra.

How do dedicated astronomical CCD's handle this?

Cheers,
~Al

Interesting stuff Al. Sorry cant help with your question. Can you tell us how you went about acquiring and processing that spectra? What did you use to break up the starlight, processing etc?

Can you tell us how you went about acquiring and processing that spectra? What did you use to break up the starlight, processing etc?

I used Vspec www.astrosurf.com/vdesnoux/ to do the processing. Basically you take the specta from the camera, change to greyscale, save as a jpeg, open that in Iris, save as a pic file then open that in vspec. A bit convoluted but it works. Perhaps there is some routine in the software to handle normalising the response of the camera across differnt wavelengths by comparing with a defined spectra of say Vega, I must read up more on it.

I've been using several different methods to try to disperse the light. For the project I am meant to use a couple of lenses, a prism and grating or a slit, so you would have a slit at the telescope then a lens to make the light paralell for passing through the dispersion device, and another lens to focus the light on the film/ccd. It is apparently possible to do this with an old pair of binoculars, by using the objective lenses as the lenses and the prism as the dispersion device. However I have had little success in getting this to work right as I end up with a lot of astigmatism in the resultant spectra which makes it useless. That happens whether I use a prism or diffraction grating (700 lines/mm; the kind I have had at ILPAC stands in the past).

What I used for the above image was a grating of 200 lines/mm placed between the scope and the camera with no collimation lenses or slit. This causes its own problems, by the light not being paralell going through the grating, but the issues it causes are less than the astigmatism seen with denser gratings. It also has the benefit of being able to have the 0th and 1st order images on the same image, which helps in the calibration, and takes the guesswork out of spectral lines. Hopefully the project supervisor will appreciate these benefits and not view it as me giving up on doing the project the proper way! - This way is low resolution but I think it will work well enough for classifying stars into spectral types.

I tried this last night, but the seeing conditions were terrible, the following spectra were taken through cirrus cloud:
At least you can see that the two A stars look the same and very different from the K star Eltanin.

Cheers,
~Al

Al - to be honest, I've no idea if this link will help. Its a very technical website about CCD'd in general. Check out the technical notes on the leftnav bar.

www.princetoninstruments.com/library_encyclopedia.shtml

Cheers and good luck!

Dave McD

Al - to be honest, I've no idea if this link will help. Its a very technical website about CCD'd in general. Check out the technical notes on the leftnav bar.

www.princetoninstruments.com/library_encyclopedia.shtml

Thanks Dave, I'll have a look. Though I think the vspec software actually handles what I need to do : www.astrosurf.com/vdesnoux/tutorial5.html "Correct spectrum from the spectral non-uniformity of the detector".

Cheers,
~Al