The book Digital photo development with darktable has been released. Check out the pdf or packaged version and, as usual, send your comments and feedback to the darktable user list.

we released version 0.9.2, with a few bugfixes on top of release 0.9.1. this is a new point release in the stable branch thus there are no new features, just updated translations tiling for memory hungry operations and as workaround for old opencl 1.0 drivers new color matrices and white balance presets a lot of stability issues have been resolved a lot of performance improvements (more sse code, better opencl code) new packages are available, check the installation instructions for more infos!

__we are excited to announce we released the first final draft of the book darktable. check out the pdf or packaged version and send your comments and feedback to the darktable user list.

we released version 0.9.1, with a few bugfixes on top of release 0.9. and 184 patches, among them new rawspeed, dcraw, libraw fixed various segfaults and deadlocks the pipeline is now more real HDR (unbounded color management, no more gamut clipping in between) fixed a nasty bug which could cause complete loss of history for an image darktable-faster now plays nicely with darktablerc (non-gconf) lots of opencl improvements updated translations second part of our GSoC: customizable keyboard shortcuts!

we released draft 0.6 of the book darktable. this is meant to be a comprehensive guide to the use of darktable for the most common development workflows. this is a book for beginners who do not want just to know how to develop an image, but looking for more information on digital photography and how it applies to darktable. check out the pdf or packaged version and send your comments and feedback to the darktable user list.

we released version 0.9, with many new features: run-time switchable opencl to exploit all the power of your GPU whenever you decide to install the driver many new modules, including a spot removal tool, better denoising (on raw pixels and non-local means) and many more blend operations, overlay your module only 20 percent if you want spot removal tool low light vision tool non-local-means denoising (relatively fast for nlmeans, but still slow) first part of the google summer of code project already merged framing module (adds postcard borders to match given aspect ratio) tonemapping a lot faster now (probably the fastest high-dimensional bilateral filter available today) changed images come with the darktable|changed tag more updates to the page/install instructions to come, but you should be good by just extracting the tarball and typing .

Today I’ll be starting on my Summer of Code project for darktable, so I thought I’d start off with a blog post about just what I’ll be doing. This Summer I’ll be focusing on UI improvements in darktable, and I have four separate tasks to complete, in this order. Removing the libglade dependence. There are two ways to construct a graphical interface in software: one is to build your interface one element at a time in your source code, and the other is to use a graphical tool to build a description of your desired interface that you can use a library (libglade in our case) to construct when the program actually runs.

As some of you might have noticed, darktable got accepted for this year’s Google Summer of Code (warning, site doesn’t work with every browser …). Currently we are seeing a few possible students lurking around in IRC, preparing their proposals in our Wiki and getting familiar with the program, code and community. If everything keeps going so smoothly this will be just f****ing awesome. If not we still have the excuse that it’s our first time, so we didn’t know what we were doing.

we released version 0.8, which obsoletes 0.7.1 in a lot of ways: much faster image loading due to rawspeed, an awesome new library by klaus post @rawstudio lots of performance improvements in our caches and pixel pipelines (together with the above like 5x–10x) gpu computing using opencl (for graphics boards that support it) for a lot of common modules, to give a huge performance boost overhauled collection module for more flexible image collections metadata editor (set author and copyright information etc) fast demosaicing now done on roi and in floating point HDR bracketing and tone mapping (somewhat experimental) flickr upload new languages: thai and japanese lots of new color matrices and white balance presets lots of bugfixes according to the git log, this release introduces over 900 new commits brought to you by (in order of commits): johannes hanika, Tobias Ellinghaus, Henrik Andersson, Pascal de Bruijn, Ger Siemerink, Bruce Guenter, Jose Carlos Garcia Sogo, Boucman, Alexandre Prokoudine, Simon Spannagel, Olivier, Jochen, Karl Mikaelsson, Jochen Schroeder, Brian Teague, Pascal Obry, calca, Ville Pätsi, Uli Scholler, Thierry Leconte, Pacsal de Bruijn, and Alex Chateau.

