Re: [darktable-user] dt 4.4 on Debian 11.7, apt update issue
On 6/30/23 08:46, Bernhard wrote: Remco Viëtor schrieb am 30.06.23 um 11:41: (Even fairly recent distributions can run into problems: there's no build for OpenSuse LEAP 15.4 or 15.5, due to an issue with OpenEXR, I disabled openEXR for my private build, but that's not a good option for a repository, IMO) there even is not a build for current debian 12 stable ... https://software.opensuse.org/download.html?project=graphics:darktable=darktable only testing and sid ... Until repository is updated, binary package for Testing from the above location works fine on stable Debian 12: $ cat /etc/issue Debian GNU/Linux 12 \n \l $ darktable --version this is darktable 4.4.0 ... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] obs has moved or something
On 6/26/23 08:27, Michael wrote: I went here <https://www.darktable.org/install/> and then to the OBS page to make sure I had the correct source. But instead of getting the options of operating systems I got this <https://build.opensuse.org/> Instead of clicking on “OBS”, just look several lines below for “The available packages are:” -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] Only 1 CPU (or better core) is used during export
On jeudi 4 mai 2023 08:11:41 CEST you wrote: Remco Viëtor schrieb am 04.05.23 um 07:27: On mercredi 3 mai 2023 22:39:06 CEST Bernhard wrote: Remco Viëtor schrieb am 03.05.23 um 17:57: (...) I see 4 cores running at 100% during export on my system ... Are you using openCL? no, not available on my machine since darktable does not support Intel GPU and with external GPU opencl always crashes after suspend-to-ram. In my experience, Intel GPUs + OpenCL + darktable work very well. It's on Debian/Ubuntu. All that needs to be installed is 'intel-opencl-icd'. No vendor repositories or kernel module compile necessary. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] OpenCL not working, Nvidia driver - could not get platforms: -1001
On 5/2/23 10:26, Marco wrote: Hi, I've recently acquired a new GPU and fail to get it to work with darktable. Here are some infos: System: Debian stable 5.10.0-21-amd64 GPU: NVIDIA GeForce GTX 1650 I installed the driver using apt install nvidia-driver as suggested by nvidia-detect. The modules are loaded: lsmod | grep nvidia nvidia_drm 73728 4 drm_kms_helper278528 1 nvidia_drm nvidia_modeset 1204224 8 nvidia_drm nvidia 35528704 367 nvidia_modeset drm 626688 8 drm_kms_helper,nvidia,nvidia_drm But darktable doesn't use the GPU: darktable-cltest 0.120796 [opencl_init] opencl related configuration options: 0.120811 [opencl_init] 0.120813 [opencl_init] opencl: 1 0.120814 [opencl_init] opencl_scheduling_profile: 'default' 0.120816 [opencl_init] opencl_library: '' 0.120817 [opencl_init] opencl_memory_requirement: 768 0.120819 [opencl_init] opencl_memory_headroom: 300 0.120820 [opencl_init] opencl_device_priority: '*/!0,*/*/*' 0.120822 [opencl_init] opencl_mandatory_timeout: 200 0.120823 [opencl_init] opencl_size_roundup: 16 0.120824 [opencl_init] opencl_async_pixelpipe: 0 0.120825 [opencl_init] opencl_synch_cache: false 0.120827 [opencl_init] opencl_number_event_handles: 25 0.120828 [opencl_init] opencl_micro_nap: 1000 0.120829 [opencl_init] opencl_use_pinned_memory: 0 0.120830 [opencl_init] opencl_use_cpu_devices: 0 0.120831 [opencl_init] opencl_avoid_atomics: 0 0.120832 [opencl_init] 0.123537 [opencl_init] could not find opencl runtime library 'libOpenCL' 0.124051 [opencl_init] could not find opencl runtime library 'libOpenCL.so' libOpenCL.so is provided by ocl-icd-libopencl1 package. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
[darktable-user] Re: [darktable-user] darktable 4.2.0 “Lens Correction” how it works
On 23/12/2022 18.46, Terry Pinfold wrote: I would just like to thank the developers for another great development of DT. After reading this post I see more capabilities in the lens correction module than I realised existed. It is time to read that section of the manual again. I also note that some cameras only have the option of lensfun while others have the choice of lensfun or embedded. Which would normally be the best option or does it vary between images? Some cameras embed lens correction data supplied by lens manufacturer. For some from those cameras, that data has been deciphered and is offered as an option in lens correction module. How does a particular manufacturer come up with that data we can only guess. It is what is used by creating in-camera JPEGs. Traditional option in lens correction module uses data from Lensfun database. That data is a collection of empirical distortion measurements from actual lens samples (recently some data was also added by extracting embedded correction data from RAW). Which one of those correction options is better, is for you to judge. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] advice asked
On 10/5/22 05:52, Ruud Baart wrote: [...] darktable-cli xxx.nef yyy.xmp xxx.jpg The yyy.xmp file have been created by developing one photo in darktable (gui version of course) and saving the settings in yyy.xmp . However, there are so many possibilities and conditions (night, fog, rain, snow, sun to name a few) that I wonder if this is the right way. Is there anyone who has experience with something similar and/or can assess my method ? One way would be to query some meteo web service for conditions during the time an image is taken. Then apply [snow,fog,rain,clear,...].xmp accordingly :) I'm using api.openweathermap.org in a Munin plugin for local weather. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] Export xmp metadata with correct lang
On 9/29/22 13:13, Volker Lenhardt wrote: But I'm sorry, but to include iptc tags doesn't seem to have any effect on the xmp tags. I do get xmp.dc.title and xmp.dc.description, but gimp shows an empty string whenever there is a special character within. I'm not sure if it is a gimp problem, but I cannot find the trick to tell it to use utf-8 with metadata. I hope that dt exports them as utf-8. You can check exported images in Finder - Command I (File>get info). In my quick test on Monterey it does show accented characters in Title and Description, which where entered dt. Gimp does show them as well. macOS is the default US English install, no International settings. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] Memory clock speed, any impact on darktable?
