I wanted to compare the images generated by dcraw_emu.exe and the latest version of dcraw, which I compiled on Windows 7 using:
gcc -o dcraw.exe -O4 dcraw.c -lm -lws2_32 -DNO_JPEG -DNO_LCMS -DDJGPP
I then ran both dcraw.exe and dcraw_emu.exe with the same command line parameters:
dcraw.exe -g 1.0 1.0 -W -q 3 data.cr2
dcraw_emu.exe -g 1.0 1.0 -W -q 3 data.cr2
using this image:
I would have thought that the output .ppm file would be binary identical from both tests but I can easily see that the result from dcraw_emu.exe always appears brighter than dcraw.exe across the entire image. Is there a step that I am missing or can you explain what is going on here please?
LibRaw (and, so, dcraw_emu)
LibRaw (and, so, dcraw_emu) contains 'auto-maximum' feature to deal with 'pink clouds' problem.
If you want to produce files which are binary identical to dcraw's results, you should switch this option off. Use -c 0 switch of dcraw_emu to do that.
-- Alex Tutubalin @LibRaw LLC