Closed Beep6581 closed 9 years ago
Patch applied to staging for testing.
Reported by wyatt.olson
on 2010-06-04 03:58:49
PatchSubmitted
Oops! a bug in a couple of parameters; here is a patch to fix that...
Reported by ejm.60657
on 2010-06-04 04:43:16
Why in CmakeLists.txt was
if (OPTION_OMP)
find_package(OpenMP)
...
commented out?
This was added to allow compilation with OMP on Linux and disable it on other platform.
Reported by ffsup2@yahoo.it
on 2010-06-04 11:58:24
Committed patch and uncommented OpenMP conditional. Emil -- use -D OPTION_OMP=false
to disable OpenMP, no need to comment anything in the CMakeList.txt file.
Cheers
Reported by wyatt.olson
on 2010-06-04 14:07:38
I commented it out because on my machine, cmake was aborting with complaints that
OpenMP was not found. With the commented code in CMakeLists.txt, I was able to
build. Apologies if this screwed things up for others, I just want to be able to
work on my stuff, and am willing to entertain expedient hacks to do so.
I will have another minor patch in a moment (parameter tweaks, overshoot protection
on edge corrections)...
Reported by ejm.60657
on 2010-06-04 14:46:52
NP, but you should be able to just use the cmake flag to remove it... it is working
fine on my OSX box at least, if yours does not work with the flag, please let me
know, as that would be a major bug (to my knowledge this works on all platforms, I
know for sure that someone has compiled a Windows version using it).
Reported by wyatt.olson
on 2010-06-04 14:49:48
As promised, bugixes and code improvements for CA autocorrect (and also a couple for
AMaZE).
Reported by ejm.60657
on 2010-06-04 18:34:58
Pushed to staging. Thanks!
Reported by wyatt.olson
on 2010-06-04 18:37:17
I am getting somewhat poorer results for some images with CA autocorrection in RT
than I am getting with the same code applied in dcraw. One possibility is different
order of processing. In dcraw, scale_colors is applied before before CA correction
and thus before demosaic; this is a kind of WB preconditioning of the data which is
important.
What is the order of processing in RT? Is there any WB pre-conditioning of the data
before demosaic (there should be)? If so, where is it in the code?
Reported by ejm.60657
on 2010-06-05 16:33:29
What is the status of this issue? I know there has been many patches for Amaze and
/ or CA correction since the last post here. If you feel this is working, please let
me know and I can update the ticket.
Cheers
Reported by wyatt.olson
on 2010-07-11 19:42:12
PatchApplied
Yes the ticket can be closed.
Reported by ejm.60657
on 2010-07-12 03:04:56
Reported by wyatt.olson
on 2010-07-12 03:06:33
Fixed
The last time i've tried, it still crashed, on Windows at least.
It only crash when using GCC optimizations... I'll check again as soon as i've finished
my current patch. If it's still ther, i'd like to talk to Emile to solve this problem
(if it's still there).
Reported by natureh.510
on 2010-07-12 12:29:40
Sounds good... I will leave the bug closed until I hear back from you, and we can decide
from there.
Cheers
Reported by wyatt.olson
on 2010-07-12 14:23:44
Originally reported on Google Code with ID 135
Reported by
ejm.60657
on 2010-06-04 03:48:51