On 25/09/2021 03.02, Michael Staats wrote: Question: Is it worth to play around with some settings between 2400 MHz and 3000 MHz? Is there any downside if, say, 2800 MHz works? And most of all: Does it make any difference for darktable? Darktable is the only program which challenges my hardware, I'm not a gamer or so. Faster memory will make image processing quicker. If 3GHz doesn't work, you can try lower frequencies until one that works. Here is one example of processing the same image in darktable on the same system, but with different CPU and memory “speeds”: Ryzen 7 2700X @3.7GHz, RAM @2.133GHz — 15.1s Ryzen 7 2700X @4.0GHz, RAM @3GHz — 11.8s As already mentioned however, the best improvement would come from using OpenCL in GPU. Here is the same image processed on the same system, but with OpenCL in Radeon Vega: Radeon RX Vega64 8GB — 2.7s -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
[darktable-user] Re: [darktable-dev] News about darktable 3.6
Such a waste of time and effort. Make `darktable` a symlink to `darktable-cli`. Done. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] exiv2, darktable, AVIF, metadata, and ExifTool
On 2/5/21 3:26 PM, Top Rock Photography wrote: Thank you, Patrick. That made me investigate. There exists an AVIF plugin for QT5. This plugin requires ECM version 5.70.0 or greater. Ubuntu LTS (20.04) has version 5.68.0. 20.10 has 5.74. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
[darktable-user] Iris Xe shows some OpenCL oomph
Hi, I just tested darktable 3.4 on new (to me) 11th gen. Core i5 and with Intel's Gen12LP (LP: low power) GPU. The latter seems to cut processing time by >50%, from 12s to 5s. This is on Dell XPS 13 Dev. Ed. Ubuntu laptop. All open source. More details on https://math.dartmouth.edu/~sarunas/darktable_bench.html -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] exiv2, darktable, AVIF, metadata, and ExifTool
On 2/4/21 4:13 PM, Top Rock Photography wrote: ... So, thank you, darktable developers, for giving me my EXIF data in my AVIF (and, of course, WebP) files. Now if I can only find a viewer —other than Chrome, Chromium, & Firefox— which can view an AVIF, and also, which can display its metadata. ;-) Geeqie? -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] optimization of opencl performance for NVIDIA GTX .... : settings in file darktablerc
On 11/20/20 2:57 PM, Marc Cabuy wrote: > Indeed Patrick, > > the default in the current version is: > opencl_synch_cache=active module > > The option I would advise to test for response time comfort during finetuning > in a 2nd pass is: > opencl_synch_cache=true. > > That is if it does not compromise too hard overall response time. Which is ok > on my system/configuration. Whatever optimizations, it would be nice if it was mentioned how they differ from darktable defaults. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
[darktable-user] Ubuntu 20.04, OpenCL, DT3.2.1, Radeon HD5400
On 11/15/20 3:28 PM, GianLuca Sarto wrote: > thanks, Šarūnas, > > I have two Darktable systems, based on Lenovo TS140, one with AMD, the > other with Nvidia Quadro 400. > Neither of the two manage OpenCL.. > > 0.039390 [opencl_init] found opencl runtime library 'libOpenCL' > 0.039412 [opencl_init] opencl library 'libOpenCL' found on your system > and loaded > 0.042560 [opencl_init] found 1 platform > 0.042582 [opencl_init] found 1 device > 0.042727 [opencl_init] device 0 `Quadro 400' has sm_20 support. > 0.042778 [opencl_init] discarding device 0 `Quadro 400' due to > insufficient global memory (511MB). Too small video memory for useful OpenCL. > I would like to upgrade one of the two systems to a 4K display, so was > already decided to purchase a new video card. > > Is there a tested solution that works out of the box, or a list of DT > OpenCL compliant cards for Linux? I don't know whether such a list exists[1]. The more memory and more parallel processing units (“GPU cores”) a video card has, the more useful it will be for OpenCL processing. The useful minimum these days might be 2GB, but I would look for 4GB and more. “Out of the box” would probably only happen if you buy a computer from a company that sells them configured with Linux. Whether you install AMD or Nvidia GPU, there will be additional steps. In case of AMD, Linux kernel already supports AMD cards with the open source ‘amdgpu’ module, so that part will be “out of the box”. OpenCL support will have to come from either 1) open source ROCm or 2) proprietary AMDGPU-PRO. In case of Nvidia, Linux kernel's ‘nouveau’ module will need to be replaced with the proprietary ‘nv’ one from Nvidia, plus OpenCL part from the same Nvidia. Ubuntu has them in standard repositories. One can also use Nvidia repositories for perhaps slightly newer software. Intel appears to have a completely open source system, but usable GPUs are still to come. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] Ubuntu 20.04, OpenCL, DT3.2.1, Radeon HD5400
On 11/14/20 3:41 PM, GianLuca Sarto wrote: > Hello All, > > DT 3.2.1 here, running on Ubuntu 20.04, Radeon HD5400. > > OpenCL should be working, however DT complains "could not get platforms" > (see below). Support for Radeon HD5xxx may have ended with fglrx (Ubuntu 16.04?). You may check what is currently supported, by proprietary AMDGPU-PRO: https://www.amd.com/en/support by open source ROCm: https://github.com/RadeonOpenCompute/ROCm#Hardware-and-Software-Support -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · OpenPGP_signature Description: OpenPGP digital signature
Re: [darktable-user] Graphic card opensource drivers
On 10/2/20 1:57 AM, Marco wrote: > On Thu, 1 Oct 2020 13:58:38 +0100 > Robert Bridge wrote: > > Thank you and the others for the responses. > > I figure that the ROCm driver is the only opensource solution that > provides OpenCL functionality. But it's fiddly to install and might > not work under Debian at all. The only other options are amdgpu > (opensource but OpenCL not working) or amdgpu-pro (proprietary but > OpenCL works). Yet another option is combination of the latter two. You can use open source amdgpu (Linux kernel module) with OpenCL libraries from AMDGPU-PRO (formerly “compute”, now “headless” part): amdgpu-pro-install --opencl=pal --headless --no-dkms -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Graphic card opensource drivers
On 10/1/20 8:06 AM, Marco wrote: > Hi, > > this is a very similar question than the one asked yesterday. But the > focus is on opensource drivers. > > I need a new graphic card. So I searched the manual and mailing list > what are the options, but I'm still unsure. A lot of information is or > might be outdated. Although this questions has been asked a dozen times > already, I couldn't find a definite recent answer. > > My needs > > - opensource drivers (if at all possible) > - should work out-of-the-box with no (or little) fiddling > > My system: > > - Debian stable (might be able to upgrade to testing, if required) > - currently kernel 4.19 AMD ROCm [1] is open source. It is probably the only such. I have just tried installing [2] it on Debian stable and current testing and unfortunately it fails with unsatisfied dependencies. I didn't try anything beyond `apt install ...`. 1. https://github.com/RadeonOpenCompute/ROCm 2. https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] could someone direct me....
On 9/5/20 9:29 AM, Timur Irikovich Davletshin wrote: > Better not to mess with updating system-wide database via sudo lensfun- > update-data. Why? Over years, I have never got into any mess with system-wide update. > On Sat, 2020-09-05 at 09:13 -0400, Šarūnas wrote: >> sudo lensfun-update-data signature.asc Description: OpenPGP digital signature
Re: [darktable-user] could someone direct me....
On 9/5/20 12:00 AM, Michael wrote: > sorry... it is late here and I forgot to say that I run Mint20 and I > don't know what you mean by "the package(s) they provide." I just d/l > version 0.3.95.tar.gz if that is what you mean. To install lensfun in Mint, in the terminal type: sudo apt update sudo apt install liblensfun1 liblensfun-bin sudo lensfun-update-data You can also use on of the graphical software managers — Synaptic Package Manager, Software Manager (look/search within application menu). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Base curve
On 7/27/20 12:50 PM, Pascal Obry wrote: > Le lundi 27 juillet 2020 à 09:51 -0700, David Vincent-Jones a écrit > : >> After you have read the replies to your initial question, I am >> sure that you must be quite confused. Let me try to more correctly >> find an answer that may make sense. >> >> There are two possible development 'routes' in darktable. 1.) uses >> the 'base curve'; 2.) uses the 'filmic-rgb' module. > > And a third one... Nothing, no base curve, no filmic. I mostly stopped applying base curve long ago, before filmic existed. I don't use filmic either (except for following a tutorial once and several image tests). Typical list, besides always-on modules, is: levels, color zones, local contrast, denoise, sharpening. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Import: I'm failing at step 1; advice please
