Question about using ecidentify
reynolds wrote on Jul 04, 2007
When using ecidentify for wavelength calibration on echelle spectra, the plot starts out with the scale in units of pixels. However, after I identify lines in a few orders and do the first fit, the scale changes to wavelength. When I go back to identify additional lines I want the units in pixels again but can't change it from wavelength. How can I change the plot units back to pixels after doing the first fit?
Thanks.
Thanks.
Francisco Valdes wrote on Jul 04, 2007
Hi,
I'm afraid you can't. Normally one wants a wavelength scale since line lists are in dispersion units and so it makes it easier to find lines once you have an approximate scale. Out of curiousity, what is your reason for wanting pixels?
Yours,
Frank Valdes
I'm afraid you can't. Normally one wants a wavelength scale since line lists are in dispersion units and so it makes it easier to find lines once you have an approximate scale. Out of curiousity, what is your reason for wanting pixels?
Yours,
Frank Valdes
reynolds wrote on Jul 04, 2007
Frank,
Thanks for the info. I'm glad you're in this forum because I always get referred to you and you've answered several questions for me in the past. I've been struggling for a while (due to other commitments) getting a good wavelength solution for spectra obtained with a fiber-fed echelle in which a single 2k x 4k CCD is used to cover the entire wavelength range from ~4000-10,000 A. Because of the wide coverage the ends of some of the red orders fall off the chip leading to gaps in the spectrum. This I think causes the IRAF solution to be inaccurate in the red even though I identify as many of the bright Argon lines as possible. Since the instrument was under development when I took over I don't have a 'line map' for it. I did develop a spreadsheet tool that I use to identify the position of lines in pixel coordinates (this is how I identify lines when I first start ecidentify) and I thought if I had the plot units back in pixels I would have a better chance of identifying lines in widely separated orders where some of the wavelength identifications appear to be way off. What do you think of this approach? I've never done a wavelength solution for a new instrument like this before; can you tell me how others have done it in the past?
Thanks in advance for your help.
Thanks for the info. I'm glad you're in this forum because I always get referred to you and you've answered several questions for me in the past. I've been struggling for a while (due to other commitments) getting a good wavelength solution for spectra obtained with a fiber-fed echelle in which a single 2k x 4k CCD is used to cover the entire wavelength range from ~4000-10,000 A. Because of the wide coverage the ends of some of the red orders fall off the chip leading to gaps in the spectrum. This I think causes the IRAF solution to be inaccurate in the red even though I identify as many of the bright Argon lines as possible. Since the instrument was under development when I took over I don't have a 'line map' for it. I did develop a spreadsheet tool that I use to identify the position of lines in pixel coordinates (this is how I identify lines when I first start ecidentify) and I thought if I had the plot units back in pixels I would have a better chance of identifying lines in widely separated orders where some of the wavelength identifications appear to be way off. What do you think of this approach? I've never done a wavelength solution for a new instrument like this before; can you tell me how others have done it in the past?
Thanks in advance for your help.
Last post on Jul 04, 2007