4.2 Preparing the SXT Data for Analysis (SXT_PREP)

It is the desire of the SXT team to have the routine SXT_PREP used for most of the preparation of the SXT data. The SXT_PREP routine performs a variety of tasks, some of which are described in the following list.

1. Flag the saturated pixels
2. Calculate the decompression error uncertainties for each pixel
3. Perform standard corrections to the data
   • Decompress the data
   • Subtract the dark current
   • Subtract the pin-hole visible stray light
4. Align the images and build the observing region (or mosaic of PFI images). The routine can handle PFI, FFI, or a mix of images. It can extract subportions out of FFI images. The alignment can be done
   • in heliocentric coordinates
   • relative to sun center
   • relative to CCD pixels
5. Optionally fill in the horizontal gaps in the observing region (or)
6. Optionally exposure normalize to a one second exposure

Some important elements of SXT_PREP are:

• The default when registering image is to convert all images to full resolution. This can be overridden by the optional keyword outres=outres.
• The signal is rescaled if the resolution is changed. This occurs whether the exposure normalize option is selected or not. The signal is divided by 4 when going from HR to FR, or from QR to HR (and divided by 16 when going from QR to FR).
• The optical images are shifted to be aligned with the X-ray images starting with SXT_PREP version 1.20 (4-Jan-94). The shift is less than 1.5 FR pixels.
• The S/C roll and SXT roll offset can be corrected so that solar north is straight up by using the keyword option /roll
• The /sfd keyword should be set when using SXT_PREP on SFD files
• The units out of SXT_PREP are data number (DN) unless the /normalize options is used, in which case they are DN/s.

The steps to using SXT_PREP can be somewhat involved. The basic steps are described below, but the routine XSXT_PREP is an easy way for a user to specify what processing should be performed, and XSXT_PREP is described at the end of this section.

1. Select the files which have the data of interest. You can use YODAT to do this, or some other method.

2. Select the images that you wish to process. Again, YODAT can do this, or you can use SSWHERE or another method for selecting the images. If you are not using YODAT, then one option is to read the roadmaps of the files, and pass the roadmap to SSWHERE. The following example assumes the list of file names (as a string vector) is in the variable infil.
   IDL> rd_roadmap, infil, rmap
   
   IDL> ss = sswhere(rmap)
   

3. To register images, it might be necessary to determine the coordinates for the center of the aligned images. Some methods for specifying the coordinates are:
   • Allow the routine to find the center of all images, and take the average. This works fine if the selection has been done carefully and the images selected are from one location. This also works fine if you are registering FFI images.
   • Specify the heliocentric coordinates specifically helio=[-20.2,50.3], helio_date='8-dec-93 4:43'
   • specify the sun coordinates specifically suncoord=[400,600]
   • Use a PFI image as a reference image image_ref = pfi_index(40)
   • Use a PFI image as a reference image and select heliocentric tracking. image_ref = pfi_index(40), /helio

   If you only selected PFI images, and the PFI images you selected are all for the same active region, then you do not need to specify the coordinates of an aligned image (it will automatically derive the optimal coordinates by averaging all PFI center locations). If they are not for the same region, then you should either modify your selection so that it only includes one location on the sun, or select a image which all of the data should be aligned to and use the ref_image=index(ialign) option, or you can specify the absolute coordinates (the procedure for establishing those coordinates is described below).

   If you have mixed PFI and FFI images, then you can specify one of the partial frame images, and then all images will be registered to that image, and the appropriate portion of the FFI image will be extracted and aligned.

   If you selected only FFI images, but you want to extract a portion to make a registered movie, then you will need to determine the coordinates for the registered image. A simple method to do this is to use SXT_GRID. If you want to simply remove the S/C jitter and the SXT pointing changes (register relative to sun center), then display a full frame image and call SXT_GRID. An example is:
   IDL> tvscl, data(*,*,5)
   
   IDL> sxt_grid, index(5), /read_out, last=loc, /angle
   As you move the cursor around, you will see that the values in the small window are changing. When the cursor is located at the location where you want the center of your image, press the right button. If you want to register images and track on a heliocentric coordinate, then you can use SXT_GRID again to get those coordinates. An example is:
   IDL> tvscl, data(*,*,5)
   
   IDL> sxt_grid, index(5), /read_out, last=loc
   It is necessary to specify the size of the output image when extracting a partial frame out of a full frame.

4. Run SXT_PREP. If you are reading the data and passing the index and data to SXT_PREP, then the observing regions should not be assembled (answer ``No'' to this question in YODAT). This is important because jitter and drift can cause pointing changes between the different exposures that comprise of one PFI observing region.

   It is possible to pass index and data into SXT_PREP, or to pass the file name array (infil) and the data sets to extract (ss). The results can be returned in the third and fourth parameters (index2 and data2) or they can be written directly to an output file (outfil=outfil).

   Below are some examples of how to call SXT_PREP.

XSXT_PREP provides a graphical user interface to SXT_PREP. The XSXT_PREP routine simplifies the access to the many SXT_PREP options and varied calling sequences. In addition, XSXT_PREP provides a front end to some additional data selection and filtering tools which may be used to fine tune the data set and allows graphical definition of all SXT_PREP input parameters. The following graphical tools are provided:

• selection of alignment coordinates and definition of output data size by display and mouse selection [optionally overlay on SXT Full Frame Image].
• filtering and re-filtering of input data (can spawn call to SSWHERE).
• ability to step through input data to select alignment reference image.
• given a time, user may graphically select AR from FFI (supplied or from SFM data) to identify FFI and PFI images which contain that AR - process if data are available online or setup background job.
• selection and display of all SXT_PREP parameters and display associated SXT_PREP calling sequence.
• spawn SXT_PREP on request or spawn a background task (file output).
• maintain log of SXT_PREP input parameters and calling sequence so that work can be repeated or reviewed at a later date.

The combination of graphical interface and online help make this a good routine to learn through use. It is a main level routine. For starters, try the following calling sequence:
IDL> .run yodat ; select SXT data cube to process

IDL> .run xsxt_prep ; play around