On 6/25/20 1:38 PM, tony Hamilton wrote: > ... > So I failed at the first step in trying to use DT and I don’t yet see an > effective, smooth, low hassle way to get beyond that first step, ... Copy RAW files from cameras or external storage into your computers using OS file managers. Use RAW processing software for processing RAW files into something else. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Side panel too small to select export directory
On 6/3/20 11:23 AM, Al Bogner wrote: > I use darktable 3.0.2-1.1 with XUbuntu 20.04 and a 32° UHD display. > It is not so easy to get an usable fontsize for *all* programs. At > the moment the fonts are too big with dt and default values, but this > fits for most other programs. You can experiment with the value of screen_dpi_overwrite= in ~/.config/darktable/darktablerc I found ‘130’ works well for 24" 4K display. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] darktable & lensfun
In Ubuntu, standard distribution darktable package depends on lensfun (see below). Of course, when installing from somewhere else or compiling from source, dependencies are up to the person installing… > apt show darktable > Package: darktable > Version: 1:2.6.2+13~g0bef6bb76-0pmjdebruijn1~disco > Priority: optional > Section: graphics > Maintainer: Pascal de Bruijn > Installed-Size: 17.1 MB > Depends: libc6 (>= 2.29), libcairo2 (>= 1.14.0), libcolord-gtk1 (>= 0.1.20), > libcolord2 (>= 1.4.3), libcups2 (>= 1.7.0), libcurl3-gnutls (>= 7.56.1), > libexiv2-14 (>= 0.25), libflickcurl0, libgcc1 (>= 1:4.0), libgdk-pixbuf2.0-0 > (>= 2.22.0), libglib2.0-0 (>= 2.39.4), libgomp1 (>= 4.9), libgphoto2-6 (>= > 2.5.10), libgphoto2-port12 (>= 2.5.10), libgraphicsmagick-q16-3 (>= 1.3.5), > libgtk-3-0 (>= 3.21.5), libilmbase23 (>= 2.2.0), libjpeg8 (>= 8c), > libjson-glib-1.0-0 (>= 0.13.2), liblcms2-2 (>= 2.2+git20110628), liblensfun1 > (>= 0.3.2), liblua5.3-0, libopenexr23, libopenjp2-7 (>= 2.0.0), > libosmgpsmap-1.0-1 (>= 1.1.0), libpango-1.0-0 (>= 1.14.0), > libpangocairo-1.0-0 (>= 1.14.0), libpng16-16 (>= 1.6.2-1), libpugixml1v5 (>= > 1.6), librsvg2-2 (>= 2.14.4), libsecret-1-0 (>= 0.7), libsoup2.4-1 (>= > 2.47.4), libsqlite3-0 (>= 3.6.0), libstdc++6 (>= 5.2), libtiff5 (>= 4.0.3), > libwebp6 (>= 0.5.1), libx11-6, libxml2 (>= 2.7.4), libxrandr2 (>= > 2:1.2.99.3), zlib1g (>= 1:1.2.0) > Recommends: darktable-dbg, gvfs darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Photos with default settings for Canon M3 are very dull
>> Am 01.03.2020 um 10:26 schrieb openhab.doc: >>> Since darktable 3.0 I am no longer satisfied with the default settings for >>> my Canon M3. The colors of the photos with the default settings of >>> darkroom >>> are very dull compared to JPG images or JPG previews. With darktable 2.6 >>> the difference was less. Darktable 3.2 will probably have a Legacy mode, which would keep the module processing order in the same sequence as 2.6 for images already processed in 2.6. You may compile 3.1 from source, which already has the Legacy ordering option. I don't know if it will made available in 3.0.x patch releases. In my case, upgrading from 2.6 to from 3.1 with the new iop_order patch, keeps the looks and therefore preserves the work already done on thousands of images in 2.6, now “legacy” mode... For me, 3.1 with the new iop_order is actually the first usable darktable 3. Thanks to those who came up and worked on this non-trivial change. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Fontsize with darktable 3.0.0~git325.4538201be
On 2/1/20 3:02 PM, Christian wrote: > Hi, custom DPI is scaling all GUI elements, while CSS font-size > mainly affects the font size. > > What's better? Don't know. Matter of taste... :-) Yes, of course. I tried 3.x only briefly and all interface elements and fonts look much too bulky for my taste. Using screen_dpi_overwrite I was able to bring it back to a more economic 2.x looks, without bothering with fonts. Šarūnas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Opencl support for DarkTable
On 1/31/20 8:21 PM, frieder wrote: > Am Donnerstag, den 30.01.2020, 22:04 + schrieb Alex: >> I'm using Ubuntu 18.04 and I intend to install that on the new PC and >> wait until 20.04 has been out a while before upgrading. Although if I >> have to use another distro I'll consider it. > When you use Ubuntu 18.04 you can consider the AMD Graphic for > darktable. Darktable needs extended open-cl support which is not part of the > open source amdgpu- driver package. It runs only with the gpupro-driver > which is accessable via amd-webiste: > https://www.amd.com/en/support/kb/release-notes/rn-prorad-lin-18-20 > > But only installs with the Distribution it is inteded for. I'm using it > with a RX 560 card and running fine, albeit nowadays considered rather > low end. > regards, > F. In my experience, open-source amdgpu kernel module and ROCm userspace (including OpenCL) work in any Linux from kernel from 4.18 to 5.3. I haven't noticed anything missing regarding OpenCL/darktable operation. Šarūnas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Opencl support for DarkTable
On 1/30/20 4:45 PM, Alex wrote: > It looks like the GTX-1660 is the one to go for, as it seems popular, > supported and relatively inexpensive! > > Specifically this one: Gigabyte GeForce GTX 1660 SUPER GAMING OC 6G. Note it's PCIe 3 (you seem to be building a PCIe Gen4 system). Not sure however if that will affect OpenCL performance, haven't had a chance to compare... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Monitor
On 1/18/20 6:53 PM, dt-l...@stefan-klinger.de wrote: > Jesus Arocho (2020-Jan-10, excerpt): >> I plan to buy a new monitor > ... > In 2017 I've directly contacted Eizo about hardware calibration under > Linux, see [1]. I'm still looking for a new monitor, but doubt it > will be Eizo. NEC SpectraView software works fine on Linux. Šarūnas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Base curve and white balance issue with DT 3.0
On 1/14/20 7:35 AM, Viktors Krasovskis wrote: > ... > Darktable 3.0 would be a fantastic tool if it would allow the users to > choose between the classic and advanced workflows. That would speed up > editing where camera manufacturer raw interpretation worked well and for > more complicated cases filmic RGB module would be a nice help for > experienced users. Most users in most cases need that their raw files > are rendered well enough and the rendering should be predictive. Only > geeks and some others (the minority) needs the ultimate raw rendering > which takes more time and in practice is less universal. Filmic RGB is > great for landscapes and architecture where the camera's dynamic range > is on it's limits, however Nikon base curve presents give far better > results on human portraits on DT 2.6.x by default. What do others think > and do they have similar experience? I hear what you are saying. I use Olympus cameras since one of their first digital SLR, E-500, and like Olympus's out of camera (OOC) color rendition very much. It would be my preferred starting point while working on RAW. However, none of the RAW conversion software (except Olympus's own) comes close to OOC colors in some universal/preset way. Base curves for Olympus didn't do that either, and I stopped using them in darktable 2.x a while ago. The closest I was able to get to OOC in some automated way was by creating a preset with a camera profile, produced by photographing IT8 color target. Here is the latest write-up on the technique by Andreas Schneider: https://pixls.us/articles/profiling-a-camera-with-darktable-chart The resulting preset worked quite well for some exposures/lighting conditions, but for quite a few it didn't, so in the end I gave up on that either. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] opensuse repo for Ubuntu 18.04 inconsistent contents (was: Looking for Darktable 3.0)
On 1/13/20 2:15 PM, Michael Staats wrote: > The repository at > > http://download.opensuse.org/repositories/graphics:/darktable/xUbuntu_18.04/ > to be more precise, the deb package > > http://download.opensuse.org/repositories/graphics:/darktable/xUbuntu_18.04/amd64/darktable_2.6.2-1.1_amd64.deb > > contains a darktable 3.0.0, although it's called 2.6.2. > > Is my observation correct, or have I completely blown a fuse? You are correct. After installing the above darktable_2.6.2-1.1_amd64.deb: $ apt policy darktable darktable: Installed: 2.6.2-1.1 Candidate: 1:2.6.2+13~g0bef6bb76-0pmjdebruijn1~disco Version table: 1:2.6.2+13~g0bef6bb76-0pmjdebruijn1~disco 500 500 http://ppa.launchpad.net/pmjdebruijn/darktable-unstable/ubuntu disco/main amd64 Packages 3.1.0~git353.af569327e 500 500 http://download.opensuse.org/repositories/graphics:/darktable:/master/xUbuntu_19.10 Packages *** 2.6.2-1.1 100 ... $ which darktable /usr/bin/darktable $ darktable --version this is darktable 3.0.0 ... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Using a 4K monitor display
