Bonjour,
I just delivered my submitted version of DNG.c and RAW.c (
http://chdk.setepontos.com/index.php?topic=650.msg60887#msg60887) after I succeed to modify them to save a big-endian dng file (
http://chdk.setepontos.com/index.php?topic=5940.msg58697#msg58697) but without good result.
Now, I am convinced, in fact I convinced myself, that the best performing way to benefit from DNG advantages, is to jointly post-process chdk-raw files and Canon-jpeg files as per DNG4PS2 is doing.
For me, personnally, it's adding an intermediate process between capture phase and my photo softwares (lightroom, photoshop, autopano pro, Dynamic HDR, ...) that I use in parallel way. In fact, instead of simply importing captures, using integrated import function, I will process-import captures and post-processes dng files. Well, it's not so bad.
But, to perform conveniently, my project is to add "OpcodeList1" tag as a separate file to enumerate bad pixels to post-process. I expect to create that "partial tag" file during "badpixels.bin" creating process.
The first point here is: "I expect that some day DNG4PS2 developers will integrate that tagged bad pixel file, as it integrates jpeg tags, into his final dng file".
The second point is: "To have a complete bad pixel list, badpixels.bin and badpixels.txt shall form the final tagged bad pixels and therefore consider both internal list as the bad pixel list." I am relatively comfortable with the interpretation of these files (
http://chdk.setepontos.com/index.php?topic=5939.msg58693#msg58693).
Before starting working on this, I would greatly appreciate your comments.
Thanks.
