1. 04 Mar, 2019 1 commit
  2. 23 Feb, 2019 1 commit
  3. 19 Feb, 2019 1 commit
  4. 07 Feb, 2019 1 commit
  5. 01 Feb, 2019 2 commits
  6. 28 Jan, 2019 1 commit
  7. 23 Jan, 2019 1 commit
    • Wolthera van Hövell's avatar
      Temporarily push D18467 · e3c2f5be
      Wolthera van Hövell authored
      This patch changes the way how opacity and flow interact. This can mean
      that a lot of presets work way differently from before. Therefore artists
      need to test so we can get some feedback on how the patch feels.
      
      For more information, and to give feedback, check https://phabricator.kde.org/T8576
      
      This is so that artists can test it via the nightly builds.
      
      We'll try to revert it on Monday 28th of January.
      
      Patch by William Brown.
      
      Note: We still need to do a proper code review!
      
      CCMAIL:kimageshop@kde.org
      Ref T8576
      e3c2f5be
  8. 11 Jan, 2019 1 commit
  9. 09 Jan, 2019 1 commit
  10. 08 Jan, 2019 1 commit
  11. 07 Jan, 2019 1 commit
  12. 06 Jan, 2019 1 commit
  13. 05 Jan, 2019 1 commit
  14. 03 Jan, 2019 1 commit
  15. 26 Dec, 2018 2 commits
    • Dmitry Kazakov's avatar
      Remove ad-hoc implementation of converting to p2020-pq in PNG saving code · 16203f97
      Dmitry Kazakov authored
      Now it just uses the stuff from pigment
      16203f97
    • Dmitry Kazakov's avatar
      Implement proper conversions to Rec. 2020 PQ color space · 3f31062e
      Dmitry Kazakov authored
      ICC engine cannot handle PQ-shaped color spaces properly. All the
      HDR data will be lost. Therefore, we should implement a custom
      transformation for that.
      
      1) IccColorSpaceEngine now has a special method supportsColorSpace(),
         which declares that "SMPTE ST 2084 PQ" should not be handled by
         the engine.
      
      2) LcmsRGBP2020PQColorSpaceFactoryWrapper adds custom conversions
         for this special color space. Before passing it to LCMS, the
         color space will be converted into Rec. 2020 Linear. After that
         ICC color management system can handle it correctly.
      
      3) The conversions graph looks like that:
      
      p2020-pq-u8 <-> p2020-g10-f16
      p2020-pq-u16 <-> p2020-g10-f16
      p2020-pq-f16 <-> p2020-g10-f32
      p2020-pq-f32 <-> p2020-g10-f32
      
      4) There is still a problem: if one decides to convert p2020-pq-f16
         into p2020-pq-u16, then the conversion will go via p2020-g10-f32.
         This is not optimal, but it can be fixed later.
      3f31062e
  16. 20 Oct, 2018 1 commit
  17. 17 Oct, 2018 1 commit
  18. 16 Oct, 2018 1 commit
    • Ivan Yossi's avatar
      BUG:398920 · 6f05272d
      Ivan Yossi authored
      Summary:
      Fixes Invert filter for all but CMYK and LAB 32bit float
      It seems the values are not normalized on inversion step.
      
      Reviewers: #krita
      
      Tags: #krita
      
      Differential Revision: https://phabricator.kde.org/D16207
      6f05272d
  19. 12 Oct, 2018 1 commit
  20. 07 Oct, 2018 3 commits
  21. 05 Oct, 2018 6 commits
  22. 04 Oct, 2018 2 commits
  23. 29 Sep, 2018 1 commit
  24. 19 Sep, 2018 1 commit
    • Boudewijn Rempt's avatar
      Properly sort the colorspace depth combobox · b0b55278
      Boudewijn Rempt authored
      We have a fixed list of possible depths, so we can hard-code
      the order in KoColorSpaceRegistry; and after having a well-sorted
      list, we should make sure that KisCmbIdList doesn't re-sort our
      list.
      b0b55278
  25. 23 Aug, 2018 1 commit
  26. 17 Aug, 2018 1 commit
  27. 15 Aug, 2018 1 commit
  28. 14 Aug, 2018 3 commits