On 11/26/19 10:51 PM, David Vincent-Jones wrote: > looking in darktable.css I note that *font-size: 1em .* > > does this scale with the screen resolution?* Yes, it does, at least in KDE Plasma. Don't have Gnome to test at the moment. Not with screen resolution, but with pixel density. Though 1em, or whatever is the default in 3.x, (there are numerous font-size settings in darktable.css) is a bit too large for my taste. Other UI elements seem a bit bulky as well... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Using a 4K monitor display
On 11/26/19 7:04 PM, David Vincent-Jones wrote: > Is there an established reference on setting up dt with a 4k display. > Where to scale fonts and other needed changes? In my experience, on Linux desktops, darktable responds correctly to whatever scaling is necessary (and can be configured in window manager's settings) fo high-DPI screens. That is, it's automatic and I never had to do anything specific for darktable, except for maybe adjusting the width of side panels in darktable preferences → GUI options. Šarūnas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] OpenCL Manjaro
On 11/25/19 4:34 PM, darkta...@911networks.com wrote: > Manjaro & Darktable 2.6.3 > > I'd like to "upgrade" to opencl. I have a Radeon R570 with 8Gb. > Arch has a "whole bunch" of opencl. Which opencl to use? > > * opencl-mesa: free runtime for AMDGPU and Radeon > * opencl-amd: proprietary standalone runtime for AMDGPU (pal and > legacy stacks in a single package) > * rocm-opencl-runtime: Part of AMD's fully open-source ROCm GPU > compute stack, which supports GFX8 and later cards(Fiji, Polaris, > Vega) > * opencl-amdgpu-pro-orca: proprietary runtime for AMDGPU PRO > (supports legacy products older than Vega 10) > * opencl-amdgpu-pro-pal: proprietary runtime for AMDGPU PRO (supports >Vega 10 and later products) I would start with ROCm. I'm not very familiar with Manjaro and AUR packaging, but installing rocm-opencl-runtime will likely pull in more ‘stuff’. If using proprietary AMDGPU-PRO, your RX570 should work the ‘legacy’ option (or just use opencl-amd). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Artifacts in exported image
On 11/26/19 1:18 AM, Niranjan Rao wrote: > Ubuntu 18.04, darktable 2.6.2 > > More and more research I do, I am getting convinced that this is a > problem with one of the updates of darktable or somehow my settings are > screwed up. > > I have photos from same raw exported at different times. Please check > the photos of two giraffes at > https://photos.app.goo.gl/K35BT8X5HGhSzwQj6. One was exported back in > August. I don't know if I got a darktable update since then or not. > Exported same photo with same style today and it shows artifacts. None > of the older images show these geometric patterns and same image > exported today shows it, no style changes since then. Most likely a result of overly strong compression on export. To get your good.jpg to 260kB I had to use quality=10 in Gimp. The reslt is as bad as your bad.jpg. This of course does not explain why out of a sudden such a compression occurs in your exports... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] What could be wrong?
On 10/31/19 11:42 PM, Niranjan Rao wrote: > Darktable 2.6.2 on Ubuntu 18.04 > > Chromium 77.0.3865.90 (Official Build) Built on Ubuntu , running on > Ubuntu 18.04 (64-bit) > > I'm facing a interesting challenge and don't have any solution yet or > what exactly could be wrong. I had 4 raw files with some minor > processing done and exported. Every time I try to upload one particular > file, chromium crashes. Same style applied to 4 images, 3 get uploaded > fine and last one causes the crash. > > Every image viewer I had on machine, including firefox, chromium, > chrome works fine. Its only during uploading with one particular image > browser crashes. I have tried uploading this particular file only and it > sill crashes. As I don't like sending attachments to this list, it's a > dilemma. > > I have tried rexporting the file again - no luck. This file is smallest > of all other files, so I think I can rule out the size factor. > > Just wondering if anyone has encountered similar situations? I certainly > have not encountered such situation. I don't want to try uploading file > against any other browser right now for multiple reasons which has to do > with carefully setup profiles in each browser. Too many variables. Move your carefully setup profiles aside and start browsers anew, with their default profiles, no custom settings, no extensions. Try uploading your file, while viewing the process in browsers' developer tools/consoles. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] OpenCL questions
On 10/13/19 4:41 AM, Peter McD wrote: > Am 08.10.19 um 19:55 schrieb darkta...@911networks.com: >> DT 2.6.2 on Manjaro >> >> How do I install OpenCL. I have amdgpu (the open source). My hardware >> doesn't qualify for rocm2. I have an FX-8350 (8-cores). >> >> Has somebody already done it? Any good write up? >> >> When I was searching, everything pointed to amdgpu-pro but it's for >> ubuntu/debian, centos and sles not for arch/majaro I can confirm, that amdgpu-pro's opencl part works (clinfo, darktable-cltest) fine on Manjaro, with AUR enabled. Latest download and fresh install of Manjaro Gnome, Linux 5.2, then: pamac build opencl-amd > To add to that question. > I use amdgpu-pro, the openCL part with openSUSE Leap 15.1. > > It works, but I would prefer the OpenSource openCL. > > Anyone out there who uses the OpenSource openCl with Darktable? Yes, ROCm with Ubuntu (github.com/RadeonOpenCompute/ROCm). Can be built from source too. Šarūnas math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] GFX card recommendations
On 9/13/19 8:40 AM, Christian von Kietzell wrote: > Hi everyone, > > I'm currently thinking about putting together a new (Linux) PC with > special attention to being a workhorse for darktable. > > A fast SSD and enough RAM being a given, I'm not sure which graphics > card to get. What's the current thinking on what to get? > > Specifically, I have the following questions: > > * Which works better in Linux - Nvidia or AMD? Both work well. In case of AMD there is an open source option. > * Which drivers work (best) for using OpenCL in darktable? Are the > open-source drivers okay or do I need the binary drivers from AMD or > Nvidia? Open source drivers are okay (ROCm for AMD), including upstream kernels. Just tested with 5.3 RC8. > * Is it worth getting a high-end card (say 5700 XT for AMD or 2080 > SUPER for Nvidia)? IMO, yes. But still I would make CPU, RAM and something like PCIe 4 chipset+NVMe storage a priority. > * Are the performance differences in darktable noticable, compared to > a card that is one or two steps down from that - especially when you > consider the rather larger amount of money I'd have to pay for it? Yes, it is noticeable. Can't judge money amounts... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Tamron SP 70-200 f2.8 DI VC USD aka A009
On 9/2/19 11:19 AM, Ralf Saalmüller wrote: > Hello and many thanks to the darktable team, > > I use darktable on at least 2 GNU/Linux computers, with devuan ascii, > Ubuntu 18.04 & 19.04 and debian buster. On debian buster already > darktable 2.7.0+1662 > > Since I've got the lens Tamron SP 70-200 f2.8 DI VC model number A009 > there was a problem with darktable (dt). dt doesn't recognize the > Tamron. I have to manually select the Lens Model from the (updated) > lensfun database. > > The exiftool does give me the following information from a picture made > with the Tamron lens. I only included the output with the keyword "Lens": > > exiftool 613A5345.CR2 | grep Lens [...] It's exiv2 that is used by lensfun/darktable. exiv2 -pt 613A5345.CR2 | grep -i lens would show what exiv2 “knows”. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas · https://useplaintext.email · signature.asc Description: OpenPGP digital signature
Re: [darktable-user] OpenCL error starting dt
On 8/17/19 12:43 PM, David Vincent-Jones wrote: > if the devs can do better with the std cpu than can be achieved with > opencl that really simplifies life on all types of systems. > Incredible! Maybe it does, maybe not. https://www.math.dartmouth.edu/~sarunas/darktable_bench.html Šarūnas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Help updating lensfun?