we released version 0.7.1, a small bugfix release to fix up some nuisances in 0.7: some more white balance presets layout fixes for overlong profile names styles now actually work extensive documentation in form of screencasts enjoy the release, but be aware that you’ll be missing out on a lot of significant speed improvements and cool new features when using the release instead of git :)

we released version 0.7 this weekend. some of the changes: .xmp is now used instead of .dt for sidecars libraw with dcraw 9.05 plug-ins of darkroom mode are now in groups/tabs new darkroom plug-ins: zone system, relight, graduated neutral density, watermark changes can now be saved into named processing styles initial preferences dialog was added custom shooting settings can be added in tethering mode initial user manual was written, available in english, german, french, dutch, and swedish moved to xdg config file standard languages: ca cs de es fi fr gl it nl pl ru sq sv reset labels: double click the label to only reset this option, not the whole plug-in a cropped image can now be recropped without resetting the whole plug-in (press backspace) lots of new cameras with enhanced support (input profiles and base curves) recursive image import 962 commits since release-0.

while waiting for 0.7, we have put up an early draft usermanual for you, to push forward this great effort, get some feedback, and reveal some details to you. thanks to henrik and olivier who worked hard on this one!

finally a new release … it has been so long that i hardly remember all the changes. let’s try to list the most important ones: libraw 0.10.0-beta3 tethered shooting mode import from camera via gphoto2 new, improved modules: vignetting, velvia, grain, denoise (via bilateral filter), color transfer, channel mixer … crop/rotate: straighten tool, perspective correction, guide lines (ported from digikam) lots of new input color matrices and base curves openexr export database format changed, which greatly improves speed (and cuts down used disk memory) film strip view in darkroom mode for quick image switching cool alternative demosaicing algorithms (ported from rawtherapee): dcb, amaze, vcd, c/a autocorrection customizable preset system for all darkroom modules with auto application to matching images, selected by exif reworked export system to modularly support export to picasa webalbum, email, or disk in jpg, png, 8/16-tiff, pfm, exr, or 16-ppm.

still no release? doh! but there is now a nice, detailed doc about how to profile your camera for darktable to get rich and precise color rendition out of your images (ever been sad about the flat reds in your expensive 5D?) read how to contribute a color matrix!

no new release today, but some great videos about dt: introductory screencast (english) short review with video tutorial (german) watch alexandre prokoudine at LGM 2010 watch meetthegimp #135, a bit older, but still … not really video, but a nice review (russian)

features, features, features ;) image tagging flexible database queries by exif and custom tags powerful color zones plug-in, to selectively alter only some colors french translation libraw 0.8.5 new light table mode layout: file manager inotify support tiff 8-bit and 16-bit export new, more powerful widgets to replace sliders and spin buttons (derived from libphat) reworked, fully color managed processing pipeline supplied some useful base curves image cache can now enforce a strict memory limit snapshots: side-by-side comparisons in darkroom mode fixed aspect ratio and horizontal/vertical flipping in crop module spot auto exposure highlight reconstruction white balance presets with fine-tuning (from ufraw) … countless small improvements

as usual, the new release comes with lots of new features: lens defect correction (lensfun) internationalization (russian and swedish translation) backup files for image editing history and tagging (.dt and .dttags) apart from database gconf preferences handling lighttable mode modules libraw 0.8.4 and low-level option interface from dt color filter-based monochrome conversion dt now reads _ICC_PROFILE atom from X server more sophisticated white balance core database layout prepared for tagging system (to come in 0.

there have been some major internal changes in dt since 0.2, and some of them result in cool new features for the user, so it is time to pass it on to the non-git audience. this includes: most processing is now being done in a new color space (L a/L b/L). this results in nicer exposure/tonecurve/denoise/color correction results. the pixel pipeline is now free to change dimensions of the image, which makes a crop/rotate operation (and lensfun in the future) possible.

darktable moved from svn to git. get it from git://darktable.git.sf.net/gitroot/darktable/darktable. the svn repository is still around, but i will probably remove it in the next couple of weeks. we also have nicer forums now.

in the hope to start a small community and to avoid duplicate email traffic, i created this mailing list. also note that there are the forums.

there have been some major changes in the last months. first, the processing backend has been replaced completely, based on an interface which is able to use libgegl (but currently doesn’t, until gegl is fast enough). all operations are encapsulated in run-time loaded plug-ins. raw reading is now based on libraw-0.8. the lighttable got a slightly different look, and more image operations have been implemented (e.g. luma/chroma denoising). this release is still marked beta, which should indicate that not all features are yet in (especially some lighttable related tasks such as filtering, sorting etc.

darktable is going online … so this is the very first version of this page, be patient with it.