[Digikam-users] HDR Further tests

[David Vincent-Jones]

I am away from a 'normal' internet connection for some time, so am
unable to send images regarding this problem but here is a follow-up on
my prior report.

Digikam 2.0 through the ppa, Ubuntu 10.10 KDE 4.6.2, from Canon 50D

Running (trying to run) a HDR through Digikam as well as directly using
'ExpoBlending', all attempts end in failure with a variety of input
material (using several different image sets).

CR2 files show correct EV values for the images during the process.
PNG files duplicate one of the EV values
TIF files indicate EV values as "unknown".

All file types appear to process the 3 files and do produce an image on
the screen.

With the 'blended' image on the screen I have tried using the
'Automatic' settings as well as manually making a series of adjustments
to all of the sliders with all image types. It appears that none of the
sliders are working ... at least not on my system.

The tests with the CR2 (RAW) data displayed the best image probably
since the EV values were correctly shown.

With CR2 and PNG tests the error information was displayed when I tried
to save the image. With the TIF images there was no error message.

Something to note: The error reported (see below) by the system shows
"tif" files being used ... in fact the files are Canon CR2 files
*************************************************************
General Error Message (all file types):
"Failed to process targets of bracketed images"
*************************************************************
Details of error displayed for CR2 and PNG type files:
"Cannot run enfuse:
 enfuse: info: input image
"/tmp/kde-david/kipi-expoblending-preprocessing-tmp-1306137814i6YhZB/.IMG_4278.tif", layer 1/2 does not have an alpha channel;
enfuse: info: assuming all pixels should contribute to the final image
enfuse: info: input image
"/tmp/kde-david/kipi-expoblending-preprocessing-tmp-1306137814i6YhZB/.IMG_4278.tif", layer 2/2 does not have an alpha channel;
enfuse: info: assuming all pixels should contribute to the final image
enfuse: input image
"/tmp/kde-david/kipi-expoblending-preprocessing-tmp-1306137814i6YhZB/.IMG_4278.tif", layer 2/2 has pixel type UINT8,
enfuse:   but previous images have pixel type UINT16"
**********************************************************

I would appreciate another DK user making a similar test to confirm my
findings. Looks like ExpoBlending is the culprit here.

David