On 8/14/19 8:05 PM, Richard Trinkner wrote: > HI folks, > > I'm using darkable 2.6.2 on Ubuntu Linux 19.04. > > My lens (NIkon 200-500mm f/5.6 ER VR) does not appear in the lens > correction choices. > > I downloaded the latest master branch of lensfun from github, > compiled it and installed it, apparently successfully. > > The new version of lensfun does in fact have correction data for my > lens. It's been installed to > /usr/local/share/lensfun/version_2/slr-nikon.xml. > > Nevertheless, when I restart darktable, the lens choice list does > not include my lens. To get newly added lens corrections it suffices to update just the lensfun database. It can be done by running `lensfun-update-data` utility (part of liblensfun-bin package). This will bring in corrections for your 200-500mm Nikkor. As for the latest lensfun compiled from source, you may find that darktable is complaining about lensfun database version (if you start darktable from command line, for example). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Sharpening
On 8/6/19 10:51 AM, darkta...@911networks.com wrote: > ... > Any suggestion for good sharpening tutorials? I keep this in bookmarks: https://redskiesatnight.com/2005/04/06/sharpening-using-image-magick -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] opencl
On 8/5/19 4:49 PM, darkta...@911networks.com wrote: > On Mon, 5 Aug 2019 15:25:22 -0400 > Šarūnas wrote: >> darktable-cltest ? >> > > It's the same: > >> darktable-cltest > > 0.025889 [opencl_init] found opencl runtime library 'libOpenCL.so.1' > 0.025912 [opencl_init] opencl library 'libOpenCL.so.1' found on your system > and loaded > 0.098914 [opencl_init] found 1 platform > 0.098938 [opencl_init] found 1 device > 0.098966 [opencl_init] discarding device 0 `Radeon RX 570 Series (POLARIS10, > DRM 3.27.0, 4.12.14-lp151.28.10-default, LLVM 7.0.1)' due to missing image > support. 'missing image support' is key. You probably won't get that support with Mesa. AMDGPU-PRO [1] or ROCm [2], not sure what's available for Linux kernel 4.12 though. Šarūnas 1. https://www.amd.com/en/support 2. https://github.com/RadeonOpenCompute/ROCm signature.asc Description: OpenPGP digital signature
Re: [darktable-user] opencl
On 8/5/19 3:11 PM, darkta...@911networks.com wrote: > DT 2.6.2 on opensuse leap 15.1 > > I can't get opencl working. I must be missing something but I don't know > what... > > > $ darktable -d opencl > > 0.065273 [opencl_init] opencl related configuration options: > 0.065296 [opencl_init] > 0.065299 [opencl_init] opencl: 1 > 0.065303 [opencl_init] opencl_library: '' > 0.065306 [opencl_init] opencl_memory_requirement: 768 > 0.065309 [opencl_init] opencl_memory_headroom: 300 > 0.065313 [opencl_init] opencl_device_priority: '*/!0,*/*/*' > 0.065316 [opencl_init] opencl_mandatory_timeout: 200 > 0.065320 [opencl_init] opencl_size_roundup: 16 > 0.065323 [opencl_init] opencl_async_pixelpipe: 0 > 0.065326 [opencl_init] opencl_synch_cache: 0 > 0.065329 [opencl_init] opencl_number_event_handles: 25 > 0.065335 [opencl_init] opencl_micro_nap: 1000 > 0.065337 [opencl_init] opencl_use_pinned_memory: 0 > 0.065340 [opencl_init] opencl_use_cpu_devices: 0 > 0.065343 [opencl_init] opencl_avoid_atomics: 0 > 0.065348 [opencl_init] > 0.065488 [opencl_init] could not find opencl runtime library 'libOpenCL' > 0.065530 [opencl_init] could not find opencl runtime library 'libOpenCL.so' > 0.065640 [opencl_init] found opencl runtime library 'libOpenCL.so.1' > 0.065668 [opencl_init] opencl library 'libOpenCL.so.1' found on your system > and loaded > [...] darktable-cltest ? -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] What about adding more IPTC fields to the metadata editor?
On 5/13/19 5:34 AM, dt-l...@stefan-klinger.de wrote: > tl;dr: If it can be done with 3rd party software, you should like it. > > Hi, sorry, I cannot solve your problem, since I do almost no metadata > editing. But I'd like to comment on one important misconception: > > Kneops (2019-May-13, excerpt): >> I really would like to be able to do this in one program (DT) > > This desire, voiced by many users and not limited to DT, has led the > IT world to bloated, unmaintainable and inflexible monolithic > applications, much pain and frustration. > > Building an application that does everything you need will lead to > feature creep [1] inevitably: Other people probably want other > features which must be added as well (unless you're special, e.g. the > maintainer or the sole paying customer). > > Applications implementing a certain feature are unlikely to be > flexible enough to hand this task off to an application which is > better at it. So if your does-all-application A also solves a minor > side task X, you won't be able to use another application B, which > specializes on X and solves it better. You're locked into using A > because the other things it does, you're bound to the crappy solution > A provides. > > The Unix Philosophy [2] and KISS Principle [3], while debatable in > their most extremist interpretations, lead a way out of this: Do not > add features to a program that could be easily handed of to an > external tool. > [...] Stefan, I agree with your points in general, but I don't see adding fuller support for metadata in dt as a feature bloat. To me it looks like a rather basic and essential feature, part of the of good digital photography practice... I would be glad to be able to code what's needed, but unfortunate that's not likely anytime soon... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Error when exporting
On 4/19/19 12:09 PM, Christian Gruber wrote:
> Yes and no. When I start darktable on the command line, it also only
> says this, only in other words.
Then there is `darktable -d` for more info, in case you haven't tried yet:
-d {all,cache,camctl,camsupport,control,dev,fswatch,input,lighttable,
lua, masks,memory,nan,opencl,perf,pwstorage,print,sql}
> Am 19.04.19 um 14:22 schrieb Šarūnas:
>> On 4/19/19 5:39 AM, Christian Gruber wrote:
>>> Hello!
>>>
>>> When I try to export a certain picture, darktable prints a small
>>> window that it could not export this file to
>>> darktable_exported/bla.jpg. Other files work. How can this
>>> failure be debugged? I already see that this file is not
>>> exported, so this error message gives me no new information.
>>> *Why* it can't write this file would be a very useful
>>> information. By the way, I can touch this file, fill it with
>>> content, and there are 26GB free in this partition. I have also
>>> exported another picture, and it worked without a problem.
>>
>> Perhaps if you start as `darktable` from the terminal window,
>> you'll see more info there...
signature.asc
Description: OpenPGP digital signature
Re: [darktable-user] Error when exporting
On 4/19/19 5:39 AM, Christian Gruber wrote: > Hello! > > When I try to export a certain picture, darktable prints a small window > that it could not export this file to darktable_exported/bla.jpg. Other > files work. > How can this failure be debugged? I already see that this file is not > exported, so this error message gives me no new information. *Why* it > can't write this file would be a very useful information. > By the way, I can touch this file, fill it with content, and there are > 26GB free in this partition. I have also exported another picture, and > it worked without a problem. Perhaps if you start as `darktable` from the terminal window, you'll see more info there... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] OT: Any Hugin experts here?
