Skip to content

DeepSkyStacker 4.2.1
Compare
Choose a tag to compare
@perdrix52 perdrix52 released this 21 Jul 14:00
· 2521 commits to release since this release
4.2.1-Release
011fbcb
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

Welcome to DeepSkyStacker 4.2.1

The main changes from 4.2.0 are:

  1. Fixed a memory leak in the new code that uses LibRaw.

  2. Further performance enhancements to Median-Kappa-Sigma processing (20% faster than 4.2.0).

Welcome to DeepSkyStacker 4.2.0

The main change in this release is to use LibRaw version 19.2 instead of dcraw to capture images from RAW image files.
This should provide support for many recent cameras which wasn't possible using dcraw.

The raw data post-processing code has had some extra performance squeezed out if it by Vitali Pelenjow who took
my initial attempt at using OpenMP to parallelise the code and made it perform as it should (new in 4.2.0 Beta 4).

The resulting images should be very much the same (if not 100% identical) to the results obtained using DeepSkyStacker 4.1.1.

There is one thing that is incompatible with DeepSkyStacker 4.1.1:

The "Use Auto White Balance" white balance button has been disabled.

You can either accept the default which is "Daylight White Balance", or select "Use Camera White Balance". If you select the latter, and camera was set to "Auto White Balance" when the images were taken, then the camera calculated "Auto White Balance" co-efficients will be used. We don't recommend the use of Auto WB as each frame gets its own unique white balance which can prevent calibration frames such as flats and darks from working correctly.

We've also fixed a number of problems and made some other enhancements since 4.1.1

  1. Support for images from cameras that use the Fujitsu X-Trans sensor (new in 4.2.0 Beta 2).
    The image is converted to a 3 colour RGB image and interpolated using Frank Markesteijn's algorithm using LibRaw::Process().
    Please note that the interpolation is quite slow (about 40 seconds per image on my Core i7 laptop).
    The image is then imported as if it were a Foveon colour image.

  2. BSD 3 clause license file added.

  3. A fix was made to correct crashes caused by a thread-unsafe assignment in the smart pointer code.

  4. The "Stop" button on the progress dialogs should now actually work!

  5. The space key can now be used to check/uncheck images in the picture list.

  6. The user's default browser will be used instead of IE to open URLs.

  7. Miscellaneous code cleanups and performance improvements including Median-Kappa-Sigma processing.

Assets 4
Loading