I am developing an interface between javafx and libraw. Generally it is now working but only if I write directly to a TIFF or PPM file. I do not want writing first into a temp file and reading again from that file. Instat I want to use the memory image from libraw.
The problem actually I am facing with it is accessing the data from the call "libraw_dcraw_make_mem_image". From the documentation it should be an array of rgb pixels but all I am getting is one character back from an memorysegment.
My app is written with JavaFX and therefore I want to use libraw via the newly ability of JDK16 to access native libs (C Libs) without writing any glue code.
It looks like there is a subset of Lumix DMC-LX100 II images that are coming out magenta. I intercepted the Bayer pattern outside of LibRaw and it is coming out as:
12
01
if I manually change it to
21
10
then the image is demosaiced correctly. It looks like the pattern is off by one.
I should mention FastRawViewer is also showing the incorrect color.
Is this a known issue? Is there a patch or could you point me to where in LibRaw it could be fixed?
maybe someone can help me with this:
-> I am trying to get the actual LibRaw 0.20.2 into FreeImage 3.18.0. Due to the reorganized code in LibRaw 0.20.2 i get the following error in the PluginRAW.cpp of FreeImage: "Identifier 'substream' not declared"
It seems the 'substream' has been declared within LibRaw before, with the version 0.19_201910_snapshot and the 0.19.5 rc it works fine to put the LibRaw code into FreeImage 3.18.x and build that library.
Hopefully someone has an idea, as a I am trying to figure this out for days...
I have a problem with LibRaw(0.20.2, MSVC2012) using std::map
Here is the test code
----------------------------------------------------------------------
std::map test;
test["elem1"].imgdata.params.user_qual = 10;
----------------------------------------------------------------------
There is no compile error, but runtime error occurs at
"test["elem1"].imgdata.params.user_qual = 10;"
I don't know why std::vector is work well but std::map is NOT work well.....
Does anyone know how to fix this error?
Is there a way to force raw processing to use daylight WB? I would like to ignore WB as set in the camera and process all images using daylight WB. With no WB selected, the images I have tested seem to come out with a yellow cast. While this can be corrected using an ICC profile afterwards, it would be nice to use a starting point closer to an accurate target rendition.
Recent comments