On 3/5/19 6:36 AM, Bruce Williams wrote: > Hey guys, > Just wondering if anyone on this list is a gun at Hugin pano stitching? > I've got a sequence I shot about 10 years ago (on a tripod, but rushed, > and not framed correctly/straight), and I'd really like to reprocess it, > but it's giving me no end of grief. > My main issue is that Hugin, despite creating control points ok, seems > to fail completely when it comes to aligning the images. > Would love to pick some brains, if anyone is able/willing to help out! :) > I've just zipped the source images up, and thrown them on google drive. > If anyone is interested in having a crack at it, feel free. If you are fighting a wavy horizon, then it may help to check control points. If they were added automatically — remove all that are in the clouds or in near-by points. Add a couple of horizontal lines per image. It may help to straighten the horizon for individual images before using them in Hugin (crop and rotate module in darktable, guides: metering). As well as shadows and highlights for some images. If panorama preview still shows some misalignment, pitch and roll for individual images can be tweaked (in Hugin's Photos tab dbl-click on an image). Here is what I get after some tweaking (.pto attached): https://math.dartmouth.edu/nextcloud/s/xSWpRNEWbJKHsjA Šarūnas Burdulis math.dartmouth.edu/~sarunas PICT2002.pto Description: application/ptoptimizer-script signature.asc Description: OpenPGP digital signature
[darktable-user] Input color profiles
Hi, User Manual says for Input Color Profile module: > 3.4.3.11. Input color profile ... ... ... > PROFILE > Choose the profile > or color matrix to apply, darktable offers many widespread matrices > along with an enhanced matrix for some camera models. The enhanced > matrices were processed by the darktable team in order to provide a > look closer to the manufacturer's. Is the “enhanced matrix” the same thing that is listed in Camera Support as availability of Custom Matrix for some cameras? (The same as included in src/common/colormatrices.c?) Thanks, -- Šarūnas Burdulis math.dartmouth.edu/~sarunas 0x6B76E2B42C2BEBF4.asc Description: application/pgp-keys signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 2018-05-21 18:14, Аl Воgnеr wrote: > I want to create my own corrections as with Ubuntu 16.04 (where > a lot more lenses worked after I created my own corrections with > little changes to existing ones), so I would like to know, how the 1st > lines of a correction file should like like because there was a change > with lensfun 0.3 > > Then I will try it in ~/.local/share/lensfun The root element is still 'lensdatabase', so this should do: (your custom entries here) -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 2018-05-21 15:50, Аl Воgnеr wrote: > Am Mon, 21 May 2018 14:41:41 -0400 > schrieb Šarūnas <saru...@mail.saabnet.com>: >> There is no correction data for Sigma 70-300mm F4-5.6 DG OS on a >> "full-frame" sensor in lensfun (not in github calibration repository). > > Yes, and therefore I added a line: > > > Sigma > Sigma 70-300mm f/4-5.6 DG Macro If I understand this correctly, the matching will still be done for the above line. The only thing that the lang line does is change the display name of the matched lens for given locale... > Sigma 70-300mm F4-5.6 DG OS The only way to have a lens alias for an optically identical one is to duplicate the entire element and then change the in it. If you'd like to have that, you should probably raise an issue on github/lensfun, which is used for the correction database maintenance. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Chromatic aberration on Fuji
On 2018-05-21 14:51, komodo wrote: > Hi, > > is there any option to apply chromatic aberration correction on Fuji > (X Trans) files ? This depends on whether correction data for a given lens (on that particular crop-factor) exists in Lensfun database. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 2018-05-21 14:11, Аl Воgnеr wrote: > ... > If you think, it makes sense, I can purge lensfun and darktable and > install again. That shouldn't hurt, but I don't know why would that affect reading settings in ~/.local. > Do you have files in /var/lib/lensfun-updates/version_1/ or in other > words, did you execute lensfun-update-data as root? Yes, I did and files are there. > Maybe if the update > is done as root, files in ~/.local/share/lensfun are ignored? Should not, unless files in ~/.local/share/lensfun happened somehow to be owned by root (or some other different user) with no read permission for others. > On the other hand, there is a file with cropfactor 1 for this lens in > the lensfun-database, but this is a lensfun-configuration problem IMHO. > The posted configuration works for me, if the other similar lens > corrections are removed. > > I also wonder which lensfun correction is used, if there is a 2nd > correction in the file. I don't see a second entry, only 'alias'. Are you referring to the 'alias' comment in 1.53 cropfactor entry for 'Sigma 70-300mm f/4-5.6 DG OS'?: It is probably true then that the two lenses are equivalent, but unless you select the Macro lens manually, the corrections from its entry won't be applied. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 2018-05-21 08:57, Аl Воgnеr wrote: > Am Sun, 20 May 2018 14:30:09 -0400 > schrieb Šarūnas <saru...@mail.saabnet.com>: > >> On 05/20/2018 01:46 PM, Аl Воgnеr wrote: >>> gimp-lensfun 0.2.4-1 >>> liblensfun-bin 0.3.2-4 ii >>> liblensfun-data-v1 0.3.2-4 ii >>> liblensfun1:amd64 0.3.2-4 ii >>> python3-lensfun 0.3.2-4 apt-cache policy liblensfun-data-v1 liblensfun-data-v1: Installed: 0.3.2-4 Candidate: 0.3.2-4 Version table: *** 0.3.2-4 500 500 http://euler:/archive.ubuntu.com/ubuntu bionic/universe apt-cache policy darktable darktable: Installed: 1:2.4.3-0pmjdebruijn1~bionic Candidate: 1:2.4.3-0pmjdebruijn1~bionic Version table: *** 1:2.4.3-0pmjdebruijn1~bionic 500 500 http://ppa.launchpad.net/pmjdebruijn/darktable-release/ubuntu bionic/main amd64 Packages 100 /var/lib/dpkg/status 1:2.4.3-0pmjdebruijn1~bionic 500 500 http://ppa.launchpad.net/pmjdebruijn/darktable-unstable/ubuntu bionic/main amd64 Packages With the above, on Ubuntu 18.04, lensfun db entries from any xml files in ~/.local/share/lensfun/ are read and corrections do work. >>> I try to get a Sigma 70-300mm F4-5.6 DG OS working with >>> _fullformat_ and darktable 2.4.3. The lens works with APS-C, but >>> not with fullframe after the repo-installation. The correction >>> suggested by me with a comment in the xml-file was used with APS-C >>> (and worked). For test purposes only I took 'Sigma 70-300mm F4-5.6 DG OS' lens entry, changed cropfactor to 1 and put that single-lens xml in ~/.local/share/lensfun/test.xml Everything works fine with the provided DNG, i.e. camera+lens is autodetected and "corrections" can be applied. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 2018-05-21 08:57, Аl Воgnеr wrote: > Am Sun, 20 May 2018 14:30:09 -0400 >>> [...] >>> I try to get a Sigma 70-300mm F4-5.6 DG OS working with >>> _fullformat_ and darktable 2.4.3. The lens works with APS-C, but >>> not with fullframe after the repo-installation. The correction >>> suggested by me with a comment in the xml-file was used with APS-C >>> (and worked). >>> >>> The correction below generally works for APS-C and FF, but it is not >>> recognized in .local-folder. It is a different correction that I >>> suggested. The correction below is for FF. Sorry, I must be blind, I looked up a different lens... There is no correction data for Sigma 70-300mm F4-5.6 DG OS on a "full-frame" sensor in lensfun (not in github calibration repository). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun priority troubles with Ubuntu 18.04
