Task GAUSS: sigma parameter

Acrobat wrote on Mar 08, 2007

Hi all,

I have to use the task GAUSS to perform a convolution. My question is: how can I pass from my instrumental resolution (expressed as lambda/Dlambda or as velocity) to the parameter sigma for the gaussian, in pixels?

Many thanks in advance.

Cheers,
Elisa

---

Mike Fitzpatrick wrote on Mar 08, 2007

Elisa,

If I understand your question: The sigma parameter must be in units of pixels, so if you know only the velocity you'll need to convert that to pixels based on the resolution of the detector. If you're trying to do something like rotational broadening of spectra, do a forum search from the banner bar for 'rotbroad' and you'll find a script that can help. (Note you may need to modify this a bit to get it working). Otherwise, what kind of instrument is this? An image or spectrum? Why are you doing the convolution at all?

-Mike

---

Acrobat wrote on Mar 08, 2007

Hi Mike,

Thanks for your answer.

Yes, I had done a forum search, finding that the suggested tools were rotbroad and also dispcor. I need to convolve a very high resolution solar spectrum in order to match my instrument resolution and be able to use it in the reduction of my spectra.

Dispcor doesn't help, and, instead of rotbroad, which needs to be modified, I prefer to use a standard tool, as gauss, as it is the most common tool in literature, used for this purpose.

I think I already have all the ingredients: the resolution of the solar spectrum, the resolution of my spectra, gauss task; I just can't see the correct way to convert the resolution values to pixels.

Thanks again for your time.

Cheers,
Elisa

---

Last post on Mar 08, 2007