diff options
author | B Stack <bgstack15@gmail.com> | 2019-03-13 10:36:44 +0000 |
---|---|---|
committer | B Stack <bgstack15@gmail.com> | 2019-03-13 10:36:44 +0000 |
commit | 2c01454a3fea1c29df0aeb208862243cb928b74c (patch) | |
tree | dc226d83311470a23cdf0c02064feb525fcc5096 /xBRZ/src/xbrz_tools.h | |
parent | Merge branch '10.9' into 'master' (diff) | |
parent | 10.10 (diff) | |
download | FreeFileSync-10.10.tar.gz FreeFileSync-10.10.tar.bz2 FreeFileSync-10.10.zip |
Merge branch '10.10' into 'master'10.10
10.10
Latest changes:
* New option: synchronize selection
* Dynamically disable unsuitable context menu options
* Support MTP devices without move command
* Fall back to copy/delete when implicitly moving to different device (e.g. symlink)
* Fixed incorrect statistics after parallel move
* Fixed menu button not triggering context menu
* Fixed crash on focus change while message popup is dismissed
* Fixed crash when trying to shrink empty image
* Fixed invisible dialogs when monitor is turned off in multi-monitor setup
* Work around GetFileInformationByHandle error code 58 on WD My Cloud EX
* Changing deletion handling now correctly triggers updated config
* Support root-relative FTP file paths (e.g. FreeNAS)
* Move and rename MTP items as a transaction
* Exclude AppleDouble files (._) via default filter on macOS
* Support home path for FTP folder picker
* Use server default permissions when creating SFTP folder
* Use native OpenSSL AES-CTR rather than libssh2 fallback
* Added context information for cloud connection errors
* Updated translation files
See merge request opensource-tracking/FreeFileSync!7
Diffstat (limited to 'xBRZ/src/xbrz_tools.h')
-rw-r--r-- | xBRZ/src/xbrz_tools.h | 10 |
1 files changed, 4 insertions, 6 deletions
diff --git a/xBRZ/src/xbrz_tools.h b/xBRZ/src/xbrz_tools.h index aaa2b769..15bea025 100644 --- a/xBRZ/src/xbrz_tools.h +++ b/xBRZ/src/xbrz_tools.h @@ -184,7 +184,7 @@ void bilinearScale(const uint32_t* src, int srcWidth, int srcHeight, int srcPitc //perf notes: // -> double-based calculation is (slightly) faster than float - // -> precalculation gives significant boost; std::vector<> memory allocation is negligible! + // -> pre-calculation gives significant boost; std::vector<> memory allocation is negligible! struct CoeffsX { int x1 = 0; @@ -233,11 +233,9 @@ void bilinearScale(const uint32_t* src, int srcWidth, int srcHeight, int srcPitc auto interpolate = [=](int offset) { - /* - https://en.wikipedia.org/wiki/Bilinear_interpolation - (c11(x2 - x) + c21(x - x1)) * (y2 - y ) + - (c12(x2 - x) + c22(x - x1)) * (y - y1) - */ + /* https://en.wikipedia.org/wiki/Bilinear_interpolation + (c11(x2 - x) + c21(x - x1)) * (y2 - y ) + + (c12(x2 - x) + c22(x - x1)) * (y - y1) */ const auto c11 = (srcLine [x1] >> (8 * offset)) & 0xff; const auto c21 = (srcLine [x2] >> (8 * offset)) & 0xff; const auto c12 = (srcLineNext[x1] >> (8 * offset)) & 0xff; |