A fantastic new tool is available to help non-professional (and professional) astronomers do low-resolution spectroscopy. There are now gratings that you can drop in your filter wheel ( the Star Analyzer SA-200 for example) and use to take spectra of objects without the imaging setup for your telescope. This page is setup to provide a few free resources for people that want to start using this interesting innovation.
The advice we give here has been developed through experience with using a Star Analyzer SA-200 in the filter wheel for the imaging setup on the Barber Observatory 20-inch Telescope. We have no financial or other arrangement with the builders of this equipment so the advice we give here is from that perspective.
I have been pretty impressed with the demos of the latest versions of the RSPEC software they sell with the grating. But sorry Mac and Linux users.... it is Microsoft only. For this reason and also for the purpose of educating people about how RSPEC works, I have set up the following tutorial.
In this tutorial I determine the wavelength scale in nanometers (nm). This is purely a personal preference. Angstroms are also a popular unit. Multiply nanometers by 10 to obtain Angstroms.
Another personal preference expressed in this tutorial is to graph spectra so that longer wavelengths appear on the right of the X-axis. This is known as the "red is right" approach because longer wavelengths are red and they are on the right. The double meaning of the word "right" should not be taken literally here. Just as many professional astronomers plot long wavelengths to the left on the X-axis (because of the relation between frequency and wavelength for light). So "red is right" is just one way to do it, NOT the only correct way to do it.
This tutorial is in two parts:
How to extract spectra from images taken with your filter wheel grating.
How to wavelength calibrate those spectra.
Before you start this tutorial you will need the following:
Images taken with your grating that have been dark subtracted in FITS format. No flat fielding is necessary.
A copy of SAOImage DS9. It is free and can be downloaded here (http://ds9.si.edu/).
Watch the video and/or follow the outline below.
Steps to Extracting Your Spectrum
Load the dark corrected FITS image into DS9.
Activate the projection tool through the "Region" menu in the toolbar on the top of the DS9 window (Region -> Shape -> Projection).
Use the arrow to draw a projection for a bright star with an obvious spectrum in the image.
Adjust the end points by using the tabs on either end of the projection.
Adjust the width extracted using the tab in the middle of the projection.
Grab (at a point other than the tabs) the projection and move it to the intended target.
In the graph window for the project, select "Save" from the "File" menu to save the data from your projection
You now have a text file of ordered X,Y pairs that you can graph. The X coordinate is pixels from the start of the projection and the Y coordinate is the brightness of the projection at that pixel.
For the second part of this you need the following to begin:
A dark correct grating image of a planetary nebula, quasar, or another point-like object with strong emission lines dominating its spectrum.
Planetary nebula with smaller angular sizes are preferred. (see here)
Here is a list of the brightest quasars.
A list of wavelengths for the strong emission lines in the object you took a grating spectrum of.
Here is a list of emission lines in planetary nebula but this graphic I use in the video is probably more useful for identifying the brightest lines.
Note that if you use a quasar you need to correct the wavelengths of the lines for cosmological red shift.
Steps to wavelength calibration
Use the DS9 projection tool to extract a spectrum of the planetary nebula.
Find the pixel in the projection where the zero order image is centered. This is your zero_point.
Identify the bright features in the spectrum and record the pixels they are centered on.
Divide the known wavelength of the feature by the distance in pixels between the zero order image (zero_point) and the feature in the projection (pixel).
That ratio (nm/pixel) can now be used to solve any spectrum you extract from any image with the projection tool for the wavelength using the formula: (wavelength) = (nm/pixel) * [(pixel) - (zero_point)].
Every time you extract a new spectrum you will re-determine the position of the zero order image (zero_point) but the ratio (nm/pixel) will remain the same as long as you are using the same grating and do not alter the distance between your filter wheel and your CCD camera.
Last Modified: July 25, 2014