On 05/20/2018 01:46 PM, Аl Воgnеr wrote: > gimp-lensfun 0.2.4-1 > liblensfun-bin 0.3.2-4 ii > liblensfun-data-v1 0.3.2-4 ii > liblensfun1:amd64 0.3.2-4 ii > python3-lensfun 0.3.2-4 > > So everything is 0.3.2-4, except for gimp which is 0.2 > > I use lensfun with darktable 1:2.4.3-0pmjdebruijn1~bionic > > I try to get a Sigma 70-300mm F4-5.6 DG OS working with _fullformat_ and > darktable 2.4.3. The lens works with APS-C, but not with fullframe > after the repo-installation. The correction suggested by me with a > comment in the xml-file was used with APS-C (and worked). > > The correction below generally works for APS-C and FF, but it is not > recognized in .local-folder. It is a different correction that I > suggested. The correction below is for FF. > > > So there are 3 place where a correction can be: > > /usr/share/lensfun/version_1/ > /var/lib/lensfun-updates/version_1/ > ~/.local/share/lensfun > > What I tried with different headers and maybe there is something wrong. > [...] All you would normally need is the attached file, placed in ~/.local/share/lensfun/ Perhaps you can share a RAW sample, which doesn't work. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org Sigma Sigma 70-300mm f/4-5.6 DG Macro Sony Alpha Sigma SA Pentax KAF2 Canon EF Nikon F AF 1
Re: [darktable-user] Lens not identified by darktable, but by exiv2
On 05/11/2018 06:26 PM, Аl Воgnеr wrote: > Am Fri, 11 May 2018 18:05:06 -0400 > schrieb Šarūnas <saru...@mail.saabnet.com>: > >> On 05/11/2018 04:32 PM, Аl Воgnеr wrote: >>> Am Fri, 11 May 2018 15:03:02 -0400 >>> schrieb Patrick Shanahan <p...@opensuse.org>: >>> >>> Hi, >>> >>> it looks like it is getting complicated. There are files at >>> https://www.pentaxforums.com/articles/hands-on-reviews/pentax-k-1d-fa-24-70mm-full-size-sample-photos.html >>> I tried http://www.pfpho.com/k1/samplephotos/160513_53p.dng and dt >>> recognizes the lens, but with _my_ file it does not. I am sorry, I >>> cannot publish this file, but I will take a shot ASAP und make it >>> public, probably in about 24hrs. >>> >>> I don't talk at the moment about lensfun. With Xenial I copied the >>> file of the Tamron 24-70, which is the same lens. I discussed this >>> already with lensfun maintainers. >>> >>> ~# ldd `which darktable` | grep exiv >>> libexiv2.so.14 => /usr/lib/x86_64-linux-gnu/libexiv2.so.14 >>> (0x7f982d35) >>> >>> This version was compiled: >>> exiv2-trunk/src/.libs/libexiv2.so.26 >>> >>> So I have: >>> /usr/local/lib/libexiv2.so.26 >>> /usr/local/lib/libexiv2.so.26.0.0 >>> >>> /usr/lib/86_64-linux-gnu/libexiv2.so.14 >>> /usr/lib/x86_64-linux-gnu/libexiv2.so.14.0.0 >> >> You may have to check /etc/ld.so.conf (and its includes) — does it >> include /usr/local/? Then run `ldconfig`. With `ldconfig -v` you may >> check what libraries are actually accounted for. > > Hi, > > cat /etc/ld.so.conf > include /etc/ld.so.conf.d/*.conf > > # grep -r local /etc/ld.so.conf.d/ > /etc/ld.so.conf.d/libc.conf:/usr/local/lib > /etc/ld.so.conf.d/x86_64-linux-gnu.conf:/usr/local/lib/x86_64-linux-gnu > /etc/ld.so.conf.d/i386-linux-gnu.conf:/usr/local/lib/i386-linux-gnu > /etc/ld.so.conf.d/i386-linux-gnu.conf:/usr/local/lib/i686-linux-gnu > > > # ldconfig -v | grep exiv [...] > So should I run ldconfig? Yes, and it looks like you have already did. Now 'ldd `which darktable`' should show, that your compiled /usr/local/lib/exiv... is used. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Lens not identified by darktable, but by exiv2
On 05/11/2018 04:32 PM, Аl Воgnеr wrote: > Am Fri, 11 May 2018 15:03:02 -0400 > schrieb Patrick Shanahan <p...@opensuse.org>: > > Hi, > > it looks like it is getting complicated. There are files at > https://www.pentaxforums.com/articles/hands-on-reviews/pentax-k-1d-fa-24-70mm-full-size-sample-photos.html > I tried http://www.pfpho.com/k1/samplephotos/160513_53p.dng and dt > recognizes the lens, but with _my_ file it does not. I am sorry, I > cannot publish this file, but I will take a shot ASAP und make it > public, probably in about 24hrs. > > I don't talk at the moment about lensfun. With Xenial I copied the file > of the Tamron 24-70, which is the same lens. I discussed this already > with lensfun maintainers. > > ~# ldd `which darktable` | grep exiv > libexiv2.so.14 => /usr/lib/x86_64-linux-gnu/libexiv2.so.14 > (0x7f982d35) > > This version was compiled: > exiv2-trunk/src/.libs/libexiv2.so.26 > > So I have: > /usr/local/lib/libexiv2.so.26 > /usr/local/lib/libexiv2.so.26.0.0 > > /usr/lib/86_64-linux-gnu/libexiv2.so.14 > /usr/lib/x86_64-linux-gnu/libexiv2.so.14.0.0 You may have to check /etc/ld.so.conf (and its includes) — does it include /usr/local/? Then run `ldconfig`. With `ldconfig -v` you may check what libraries are actually accounted for. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Lens not identified by darktable, but by exiv2
On 2018-05-11 15:03, Patrick Shanahan wrote: > * Аl Воgnеr <darktable...@corr.eu.org> [05-11-18 14:47]: [...] >> [MakerNotes]0x0001 Lens >> Type : HD PENTAX-D FA 24-70mm F2.8 ED SDM WR >> [...] >> So could it be, that dt uses another exiv2 in my system? I cannot find >> one. 'ldd `which darktable`' to find out. > > I have: >lensfun-tools-0.3.2-39.5.x86_64 >liblensfun1-0.3.2-39.5.x86_64 >python3-lensfun-0.3.2-39.5.x86_64 >lensfun-data-0.3.2-39.5.noarch > > and >grep 24-70 /usr/share/lensfun/version_1/*-pentax.xml > has no output > > I would guess lensfun does not contain data for your lense > > put a raw image on a file share service for trial. I will open it in my > dt (git-master) and see what lense correction is available. Lens has been calibrated, data is in lensfun Github repository. You can extract entry from this file https://raw.githubusercontent.com/lensfun/lensfun/master/data/db/slr-pentax.xml and put it in your ~/.local/share/lensfun/ until it becomes available via `lensfun-update-data`. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] ... new Ubuntu version Bionic
On 2018-05-11 02:56, Frank J. wrote: > Am 10.05.2018 um 23:53 schrieb Аl Воgnеr: >> I had big problems with Ubuntu 16.04, but now with Ubuntu 18.04 it >> is getting more worse. > ... >> ..., then I changed to ppa:pmjdebruijn/darktable-unstable which is >> the same at the moment. > ... >> Which experiences did you make with Bionic? >> >> Al > > Hi, any other experiences? > > I'm ready to upgrade Ubuntu 16.04 LTS to 18.04 LTS, using DT from > ppa.launchpad.net/pmjdebruijn/darktable-release/ubuntu > (not from unstable) As for 18.04, with KDE, all is fine with darktable, lensfun etc. If you make use of AMD graphics with OpenCL, then you may want to use kernel 4.13 (18.04 is 4.15). I upgraded from 17.10 and did not see much difference if at all. You may see it differently. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] OpenCL performance with newer kernels?
On 2018-04-28 22:45, Šarūnas wrote: > On 04/28/2018 01:01 PM, Peter McD wrote: >> Any news on the openCL performance with ubuntu 18.04 or other >> distros with even newer kernels? related to amdgpu vs amdgpu-pro? > > No performance with newer kernel (4.15) at all :) AMDGPU-PRO wasn't > released yet. Open source ROCm works with 4.13 (which is easy to keep > either with APT pinning linux-image-generic to 'artful' or using kernels > from Ubuntu mainline PPA). With ROCm 1.7 and Linux 4.13, RX480 works > fine, Vega64 has some strange OpenCL issues in both Gnome and KDE. P.S. AMDGPU-PRO 18.10 was just released and it works on Ubuntu 18.04, including OpenCL, provided the kernel is 4.13¹. RX Vega64 + darktable do the venerable bench.SRW in 2.5s. No issues with Xorg/KDE. 1. 4.13 can be installed from kernel.ubuntu.com/~kernel-ppa/mainline or by pinning linux-image-generic to 17.10 Artful (that way one gets updates too). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] OpenCL performance with newer kernels?
On 04/28/2018 01:01 PM, Peter McD wrote: > Any news on the openCL performance with ubuntu 18.04 or other > distros with even newer kernels? related to amdgpu vs amdgpu-pro? No performance with newer kernel (4.15) at all :) AMDGPU-PRO wasn't released yet. Open source ROCm works with 4.13 (which is easy to keep either with APT pinning linux-image-generic to 'artful' or using kernels from Ubuntu mainline PPA). With ROCm 1.7 and Linux 4.13, RX480 works fine, Vega64 has some strange OpenCL issues in both Gnome and KDE. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] amd - graphicdriver - Situation with debian.
