Qtpfsgui now supports tiff file formats in:
"File->Load HDR..." (8,16,32 bit and logluv tiff formats)
"File->Save HDR..." (32 bit and logluv tiff formats)
"File->New HDR..." (8 and 16 bit tiff formats)
The memory allocation is now done via mmap (on linux and mac) and virtualalloc (on win32).
A missing newline has been added in the exif comments.
A byte alignment bug for PowerPC architectures ("blue image") has been fixed.
Visit http://billigites.blogspot.com to get an initial MacOSX universal binary of qtpfsgui. Please report any problem you may encounter.
This version introduces 2 new functionalities:
A dialog to resize an HDR.
A dialog to transplant exif data between 2 sets of files.
It also implements an initial dependency detection system when compiling in Linux
Now we write some exif data in the output JPEG image: we store the tonemapping parameters in the "Comment" tag and we mention qtpfsgui in the "software" tag.
We added an apply button in tonemapping dialog (this enables changing more than one thing before computing the result)
We also fixed several usability problems.
This release implements an initial support for RAW images formats, via dcraw's code.
It also contains a new "settings" panel and KDE/GNOME desktop integration.
Another release made of small bugfixes:
Fixed exif calculations when APEX values are used. This fixes another red-image hdr bug.
Fixed a Segmentation Fault (const/noconst uchar) bug that happens when compiling in Linux (Suse 10.2) against very recent versions of Qt4.
Fixed some small memory leaks.
Minor bugfixes in this release:
The only LDR image format supported is now JPEG, the one that has EXIF tags.
better EXIF tag check: if exif data not found user is warned and operation is aborted, this should eliminate the "red screen" bug completely. Let me know if it doesn't :).
restored PFS stream fileformat compatibility
fixed color-related bug for reinhard04 tonemapping operator.
v1.6 of qtpfsgui contains several performance improvements.
1) The code layout/design was changed in order to reduce memory usage.
Fit to window'' option now keeps image ratio.
3) Faster computation of durand tonemapping operator (via fftw3).
On Windows you need to download the updatedWindows Dependency Package'' that contains the fftw3 DLL.
Initial anti-ghosting (BETA) feature (use it with few and small input exposures).
restored the "what you see is what you save" behaviour in TMO dialog. (fix saveLDR bug)
removed buttons in the TMO section: modifying the settings implies istantaneous computation.
fixed shortcuts action functionality.
fixed save ldr dialog (if no extension specified it uses the one belonging to the current ``filter'')... read more
List of changes:
new feature, tonemapping settings: you can save the current tonemapping settings and the pre/post gamma values to a text file which you can later re-load.
Changed behaviour in the Tonemapping dialog:
Combobox moved to the top of the preview image for clarity.
Clicking save you'll get the full (100%) input hdr tonemapped with current settings.
* fixed several memory leaks and inefficiencies.
Version 1.3 fixes a bug when reading jpegs created by cameras that don't write ISO equiv tag (NIKON D50).
This led to a wrong HDR result (completely red).
Version 1.2 of qtpfsgui is now available on sourceforge File Release System.
This release includes many performance improvements (pfstools/pfstmo/pfshdrcreate are now included in the sourcecode, to reduce the overhead resulting from launching external programs)
Qtpfsgui now uses Qt4.
A windows binary is provided in the download page.
Project's webpage uploaded as well:
It's now available version 1.0 of Qtpfsgui.
The biggest update is that the code of pfshdrcreate is now integrated in Qtpfsgui. This means faster operations and less memory usage!
Release 1.0 is the last release before moving to Qt4 (more news in the next post).
Name of pfscalibration2 was changed into pfshdrcreate; this was needed in order not to create confusion.
Qtpfsgui, a Qt graphical user interface that
enables users to work on HDR images.
Supported operations include: creations of a HDR
file from a set of images of a scene taken at
different exposure setting, loading an existing
HDR image, saving an HDR, rotating HDRs, and
tonemapping an HDR image into a common LDR image
format (e.g jpeg or png).
This is only the graphical interface, you will also need:
1) pfstools ( http://sourceforge.net/projects/pfstools )
2) pfstmo ( http://www.mpii.mpg.de/resources/tmo )
3) pfscalibration2 ( http://sourceforge.net/projects/qtpfsgui )
pfscalibration2 is a branch of the package pfscalibration and the backend used by Qtpfsgui to
create an HDR from a set of LDR images.
Differences with pfscalibration include reduced
memory footprint (4:1 ratio) and a different
algorithm to create hdr images (i.e. devebec).
It does not support RAW camera image formats yet, but maintains compatibility with the pfs
bytestream generated by pfstools & co.