<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Strict//EN"> <html><head> <link rel="stylesheet" type="text/css" href="cmunipack.css"> <title>Reduction of CCD frames</title> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head><body><div class="main_head"><p class="head"><a href="index.html">C-Munipack 1.2</a> / <a href="node006.html">User's manual</a></p><h1>Reduction of CCD frames</h1></div><p>The reduction of CCD frames is a process, which transforms a set of images carried by means of a camera into a set of measurements of objects that are found in the viewfield. A result of the reduction process is used as an input for following stages - for example, in the eclipsing variable star observation, it is used to construct a light curve, a time dependency of variable star's brightness.</p><p>The reduction process, described in this chapter, consists of a sequence of several steps: selection of input files and conversion into working format, calibration, photometry and matching.</p><h2>Before you start</h2><p>Before you start the reduction of your own CCD frames, you may need to perform several pre-processing steps. Though it isn't neccessary, combining the several correction frames into so-called "master" ones is advisory, because it reduces the noise and makes the result more precise. The method of making master correction frames is described in separate chapters.</p><h2>Input files</h2><p>First of all, we have to tell the program which files we are going to work on. These files are called the input files. Their list is displayed in the table in the main application window. When the application is closed, the list of files are saved to the disk and it is restored back when the program is launched again.</p><p>Supposing that the table now consists of files from your previous task, let's get rid of them. Please, use <tt>Files</tt> -> <tt>Clear files</tt> to start a new task instead of just removing the files from the table. Besides the clearing the table of input files, this function resets all internal variables, too.</p><p>Now, we need to populate the table with the CCD frames we're going to reduce. There are two methods how to acheive that - adding a individual files or adding all files from a folder. Which way is the best for you depends on organization of your observations on the disk. I'd suggest you to make a folder for each year, a folder for each night in it, the a subfolder for a name of object or another viewfield identification and finally a subfolder named upon the color filter (if you use more of them). In this case, the "Add folder" method is more convenient.</p><p>Click on <tt>Files</tt> -> <tt>Add folder</tt> in the main menu. A new dialog appears. In the dialog, find a folder where the inputs files are stored in. Click on an entry in the <tt>Places</tt> pane to go to one of a preselected folders, double click in the middle pane enters the folder. The buttons in the upper part of the dialog shows your current position in the directory tree, you can use them to go to one of the parent folders. Enter the folder with the input files - you should see them in the middle pane. Then, click on the <tt>Add</tt> button to add files to the table of input files. The program shows the number of added files in the separate dialog. The <tt>Add folder</tt> is not closed automatically and allows a user to continue. Click on the <tt>OK</tt> button to close the dialog and return to the main window.</p><div class="mediaobject"><img border=0 src="addfolder.png" alt=""Add folder" dialog"/></div><p>If you want to reduce only a subset of files from a folder, click on <tt>Files</tt> -> <tt>Add folder</tt> in the main menu. A new dialog appears, similar to the previous one. In the dialog, find a folder where the inputs files are stored in. Click on an entry in the <tt>Places</tt> pane to go to one of a preselected folders, double click in the middle pane enters the folder. The buttons in the upper part of the dialog shows your current position in the directory tree, you can use them to go to one of the parent folders. In the middle pane, select the files using the <b>Ctrl</b> modifier to include and exclude a single file and the <b>Shift</b> modifier to include a range of files. Then, click on the <tt>Add</tt> button to add selected files to the table of input files. The program shows the number of added files in the separate dialog. The <tt>Add folder</tt> is not closed automatically and allows a user to continue. Click on the <tt>OK</tt> button to close the dialog and return to the main window.</p><div class="mediaobject"><img border=0 src="addfiles.png" alt=""Add files" dialog"/></div><p>Before the next steps, the program needs to make a copy of source files. This copy is made for a security reason - all further steps are applied to that copy and the original files are never modified. If you make an error during the calibration, you can always go back to that point, make a fresh copy of the original images and have another go. Click on <tt>Files</tt> -> <tt>Convert/fetch files</tt> in the main menu. Wait for the process to finish.</p><h2>Calibrating the images</h2><p>A raw CCD image consists of several components. By the calibration process, we get rid of those which affect the result of the photometry. In some literature, the calibration is depicted as the peeling of an onion. There are three major components which a raw frame consists of - the current made by incident light, current made thermal drift of electrons (so-called dark current) and constant bias level. In standard calibration scheme, the dark-frame correction subtracts the dark current and the also the bias. Because of the nature of the dark current, it is neccessary to use a correction frame of the same exposure duration as source files and it must be carried out on the same CCD temperature, too. Thus, the properly working temperature regulation on your CCD camera is vital. An advanced calibration scheme, described in separate chapter, allows an observer to use a single dark frame to calibrate images of different exposure duration. The calibration steps can't be performed in any sequence - always do the dark-frame correction before the flat-frame correction. The time correction can be done in anytime during the calibration process.</p><p>The dark-frame correction subtracts the dark frame from source files. The dark frame is carried out by camera when the aperture is closed. Instead of using the automatic autodark function of your CCD camera, it is better to provide five to ten dark frames at the end of observation when the camera's chip is uniformly cooled. Then, use the "Master-dark" function to make one composed frame to be used for correction. This substantially supresses the noise. The correction frame must be saved in one of supported image formats.</p><p>Click on <tt>Calibration</tt> -> <tt>Dark correction</tt>. The new dialog appears. Find and select the dark correction frame and confirm the dialog by the <tt>OK</tt> button.