Skip to content

Latest commit

 

History

History
182 lines (146 loc) · 7.9 KB

RELEASE_NOTES.md

File metadata and controls

182 lines (146 loc) · 7.9 KB
Raw

AstroPhotography Release Notes

Release notes for versions starting at v0.2.2 and later.

Version 0.5

The objectives of version 0.5 are:

  • Catch up to latest astropy (>=6.0) and affiliated package changes.
  • Working sphinx documentation!
  • Package layout reorganization to move past some limitations imposed by the original cookiecutter template, that are no longer helpful.
  • Easier and more pythonic configuration for processing a set of images. The example bash shell scripts are not sustainable.
  • The ability to process the metadata-less "Premium" datasets now available from iTelescope.

Version 0.5.1 2024-01-30

  • Improve documentation now that we can run sphinx on the module.
  • Resolved #017, added yamlkeyval mode to ApAddMetadata / ap_add_metadata that works on iTelescope premium images.
  • #021, catch up to latest astropy (>=6.0) and affiliated package changes. Fixes to ApFindStars, ApMeasureStars, ApFixCosmicRays, and ApMeasureBackground.
  • Add AP_CALIB_DIR to calibrate_all.sh script, document in doc/iTelescope_processing.md

Version 0.5.0 2022-09-18

  • The module itself has moved from ./src/AstroPhotography to ./AstroPhotography.
  • sphinx-generated documenation, including the module docstrings, now works. Change directory into the doc directory and run make html, then view the file in _build/html/index.html.

Limitations:

  • Python docstrings may be incorrect in some places, but that will be addressed at a later date.
  • The default font size used in the sphinx sizzle theme is too large.

Version 0.4

Version 0.4 returns to issues related to processing of FITS images, in particular from iTelescope.

v0.4.2 2022-09-11

  • WIP on describing how to prepare for data processing and new calibration files in iTelescope_processing.md.
  • WIP on capability to process telescopes other than T05 in calibrate_all.sh (although note this script is a temporary clunky solution for batch processing).
  • Added more options to ApMeasureBackground and ap_measure_background.py that allow user-driven finetuning of the background estimations, based on experience with T20 observations of the Cygnus Loop.
  • Added ApAutoBadcols and ap_auto_badcols to automatically detect the majority of bad columns and rows when first generating a user-defined bad pixel calibration file. (Issue #14)
  • Improved handling of failed name resolution in ap_add_metadata.py
  • Added capability to specify target name on ap_add_metadata.py in cases where file name parsing will not work.
  • ApAddMetadata will now attempt to remove the mosaic subfield identifiers generated by Telescopius, e.g. CygnusLoop_x1_y1 is no longer treated as CygnusLoop x1 y1 but as CygnusLoop instead.
  • Added ap_fix_itelescope_dirs.sh and ap_rename_files_with_spaces.sh

v0.4.1 2021-07-24

  • Updated the requirements.txt file with long term aim of supporting python39.
  • Added ap_measure_background.py and ApMeasureBackground.py to resolve issue #8. These measure and output an image of the large scale background in an image that can be used to correct for imperfect bias/dark/flat calibration or moon illumination.
  • Big changes in the ApFindStars interface to support it being used by external classes instead of doing all I/O on files.

v0.4.0 2021-05-22

  • Added ap_imarith.py and ApImArith to supply a native way of performing simple arithmetical operations on one or two FITS files.
  • Added example bash driver scripts that can process multiple images from the observation of a single target. They're somewhat clunky to set up, so they're only provided as an example.

Limitations:

  • ap_imarith.py and ApImArith assume the data is always the first extension (Primary HDU) of the FITS file. This does not allow for multi-extension files (MEF).

Version 0.3

Version 0.3 is an interim release that attempts to finally provide the basic "required" functionality to the RAW file conversion tool dksraw, before attention switches back to FITS file processing.

v0.3.0 2021-04-04

  • FITS files can now be written by the dksraw grey, dksraw rgb, and dksraw split commands. This expects you to specify file names ending in in the full .fits (not .fit).
  • There is a new command dksraw rgb, that generates 3-channel RGB images from RAW files. If the output file is FITS format the red, green and blue channels are written to separate 2-D image extensions within the same file. They can be loaded as a single RGB image in ds9 using the command line, e.g. ds9 -rgb -red test_rgb.fits[0] -green test_rgb.fits[1] -blue test_rgb.fits[2]
  • The dksraw grey and new dksraw rgb commands now have a (default) linear method for Bayer demasking and looks much better than the old direct method.
  • The dksraw grey and dksraw rgb commands have a --renormalize option suitable for generating quick-look images that fill the dynamic range of the output file.
  • RAW file EXIF metadata is now read within RawConv, using the python package ExifRead. This is less capable than a command line tool like exiv2. Nevertheless a limited set of this metadata is written to FITS files (if used as output).
  • Minor CLI improvements:
    • Renamed --warn to --loglevel for consistency with the ap_ scripts, and changed how it was specified in argparse to avoid having to specify it before the subcommand.
    • Made "black" level subtraction the default in dksraw, so now you have to specify --keepblack if you want the black levels to be retained.
  • The old TODO.md file has been removed as it was very out of date. See the github Issue for a clearer view of development priorities.

Limitations:

  • EXIF metadata writing to graphics files formats (tiff, jpg, png etc) has not been successfully implemented, as there is a disconnect between python packages that can read metadata from RAW file and those that claim to write valid EXIF metadata to graphics file formats that I have not had time to solve.
  • Sphinx documentation still does not work.
  • We are still using imageio for graphics-file format output, which is very slow for formats other than TIFF.

Version 0.2

v0.2.4 2021-02-27

The emphasis of this release improvements in the calibration process, in particular bad pixel removal and the newly added cosmic ray removal.

  • Added cosmic ray removal using ccdproc's lacosmic algorithm in ApFixCosmicRays, invoked either by ap_fix_cosmic_rays.py or by ApCalibrate/ap_calibrate.
  • Disabled the default tweak_order=2 SIP polynomial fit in astrometry.net
  • Added option to apply user-defined bad pixels in ap_find_badpix.py and ApFindBadPixels.
  • Fix issue #5, allowing darks that either have bias subtraction or have not (--dark_still_biased).

v0.2.3 2020-12-22

The emphasis of this release is on the capability to process FITS images from amatuer astronomical observations, in particular iTelescope.net.

The ap_ scripts now allow iTelescope.net data to be calibrated from raw fits frames, through bias/dark/flat correction, star detection and WCS solutions.

Outstanding Issues:

  • The ap_ scripts are not installed as console scripts, and some still don't have their Ap implementation classes available externally.
  • sphinx documentation is not working.
  • RAW conversion still doesn't output FITS.
  • The Ap classes and ap_ scripts were tested by analysis, not with automated unit tests.

v0.2.2 2020-07-12

Continued work on conversion of camera RAW files, including conversion to grey-scale 16-bit formats, splitting by Bayer channels and white balance related functions. Output is standard imaging formats, e.g. png.

Outsanding Issues:

  • Conversions to FITS format not yet implemented, although this should be trivial.
  • Python imageio is simple and easy to work with, but doesn't seem to properly support metadata.
  • Unit tests are clunky and depend on external "truth" calculations made in octave/Matlab