Skip to content

Latest commit

 

History

History
234 lines (190 loc) · 11.2 KB

iTelescope_processing.md

File metadata and controls

234 lines (190 loc) · 11.2 KB
Raw

Processing images from iTelescope.net

Processing Stages

The following table outlines conceptual pipeline processing stages for astronomical FITS imagery, primarily those obtained using iTelescope.

Stage Stage Name Description Python Class Package Used Command Line Tool?
Calibration Generation Build Master Cal Build master calibration files Yes astropy.ccdproc ap_combine_darks.py
Calibration Generation Calc Read Noise Calculate detector read noise Yes ap_calc_read_noise.py
Calibration Generation Find Bad Pixels Generate a bad pixel map Yes ap_find_badpix.py
Calibration Generation Find Bad Columns Find bad columns for badpix.yml Yes ap_auto_badcols.py
Calibration Apply Master Cal Apply bias/dark/flat correction Yes ap_calibrate.py
Calibration Add Metadata Add metadata to calibrated FITS Yes astroplan ap_add_metadata.py
Calibration Fix Bad Pixels Correct bad pixels Yes ap_fix_badpix.py
Calibration Cosmic Ray Reject Find and correct cosmic rays Yes astropy.ccdproc Not yet
Image Processing Find Stars And PSF Find stars, measure PSF Yes astropy.photutils ap_find_stars.py
Image Processing Astrometry Astrometric solution per image Yes astrometry.net ap_astrometry.py
Image Processing Quality Filter Identify "poor quality" images Yes ap_find_stars.py
and summarize set of images Yes ap_quality_summary.py
Image Processing Image Arithmetic Arithmetic on 1 or 2 images Yes ap_imarith.py
Image Processing Background Estimation Estimate sky background Yes ap_measure_background.py
Image Processing Resample Resample an image to a WCS Not yet astromatic swarp* (resample_all.sh**)
Image Combination Continuum Subtract Continuum scaling/subtraction Not yet Not yet
Image Combination RGB Color Composite Make 3-color composites Not yet astromatic stiff* (composite_all.sh**)
Process All Images Calibrate All Calibrate all raw images Not yet (calibrate_all.sh**)
Process All Images Navigate All Astrometry/WCS on all images Not yet (navigate_all.sh**)
Process All Images Resample All Resample an image to a WCS Not yet astromatic swarp* (resample_all.sh**)

Notes:

  • (*) Temporarily using an external non-python-based tool.
  • (**) A temporary bash implementation.

Expected FITS Header Keywords

Note: This section is incompete.

The following FITS header keywords are required.

Ap Class FITS Keywords Notes
ApAddMetadata (none)
ApAstrometry IMG_FILE (a) Fom ApFindStars source list file
ApAutoBadcols (none)
ApCalibrate EXPOSURE or EXPTIME
GAIN or EGAIN (o)
ApFindBadPixels (none)
ApFindStars EXPOSURE

Notes:

  • (a) Required, but generates by another stage of AstroPhotography processing
  • (o) Optional, but not ideal if not present.

Software Preparation

The requirements.txt file should handle all software requirements except Astromatic software. However use of Astrometry.net requires a key and that key is best stored in your local astroquery configuration file.

Astrometry.net Key Configuration

Note that this is the astroquery config file, not the astropy config file. Edit ~/.astropy/config/astroquery.cfg and set api_key in the [astrometry_net] section.

If this configuration file is not present on your system, use the information here (astroquery 0.4.6 documentation) to copy the file here into a new file at the listed location.

Data Preparation

After downloading the new data (or calibration data) from the iTelescope FTP site you should run ap_fix_itelescope_dirs.sh to correct the directory permissions and remove any spaces from the directory names.

Observation Data

To unpack the data, for example for a new T05 observation of MyTarget on yyyymmdd, unpack the zip files while also removing problematic space from file names (the following commands assume the bash shell):

cd T05/MyTarget/yyyymmdd/
bash $PATH_TO_AP/src/AstroPhotography/scripts/ap_rename_files_with_spaces.sh
for file in *.zip; do unzip $file; echo ""; done
bash $PATH_TO_AP/src/AstroPhotography/scripts/ap_rename_files_with_spaces.sh
rm *.zip

Calibration Data Preparation

Calibration data from the iTelescope website should be placed in a directory of your choosing. The telescope and date specfic directory structure from iTelesope should be retained under this master calibration directory.