</p><p>You always should use an individual dark frame for each observation series taken from several nights. Perform the correction step by step. In the table of input files, select the frames taken in one night, open correction dialog, check the <tt>selected files only</tt> option in the box below the navigation pane and select the corresponding dark frame and start the correction. Repeat the steps for each frame series.</p><div class="mediaobject"><img border=0 src="darkcorr.png" alt=""Dark correction" dialog"/></div><p>The flat-frame correction must be performed after the dark-frame correction. The flat frame is carried out when the camera is pointed to uniformly illuminated background. It eliminates variable gain of camera's chip CCD elements and also the optical system unbalanced sensitivity. To reduce the noise and increase the quality of photometry, you can combine a sequence of flat frames to obtain a "Master-flat" frame. The separate correction frame has to be used for each filter. Avoid eventual camera rotation on its mount. The correction frame must be saved in one of supported image formats.</p><p>Click on <tt>Calibration</tt> -> <tt>Flat correction</tt>. The new dialog appears. Find and select the flat correction frame and confirm the dialog by the <tt>OK</tt> button.</p><p>If you need to use individual correction frame of each subset of input files, perform the correction step by step. In the table of input files, select the frames, open correction dialog, check the <tt>selected files only</tt> option, select the corresponding flat frame and start the correction. Repeat the steps for each series of frames.</p><div class="mediaobject"><img border=0 src="flatcorr.png" alt=""Flat correction" dialog"/></div><p>Sometimes, you need to change the time of observation. The time correction modifies the date and time of observations by adding or subtracting a specified time interval. The correction can be used repeatedly - the effect of the function is cumulative, so the intervals are sumarised. For example, if you add 60 seconds and then add 40 seconds, the resulting observation time will be shifted by 100 seconds. You can also apply a correction to a subset of input files. In the table of input files, select the frames you want to change, open the correction dialog, check the <tt>selected files only</tt> option and set up the correction. Repeat the steps for each series of frames.</p><p>Click on <tt>Calibration</tt> -> <tt>Time correction</tt>. The new dialog appears. There are three ways how to define a time interval betwen actual and new observation times - you can check the <tt>by seconds</tt> option and enter a number of seconds to the edit box or you can check the <tt>by days</tt> and enter a number of days. There are two other options - you can shift the time <tt>to past</tt> or <tt> to future</tt>. Alternatively, you can check the third option <tt>by differention of dates</tt> and fill two dates into the edit boxes below. All frames will be shifted in time by the differention of these two dates. Confirm the dialog by the <tt>OK</tt> button.</p><p>If you want to modify several source files only, perform the correction step by step. In the table of input files, select the frames, open correction dialog, check the <tt>selected files only</tt> option, set up the time interval and start the correction. Repeat the steps for each series of frames.</p><div class="mediaobject"><img border=0 src="timecorr.png" alt=""Time correction" dialog"/></div><h2>Photometry</h2><p>The source frames are calibrated now, we are ready to detect stars on each frame and measure their brightness. This process is called the photometry and unlike the previous commands, the result is saved to a text-based file, so-called the photometry file.</p><p>First stage of the photometry is the detection of stars on a calibrated image. There are several parameters that controls this process, but there are two among them, which are most important and which needs to be adjusted by an user to accomodate the program to his specific quality of images. Although, the default values work fine in many cases.</p><dl><dt><b>FWHM</b></dt><dd><p>The FWHM parameter specify the expected width of stars on a frame. The value is the Full Width at Half Maximum in pixels. The parameter controls the behavior of the low-pass digital filter, which is used in the star detection algorithm.</p></dd><dt><b>Threshold</b></dt><dd><p>The Threshold parameter specify the lowest brightness of detected stars. Fainter objects are considered to be background artefacts and thus sorted out. The value is dimensionless coefficient.</p></dd></dl><p>Once you tune up the parameter for your environment, usually it is not necessary to adjust them for every task, unless the quality of your images varies considerably. In the first iteration, you can use the default values (FWHM = 3.0 and Threshold = 4.0) and do the photometry. Click on the <tt>Photometry</tt> -> <tt>Photometry</tt> item in the main menu and confirm the new dialog by the <tt>OK</tt>. Then, by double click on a frame in the main window open the preview window and check the results. If there are stars which have been detected as a close binary although it is not true, you should increase the <b>FWHM</b> value. If the stars you are interested in are not detected, try decrease the <b>Threshold</b> value. If it doesn't help, decrease the <b>FWHM</b>. If there is a lot of background artifacts detected as a real stars, increase the <b>Threshold</b>. By several iterations, adjust the parameters, so all the stars you are interested in are detected and there are no false binaries.</p><p>The photometry stores the result to a photometry file. It contains a list of stars found on a frame, their positions and brightness.</p><h2>Matching</h2><p>In all previous stages, the frames were processed separately. As a result of this, a star #1 in one file is not necessarily the same as a star #1 in another file. The matching is a process which takes one frame as a reference frame and it finds corresponding stars on all other frames. As a result, each star is assigned a unique identifier, which identifies a star on all frames.</p><p>Click on <tt>Matching</tt> -> <tt>Match stars</tt> item in the main menu. A new dialog appears. On its left, there is a list of frames and corresponding number of stars that have been found on it. In my experience, the frame with the greatest number of stars works the best. That's why the frames are sorted by number of stars in descending order. If you select a frame, it is displayed in the right part of the dialog. Select a frame that shall be used as a reference frame, and confirm the dialog by the <tt>OK</tt> button.</p><p>The reduction of the CCD images is now done. You can follow to make a light curve, find variables, etc.</p><div class="mediaobject"><img border=0 src="matching.png" alt="Matching stars"/></div></body></html>
