Data Reduction

Data Reduction for the GTN


Reducing CCD images taken with telescopes like those in the GTN is a somewhat involved process. There are several steps that must be completed to convert your raw images into scientifically meaningful data. Even if your primary goal is to take “pretty pictures” several of these steps are important. This page is designed to give you a rough guide for reducing your images and to point to resources that you will need or want to have for that purpose. For a more thorough discussion of the topics on this page see the AAVSO CCD Observing Manual. Another excellent resource for all aspects of astronomical image processing is The Handbook of Astronomical Image Processing by Richard Berry and James Burnell. To learn how to actually perform the steps discussed below you will have to consult the documentation for whatever software you are using for the reductions.

STEP 1 – Image Calibration

Image calibration refers to corrections that must be made to CCD data before you can extract scientific measurements from them. If you do not know what flat fields, darks and biases are, see the page on Image Calibrations before you begin this overview. For the rest of this page we assume that you have a set of calibration images and science images. We will describe how you should apply the calibrations to your science images. In certain instances, the calibration images might have been made for you. In any case, you will have to get a set of calibration images in order to reduce your data.

A Note About Image Calibration

During each of the steps outlined in our description of image calibration you should check the resulting images and make sure that the output is reasonable. A bad calibration correction will give you bad science images at the end, so you want to be sure that at every step you are not introducing any errors. Visually inspecting your images is one way to check that things are not going horribly wrong. You can also do hand-calculations with a few images to check pixel values. You definitely want to find any discrepancies now, before you havve spent hours working on science images that will have to be redone because the calibration was done incorrectly. Keeping these things in mind, click on Image Calibration to read a general overview of the steps required to calibrate your CCD data. Note that this gives only an overview of what is to be done; though the steps will be identical, the particulars of how to execute them will vary depending on the software package you use.

STEP 2 – Image Stacking

After you have calibrated your science images you must stack them. This operation is much like combining your flats, darks and biases, but there is one important difference: Your science images contain objects, and before you combine them you have to align the objects so that they add together. Aligning your images is called registering them. If your images are not registered before you combine them you will end up with multiple instances of each object, one from each of your raw images, in your final image. Of course, if your tracking is very good or if you have an auto guider then your images might already be well enough aligned (you should check to make sure). In any event, other than this very important difference, you merge science images for the same reason you merge calibration images: it helps remove cosmic rays and it improves the statistics in the final image.

The details of how to align and merge images will vary for different software packages. Some make it very easy by analyzing the field and calculating the necessary offsets for each image automatically. They then register the images and add them together. Other packages require you to determine the positions of a subset of stars from which you calculate offsets and apply shifts by hand. You should look at the documentation for your own software to find the details on how to register and combine images.

STEP 3 – Doing the Reductions

Once you have calibrated and stacked your images you are ready to begin the real reductions… and about time! If you wish to determine the brightness of objects in your field, look at the photometry section. If you want to do astrometry, go to that section – you actually might not have to bother about calibration if all you want is positions of things, though for the most accurate positions possible calibration will still be required.


Photometry is the measurement of the brightness of objects in your images. There are two types of photometry, relative and absolute. For most GTN projects you will probably want to do relative photometry, in which you compare your program objects to other objects in the image. This is also often called differential photometry. If you want to do absolute photometry instead, then you must do additional observations of photometric standards in order to put your comparison objects on an absolute brightness scale. For a description of how to do either kind of photometry you should consult a reference like the AAVSO CCD Observing Manual. To gain a basic understanding of how photometry works, have a look at our Basic photometry and Astrometry page. Again, the particular steps you must complete to do your photometry will depend on the software package you are using. Consult its manual for details.


Astrometry is the measurement of positions of objects in your images. This is always done relative to some set of reference objects within an image. In the event that you have absolute positions (RA, DEC) of these reference objects, you will be able to put all the objects in your field on this same absolute system. To do this you need a reference catalog of stars to determine the absolute position of your field. Read our astrometry page to learn how to do this.

MaximDL Photometry Exercise

We have put together a step by step tutorial that takes you through the process of reducing a set of images using MaximDL. This is one of the popular packages used on Windows. A list of other packages can be found in the table below.

Start the Photometry Exercise

Software Packages

You have many choices when deciding what software package to use for your data reductions. When choosing a package you should
consider the computer platform you will be using, your comfort level with computing and coding and how much money you want to spend,
among other things. There are currently on the market an array of packages specifically designed for astronomical data reductions. Some
of them have elaborate and well thought out user interfaces and predefined functions to make your reductions nearly automatic. Other
packages are not much more than high-level programming languages that will require you to write a lot of your own routines (though it is
often possible to download routines that other people have written off the World Wide Web). We list a few of these below. Click on the
name of each package to get a short synopsis of what we see as their strengths and weaknesses. The synopsis will show up in a popup
window, so be sure that your browser will allow popups, at least while you are viewing this page. The link at the right will take you to
the download site for the software.