To point the bash shell scripts to the calibration data set an environment variable AP_CAL_DIR pointing to your calibration parent directory. For example:

> cd /scratch/iTelescopeScratch/calibration-library/
> ls 
T05  T09  T14  T16  T20  T32  T33
> tree -d T05
T05
├── Masters
│   ├── Bias
│   │   ├── 2017-07
│   │   ├── 2017-11
│   │   ├── 2018-06-27
│   │   ├── 2018-06-29
│   │   ├── 2018-10-01
│   │   ├── 2018-10-18
│   │   ├── 2019-01
│   │   ├── 2019-03
│   │   ├── 2019-08
│   │   ├── 2020-03
│   │   └── 2021-02-14
│   ├── Darks
│   │   ├── 2017-07
│   │   ├── 2017-11
│   │   ├── 2018-06-27
│   │   ├── 2018-06-29
│   │   ├── 2018-10-01
│   │   ├── 2018-10-18
│   │   ├── 2019-01
│   │   ├── 2019-03
│   │   ├── 2019-08
│   │   ├── 2020-03
│   │   └── 2021-02-14
│   └── Flats
│       ├── 2018-01
│       ├── 2018-09
│       ├── 2019-01
│       ├── 2019-02
│       ├── 2019-04
│       ├── 2019-08
│       ├── 2020-03
│       └── 2021-02-14
└── Raws
    └── Darks
        └── 2020-03

38 directories
export AP_CAL_DIR=/scratch/iTelescopeScratch/calibration-library/

A few addition steps must be taken when downloading new calibration data from the iTelescope website:

  • Fix directory permissions using ap_fix_itelescope_dirs.sh
  • Remove spaces from calibration file names using ap_rename_files_with_spaces.sh
  • Determine whether the master dark files have had the bias values subtracted or not.
  • Generate an initial master bad pixel file from the master dark file using ap_find_badpixel.py.
  • After processing real data with the initial master bad pixel file you will likely find additional bad pixels and/or columns that arise at dates or exposures different from those used for the calibration files. Careful inspection of the processed observation data (e.g. with ds9) can be used generate a user-defined bad pixel file (in yaml format) which can then be used with ap_find_badpixel.py to generate an updated master bad pixel file. (The observation data can then be reprocessed to remove the user-identified bad-pixels and columns.)

Generating An Initial Master Bad Pixel File

The following example demonstates how to generate an initial set of bad pixel files. In this case for iTelescope T20 with the 2020-04 master files.

cd calibration-library/T20/Masters/Darks/2020-04/
python3 ~/git/AstroPhotography/src/AstroPhotography/scripts/ap_find_badpix.py -l DEBUG \
    Master_Dark_1_4008x2672_Bin1x1_Temp-15C_ExpTime900s.fit \
    Master_Badpix_1_4008x2672_Bin1x1_Temp-15C_ExpTime900s.fit 
python3 ~/git/AstroPhotography/src/AstroPhotography/scripts/ap_find_badpix.py -l DEBUG \
    Master_Dark_2_2004x1336_Bin2x2_Temp-15C_ExpTime900s.fit \
    Master_Badpix_2_2004x1336_Bin2x2_Temp-15C_ExpTime900s.fit

Comparison of the resulting bad pixel files to the darks themselves show that the default bad pixel detection works well on the isolated high data value bad pixels, but misses some potentially problematic bad columns.

User-defined Bad Pixel File

These bad columns are most easily found by calibrating some real data with the initial badpix file, and inspecting the calibrated images for artifacts.

The ap_auto_badcol.py script, which uses ApAutoBadcols, can be used to find the majority of the most obvious bad columns and rows at thise stage. Its output can then be copy and pasted into a user-defined badpixel file (see template in etc/user_badpixels.yml). Rerunning ap_find_badpix.py using both the master dark and the user-defined bad pixels generates an updated master badpix file, which can then be used to recalibrate your images. For example:

# copy template yml file to t20_user_badpixels_1_4008x2672_2020.yml

# automatically detect the worse bad columns in an initially calibrated
# image
ap_auto_badcol.py 20210708/cal-T20-davestrickland-CygnusLoop_x1_y1-20210708-220604-Ha-BIN1-E-300-001.fits

# copy output into t20_user_badpixels_1_4008x2672_2020.yml

# update the master bad pixel file
ap_find_badpix.py -l DEBUG \
    Master_Dark_1_4008x2672_Bin1x1_Temp-15C_ExpTime900s.fit \
    Master_Badpix_1_4008x2672_Bin1x1_Temp-15C_ExpTime900s.fit \
    --user_badpix t20_user_badpixels_1_4008x2672_2020.yml

You would then recalibrate your images using the updated cailbration files.

Inspecting those images will likely reveal other remaining bad columns that are weaker, or only partially bad, that were missed by ap_auto_badcols. In my experience these are most easily visible in long duration narrow-band images.

You should identify those by eye, and add them to the user bad pixel file. Then rerun ap_find_badpix.py and recalibrate your images, and iterate this process until no obvious bad columns/rows/regions are visible.

Pipeline Processing