# Spectral deconvolution

1. Oct 29, 2007

### Deckard

Hi,

I am working on long slit spectroscopy for an astronomical instrument, and I think I need deconvolution, but I am not sure, so maybe you will be able to answer me :) Currently those are only simulations. I work in near-infrared (0.950 -> 2.2 micron) with a low resolution (R=30).

My problem is that the spectrum I am supposed to observe can be very steep (ie I can have a quick flux variation between $$\lambda$$ and $$\lambda+\delta\lambda$$), but when I watch them through my instrument, this spectrum is convolved with the slit LSF (line spread function), which is a sinc function ($$sinc(x)=sin(\pi x)/\pi x$$). Of course the width of the LSF varies acording to wavelength, and then is much larger at 0.95 micron than at 2.2 micron. And when I observe a steep change in the spectrum through the spectrograph, the change is much less steeper because it has been convolved with the LSF wich is less steeper.

Are there some deconvolution methods that can correct, at least partialy what I describe?

I hpe what I say is making sense. If it's not I can post some picture to illustrate this. Thanks a lot in advance.

-- Arthur;

2. Nov 4, 2007

### leptonsoliton

Deconvolution. With applications in spectroscopy, PA Jansson - New York: Academic Press, 1984, edited by Jansson, Peter A., 1984

.... and thousands of other references. Use Google.