This is coming from my hobby project jdlraw.sourceforge.net and might
interest you as well. The file to look for is jdlDcRaw.cpp :
*) Added OMP to the VNG demosaicing.
The end result is mathematically not *exactly* equal to the non-OMP case.
However, visually, I can't see any relevant deviation.
(by the way : the reason is clear : the omp case is conceived as if
the photo is divided in N photos on which VNG is applied, this gives
differences on the boundaries versus one photo with VNG. But I believe
it is correct)
I added a user option to switch on/off OMP.
------------------------------------------------------------------------------
_______________________________________________
ufraw-devel mailing list
ufraw-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ufraw-devel
