Well, those comments are for raw converter developers.

Iliah, You say "It looks like the time spent in postprocessing to get more details and counter-act artifacts might be better spent if demosaicking prepares a cleaner and more detailed image to start with. The processing time difference between integer and floating point implementations of demosaicking most of the times is less than 5 seconds, which is substantially less compared to the time spent on image editing to get close to the same result. Try it, and you will see. "

How does one go about "trying it" to see? What about this comment:
"For float data output we need to change entire processing pipeline to floating point." from http://www.libraw.org/about?

Elle Stone

Take a look at the method "LoadRawImage" at https://valelab.ucsf.edu/svn/micromanager2/trunk/DeviceAdapters/Tethered... )

The problem is very simple:
- you have set half_size=1 *after* unpack is called. So for the very first run the half_size is not set for unpack stage but set for postprocessing stage.

But half_size option is heavily used on unpack() stage (only 1/4 of image array allocated, pixel data placement is very different).

You need to move
rawProcessor_.imgdata.params.half_size = 1;
line above of unpack() call.

I'm using LibRaw 0.13.1. I've tested on Windows and macosx; and in each case there is a difference between the first and the second decoded image.

#ifdef WIN32
// suppress sprintf-related warning. sprintf() is permitted in sample code
#define _CRT_SECURE_NO_WARNINGS
#endif
 
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <math.h>
#include <ctype.h>
#ifndef WIN32
#include <fcntl.h>
#include <unistd.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <sys/time.h>
#endif
 
#include "libraw/libraw.h"
#ifdef WIN32
#define snprintf _snprintf
#include <windows.h>
#endif
 
 
void write_ppm(libraw_processed_image_t *img, const char *basename)
{
    if(!img) return;
    // type SHOULD be LIBRAW_IMAGE_BITMAP, but we'll check
    if(img->type != LIBRAW_IMAGE_BITMAP) return;
    // only 3-color images supported...
    if(img->colors != 3) return;
 
    char fn[1024];
    snprintf(fn,1024,"%s.ppm",basename);
    FILE *f = fopen(fn,"wb");
    if(!f) return;
    fprintf (f, "P6\n%d %d\n%d\n", img->width, img->height, (1 << img->bits)-1);
/*
  NOTE:
  data in img->data is not converted to network byte order.
  So, we should swap values on some architectures for dcraw compatibility
  (unfortunately, xv cannot display 16-bit PPMs with network byte order data
 
#define SWAP(a,b) { a ^= b; a ^= (b ^= a); }
    if (img->bits == 16 && htons(0x55aa) != 0x55aa)
        for(unsigned i=0; i< img->data_size; i+=2)
            SWAP(img->data[i],img->data[i+1]);
#undef SWAP
*/
    fwrite(img->data,img->data_size,1,f);
    fclose(f);
}
 
int main(int argc, char *argv[])
{
	char filename[] = "raw_canon_550d.cr2";
	LibRaw rawProcessor_;
	int rc;
	for (int i = 0; i < 10; i++)
	{
		libraw_processed_image_t* rawImg = NULL;
		rc = rawProcessor_.open_file(filename);
		rc = rawProcessor_.unpack();
		rawProcessor_.imgdata.params.filtering_mode = LIBRAW_FILTERING_AUTOMATIC;
		rawProcessor_.imgdata.params.output_bps = 16; // Write 16 bits per color value
//		rawProcessor_.imgdata.params.gamm[0] = rawProcessor_.imgdata.params.gamm[1] = 1.0; // linear gamma curve
//		rawProcessor_.imgdata.params.no_auto_bright = 1; // Don't use automatic increase of brightness by histogram.
		rawProcessor_.imgdata.params.document_mode = 0; // standard processing (with white balance)
		rawProcessor_.imgdata.params.use_camera_wb = 1; // If possible, use the white balance from the camera.
		rawProcessor_.imgdata.params.half_size = 1;
		rc = rawProcessor_.dcraw_process();
		rawImg = rawProcessor_.dcraw_make_mem_image(&rc);
		char ppmfname[128];
		snprintf(ppmfname,1024,"%04i",i);
		write_ppm(rawImg, ppmfname);
		rawProcessor_.dcraw_clear_mem(rawImg);
		rawProcessor_.recycle();
	} 
	return 0;
}
//not truncated

Is is possible, that ESDK memory is protected from outside access and only ESDK call may use it?

Unfortunately, I've no reply from Canon-Europe yet, so cannot look into ESDK.

Good news.
If i use my memory to dowload cr2 file with edsdk, i haven't crash :D

thanks.

No-no.

libraw (.a, .so) is compiled with -DLIBRAW_NOTHREADS, so some internal decoder buffers are static. This is faster, but this library is not reentrant. You may use it in multithreaded environment, but not for decoding several files in parallel

libraw_r is compiled with -pthread and without LIBRAW_NOTHREADS define, so all I/O calls are made via thread-safe version and decoder storage is allocated within LibRaw class. So, extra locks, extra indirection for storage, but thread-safe.

Not reentrant version is several percent faster.

In the best, is that i understand.

I thinks if you use pimpl idiom on LibRaw class you can remove this problem.
If i understand correctly the error, your libraw_memmgr implementation is declared in .h and is inlined.
So when you use it in application free and malloc are not the same that use in the dll.

http://support.microsoft.com/kb/140584/en-us

Thanks for info.

It looks like I need to provide four different DLLS:
librawT.dll - /MT
librawTd.dll - /MTd
librawD.dll - /MD
librawDd.dll - /MDd

Oh! Windows make me crazy!