Ingo Weyrich
478410fa39
Merge pull request #4347 from Beep6581/lab2xyz
...
Improve accuracy of Lab <=> xyz conversions
2018-01-31 23:23:34 +01:00
Alberto Griggio
adb1392839
renamed formal parameters of LIM from (a, b, c) to (val, low, high)
2018-01-31 21:34:10 +01:00
Alberto Griggio
7dd598d9b7
bug fixes by heckflosse
2018-01-31 21:24:56 +01:00
heckflosse
8721b08346
Show correct rgb values in histogram, navigator and colour picker when using 'Rec2020'
2018-01-31 20:59:32 +01:00
Alberto Griggio
74e31dff45
histmatching: dynamically adjust the skip factor according to the side of the embedded thumbnail
...
with large the embedded thumbnails we should skip less to improve the quality of the matching
2018-01-31 18:09:28 +01:00
Alberto Griggio
b09aeaa6ae
avoid one useless copy if no cropping is required
2018-01-31 17:06:39 +01:00
Alberto Griggio
79324ed5d0
more fine-tuning of the histogram matching
...
make sure we do not place curve points in regions of the histogram that are
outside the dynamic range of the input images -- that would be wild guessing,
better to just leave this to the spline interpolation
2018-01-31 10:34:50 +01:00
Alberto Griggio
fdde31839d
incorporated Floessie's feedback
2018-01-31 08:29:19 +01:00
heckflosse
3133adb1b5
Cleanup after Flössies review
2018-01-30 21:19:20 +01:00
heckflosse
50ba33900b
some more constexpr
2018-01-30 18:59:59 +01:00
Alberto Griggio
c04150be35
histmatching: slightly improved match for the shadows part of the tone curve
2018-01-30 15:27:56 +01:00
Alberto Griggio
ffbd97960f
histmatching: make sure we use an sRGB profile for generating the target rendering
2018-01-30 10:47:18 +01:00
Alberto Griggio
5effd15bb3
histmatching: fix cropping of the target (broken by recent refactoring)
2018-01-30 10:24:19 +01:00
Alberto Griggio
71a3af0c86
improve generated curve in histogram matching on underexposed pictures
2018-01-30 10:06:57 +01:00
Alberto Griggio
8af91ce74c
faster histogram matching by using the processing pipeline in the thumbnail
2018-01-30 10:06:20 +01:00
Alberto Griggio
4ce9b0f010
avoid dependency of histmatching on the current demosaic algo
...
(by performing an internal demosaicing step in histogram matching. Obviously
this gets slower...)
2018-01-30 00:35:18 +01:00
heckflosse
e7d90ec1a2
Improve accuracy of Y to L conversion, fixes #4338
2018-01-29 21:17:34 +01:00
Ingo Weyrich
24151b313a
Merge pull request #4324 from Beep6581/remove_x86_64_dependencies
...
Removed most of the __x86_64__ dependencies. Also removed _RT_NESTED_OPENMP
2018-01-26 16:37:51 +01:00
Alberto Griggio
8a510315c2
Fattal: added anchor parameter
2018-01-25 22:13:17 +01:00
heckflosse
83bc6140bd
Avoid Lab L values which lead to negative Y values when converting Lab to XYZ, fixes #4334
2018-01-25 21:27:10 +01:00
Alberto Griggio
e4555cc51c
PP3 save: make sure that when nothing is selected, the PP3 is truly empty
...
Fixes #4330
2018-01-24 22:09:30 +01:00
heckflosse
ff20dd1a4d
Disable LIKELY and UNLIKELY for 32bit builds using gcc >= 7
2018-01-24 20:05:34 +01:00
heckflosse
9646fd84c3
Disable LIKELY and UNLIKELY for 32bit builds using gcc >= 7
2018-01-24 20:03:52 +01:00
heckflosse
96bf06a950
Removed most of the __x86_64__ dependencies. Also removed _RT_NESTED_OPENMP
2018-01-23 16:05:31 +01:00
Alberto Griggio
4cd1ad066d
imagefloat: correctly clip the values when outputting to 8- or 16-bit formats
...
Fixes #4310
2018-01-23 13:44:58 +01:00
Desmis
9a4359995f
Suppress manuel settings fot local contrsat and edge detection in wavelet
2018-01-23 10:36:50 +01:00
Alberto Griggio
bc6dad6eb9
histogram matching: fixed typo in calculating the crop
2018-01-23 00:16:19 +01:00
heckflosse
fc77fd6964
Remove all BENCHMARK defines
2018-01-22 19:04:42 +01:00
Alberto Griggio
ac1238e774
added hotfix for #3794 -- to be revised after 5.4 is out
2018-01-22 15:50:17 +01:00
Alberto Griggio
03afede882
Some tweaks to curve extraction from histogram matching
2018-01-22 13:38:23 +01:00
Desmis
dbe3d55e33
Suppress unused wavelet parameters issue4318
2018-01-22 10:06:40 +01:00
Alberto Griggio
b13db578eb
do not reset highlight compression when doing histogram matching
2018-01-21 19:21:41 +01:00
heckflosse
44984a911b
Don't set exposure to zero when 'auto levels' is disabled and 'auto matched tone curve' is enabled
2018-01-21 18:03:51 +01:00
heckflosse
88123cdd14
Set last curve point of auto matched curve always to 1;1
2018-01-21 18:01:07 +01:00
Alberto Griggio
8f76392885
histogram matching: cache the computed tone curve in RawImageSource
2018-01-19 21:43:55 +01:00
Alberto Griggio
8728e3f3f3
merge with dev
2018-01-19 13:53:37 +01:00
Alberto Griggio
4feb663f63
histogram matching: handle the case in which the thumbnail and the raw have different aspect ratios
2018-01-18 23:36:03 +01:00
heckflosse
8622efce00
Crash opening dng file from hdrmerge, fixes #4313
2018-01-18 19:53:18 +01:00
Alberto Griggio
7b3e9f7b7a
fixed bad interaction between auto levels and histogram matching
2018-01-18 17:55:14 +01:00
Alberto Griggio
abfeb4ca1d
fixed crashes in histogram matching
2018-01-18 14:17:51 +01:00
Alberto Griggio
931ee95620
use a relative height (10% of the full height) instead of an absolute one in histogram matching
2018-01-17 23:09:49 +01:00
Alberto Griggio
7bcc8ae236
removed useless lines
2018-01-17 22:25:32 +01:00
Alberto Griggio
bc854fc1e7
merge with dev
2018-01-17 22:09:47 +01:00
Alberto Griggio
942da71ef0
added some verbosity in RawImageSource::getAutoMatchedToneCurve
2018-01-17 22:02:12 +01:00
Alberto Griggio
bc89e24ab7
working version -- yay!
2018-01-17 21:16:22 +01:00
heckflosse
53ec11eea9
Cleanup fftw when closing RT, instead of after each time fftw is used. Speeds up fattal and luminance denoise
2018-01-17 20:39:01 +01:00
Alberto Griggio
52957e9eab
further experiments with histogram matching
2018-01-17 01:41:28 +01:00
Alberto Griggio
bb56d73cc8
started working on proof-of-concept histogram matching
2018-01-17 01:12:13 +01:00
heckflosse
f644d41ff2
Remove buggy (unpredictable) automatic setting of HL reconstruction when using Auto levels, fixes #4249
2018-01-16 22:34:29 +01:00
heckflosse
c139744647
RT somtimes misinterpretes image orientation of Canon CR2 files, fixes #4293
2018-01-16 20:38:29 +01:00