If you know of additional good packages that we should include in this list, please send us an email (to one of the contacts at the
bottom of the page).

Aperture Photomety Tool is a Java tool for doing simple aperture photometry. It is maintained at NASA/IPAC. Since it is written in
Java it will run on any platform. The GUI is very simple and easy to use and understand. The package also comes free of charge.

AIP4Win is a comprehensive package that comes as part of an excellent book on image processing: The Handbook
of Astronomical Image Processing
. This is highly recommended for Windows users, but it will not run natively on other platforms.

Equinox is a Mac OSX package that comes in several versions. One part is a planetarium package that can be used to
plan observations. The other is a CCD controller/imaging package that can be used for data reductions. They can be purchased
separately, or bundled together. We have not run this, so we cannot comment on how it works.

CCDSoft is the camera controller/image software package from Software Bisque. This is a full-feature package that
will do everything you need for photometry and astrometry.

Deep Sky Stacker is a free package that can be used to register and combine images.

ImageJ is a java version of the NIH Image package. It was developed to view PET scans, MRI and the like, but
it works very well for astronomical purposes. Of course, it does not contain native astronomy related tasks, but many of these have
been written and are available on the web. Since it is JAVA it runs on all platforms.

MaximDL is another full-featured package for Windows. It can both control your CCD camera and do photometric and
astrometric manipulations, pretty much all you need.

Mira is a powerful image processing package for Windows. It will do all you need to do.

Interactive Data Language (IDL) is a high-level programming language for image and array manipulation. It is used by
many professional astronomers, and libraries exist for most standard tasks. It runs on all platforms but is quite expensive.

Image Reduction and Analysis Facility (IRAF) was developed by and for professional astronomers. It will do everything
you need to do, and then some. However, it is not the easiest package to learn. Will run on any unix type OS and is available free of

Perl Data Language (PDL) is an open-source answer to IDL. It is based on the PERL and C programming languages and
claims to work with arrays as easily as IDL. We have not used PDL, but if you are the kind of person who likes to write your own code
you might want to try it. Currently under development, follow link in right column.


Astronomical Software Packages
Package Platform(s) Download
Aperture photometry Tool (APT)
Mac OSX, *-nix, Windows Aperture photometry Tool (NASA/IPAC)
SalsaJ (Simplified ImageJ)

Mac OSX, *-nix, Windows Hands-On Universe
ImageJ (Java NIH Image)

Mac OSX, *-nix, Windows NIH
Mac OSX Microprojects
Windows Berry and Burnell
Windows Software Bisque
Deep Sky Stacker
Windows Deep Sky Stacker
Windows Cyanogen
Windows MiraMetrics
Mac OSX, *-nix, Windows ITT
Mac OSX, *-nix NOAO
Mac OSX, *-nix, Windows Perl Data
Cost Key for Left Column: $ – least expensive, $$ – more expensive, $$$ – most
expensive (Check for educational discounts from all these vendors.)

000-017   000-080   000-089   000-104   000-105   000-106   070-461   100-101   100-105  , 100-105  , 101   101-400   102-400   1V0-601   1Y0-201   1Z0-051   1Z0-060   1Z0-061   1Z0-144   1z0-434   1Z0-803   1Z0-804   1z0-808   200-101   200-120   200-125  , 200-125  , 200-310   200-355   210-060   210-065   210-260   220-801   220-802   220-901   220-902   2V0-620   2V0-621   2V0-621D   300-070   300-075   300-101   300-115   300-135   3002   300-206   300-208   300-209   300-320   350-001   350-018   350-029   350-030   350-050   350-060   350-080   352-001   400-051   400-101   400-201   500-260   640-692   640-911   640-916   642-732   642-999   700-501   70-177   70-178   70-243   70-246   70-270   70-346   70-347   70-410   70-411   70-412   70-413   70-417   70-461   70-462   70-463   70-480   70-483   70-486   70-487   70-488   70-532   70-533   70-534   70-980   74-678   810-403   9A0-385   9L0-012   9L0-066   ADM-201   AWS-SYSOPS   C_TFIN52_66   c2010-652   c2010-657   CAP   CAS-002   CCA-500   CISM   CISSP   CRISC   EX200   EX300   HP0-S42   ICBB   ICGB   ITILFND   JK0-022   JN0-102   JN0-360   LX0-103   LX0-104   M70-101   MB2-704   MB2-707   MB5-705   MB6-703   N10-006   NS0-157   NSE4   OG0-091   OG0-093   PEGACPBA71V1   PMP   PR000041   SSCP   SY0-401   VCP550   100-101   70-412   VCP550   LX0-104   000-017   70-461   700-501   70-347   EX200   SSCP   200-120   VCP550   NS0-157   400-201   N10-006   000-080   000-080   70-346   3002   300-101   PMP   640-692   ITILFND   ICBB   70-246   9L0-066