On 04/24/2018 06:19 PM, frieder wrote: > Am Dienstag, den 24.04.2018, 17:06 -0400 schrieb Šarūnas: >> On 04/24/2018 04:23 PM, frieder wrote: >>> At the moment I'm mostly using Debian 8 (jessy) because of missing open >>> CL driver Support, because fglrx isn't supporten with latest kernels any >>> more (grml) with Debian 9. >>> Currently I have Debian 8 and 9 installed on my PC, on both systems >>> there were *issues* with some components I need. This is why I have to >>> switch, Debian 8 is still more importend. >>> >>> This is why I decided to buy a later graphic Card for better driver >>> support, now I have a Radeon RX560 with 4MB on my table, to get a better >>> situation. >>> >>> Can someone tell me what driver-options I do have on Debian 8 and Debian >>> 9? >>> And which ones does include OpenCl and are to recomend? >> >> If you can get kernel 4.11–4.13 on any of these Debian, then you can use >> open source ROCm OpenCL (github.com/RadeonOpenCompute/ROCm). amdgpu >> itself is already in the Linux kernel. >> > Sounds good... altough, Debian based on Kernel 3.16 and 4.9. > Backport Kernels from 4.13 are not available for X86/amd64 architecture. Sounds very stable indeed :) I had to go back to 4.13 (from 4.17) in order to be able to use ROCm… -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] amd - graphicdriver - Situation with debian.
On 04/24/2018 04:23 PM, frieder wrote: > At the moment I'm mostly using Debian 8 (jessy) because of missing open > CL driver Support, because fglrx isn't supporten with latest kernels any > more (grml) with Debian 9. > Currently I have Debian 8 and 9 installed on my PC, on both systems > there were *issues* with some components I need. This is why I have to > switch, Debian 8 is still more importend. > > This is why I decided to buy a later graphic Card for better driver > support, now I have a Radeon RX560 with 4MB on my table, to get a better > situation. > > Can someone tell me what driver-options I do have on Debian 8 and Debian > 9? > And which ones does include OpenCl and are to recomend? If you can get kernel 4.11–4.13 on any of these Debian, then you can use open source ROCm OpenCL (github.com/RadeonOpenCompute/ROCm). amdgpu itself is already in the Linux kernel. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Bug report
On 2018-04-20 15:31, Henkka wrote: > Tried to make two low pass masks. Mission impossible. Watch this video. > Easier to show than try to tell. > > https://youtu.be/wBy8nYlOpmQ As expected: each lowpass instance is blurring other instance's duck. Instead of multiple instances you might want to use multiple masks in the same instance. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] CPU Recommendations
ssor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor model name : AMD Ryzen 5 1600 Six-Core Processor total used free shared buff/cache available Mem: 15G 2.1G 11G 75M 1.9G 13G Swap: 0B 0B 0B 08:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1070] (rev a1) Linux storas 4.13.0-37-generic #42-Ubuntu SMP Wed Mar 7 14:13:23 UTC 2018 x86_64 x86_64 x86_64 GNU/Linu -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Checking for errors/debugging my openCL
On 03/07/2018 04:30 PM, Bernhard wrote: > > > Ulrich Pegelow schrieb am 05.03.2018 um 17:54: >> Additionally you may want to play with opencl_number_event_handles. >> Start with an extreme setting of zero. In case this would solve your >> issues you can then try different values in-between. >> >> Ulrich > I changed this value with no result. > > Then I learned that there is a little tool called clinfo. > I fear that it's a hardware or system issue now. > > I get > > $ clinfo > Number of platforms 1 > Platform Name NVIDIA CUDA > Platform Vendor NVIDIA Corporation > Platform Version OpenCL 1.2 CUDA 9.0.282 > Platform Profile FULL_PROFILE > Platform Extensions cl_khr_global_int32_base_atomics > cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics > cl_khr_local_int32_extended_atomics cl_khr_fp64 > cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing > cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll > cl_nv_copy_opts cl_nv_create_buffer > Platform Extensions function suffix NV > > Platform Name NVIDIA CUDA > Number of devices 1 > Device Name GeForce GTX 1050 Ti > Device Vendor NVIDIA Corporation > Device Vendor ID 0x10de > Device Version OpenCL 1.2 CUDA > Driver Version 384.111 > [...] Perhaps it is possible to try different versions of nvidia driver and CUDA, older/newer? Nvidia has a network repository at http://developer.download.nvidia.com/compute/cuda/repos/ -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Heavily disappointed about openCL
On 02/11/2018 03:06 PM, Bernhard wrote:
>
>
> Roman Lebedev schrieb am 11.02.2018 um 20:47:
>> On an unrelated note: not everyone speaks ${langname}
>>
>> Roman.
> sorry, but could figure out how to change:
>
>
> Michael Rasmussen schrieb am 11.02.2018 um 16:42:
>> If you do this what do you get?
>> $ wgethttp://www.mirada.ch/bench.SRW
>> $ wgethttp://www.mirada.ch/bench.SRW.xmp
>>
>> $ darktable-cli bench.SRW test.jpg --core --disable-opencl -d perf
>> >>> pixel pipeline processing took 40,388 secs (215,986 CPU)
>>
>> $ darktable-cli bench.SRW test.jpg --core -d perf -d opencl
>> >>> pixel pipeline processing took 8,440 secs (15,970 CPU)
>
> ~ $ darktable-cli bench.SRW test.jpg --core --disable-opencl -d perf
> [...]
> [dev_process_export] pixel pipeline processing took 28,220 secs (105,356
> CPU)
>
> ~ $ darktable-cli bench.SRW test.jpg --core -d perf -d opencl
> [...]
> [dev_process_export] pixel pipeline processing took 7,026 secs (9,684 CPU)
> [export_job] exported to `test_05.jpg'
So it looks like using GPU with OpenCL gives a nice speedup from 28s to
7s. When it works. I can't tell, why disin't it wrok in your previoaus
attempts, but here some empirical suggestion.
1. Backup you current darktable config (~/.config/darktable/darktablerc)
to somwhere safe.
2. Delete it and let darktable create it anew on ext start.
3. See what you get with processing times. If it now uses GPU, check the
difference between the new and your old darktablerc (diff darktablerc
darktablerc_old).
--
Šarūnas Burdulis
math.dartmouth.edu/~sarunas
signature.asc
Description: OpenPGP digital signature
Re: [darktable-user] lens correction for Canon Powershot G7 X under darktable 2.4.1 and Ubuntu 16.04 LTS
On 02/08/2018 04:34 PM, Michael Rasmussen wrote: > Hi Tanja, > > I have posted this to the list as well so that other readers can follow > the discussion. > > The reason behind the older version of the library on your installation > is because you a running latest LTS version from Ubuntu while I am > running Debian Unstable which means the version of my libraries are the > latest versions that are not considered experimental. But since you > have the library installed this is not the reason for your problem. > > On Thu, 8 Feb 2018 21:28:29 +0100 > "Tanja Etzelstorfer" <tanj...@gmx.at> wrote: > >> Hi Michael! >> >> The dpkg command gives the following output: >> >> liblensfun-data install >> liblensfun0 install >> >> seems this is an older version.. Does running `lensfun-update-data` make a difference? (It is provided by the liblensfun-bin package.) -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Darktable on a 4K display
On 2018-02-07 07:36, René Seindal wrote: > Hi > > Does anybody use darktable on a 4K display? Yes. It works well (144 DPI setting in KDE), but the system has more resources: 16GB RAM and a GPU with 8GB, working OpenCL. > I have tried, and it crashes a lot. Running it from a terminal just > gives the message "Killed". > > The kernel log states that it is an out-of-memory situation. > > It is not a very new computer. Its an Intel NUC with an i5 cpu with > integrated graphics, and 8Gb of ram. Your integrated graphics probably takes a good chunk out of those 8GB. > I've tried to monitor ram usage using top while darktable is > working, but it never goes over 10% with lots of memory still > available. > > There must be some sudden spike in memory allocation that causes the > kernel to run out of memory, that I cannot see using top. If you haven't yet, you may run 'darktable -d memory' to see how much memory it uses. > It almost certainly happens in the pixel pipeline, either when > rendering a photo for display, or during export. > > If I switch the display to normal FHD 1920x1080 resolution, there are > no crashes. 4K is roughly 4 times the pixels of FHD, which probably translates into memory needs... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] CPU Recommendations
On 02/04/2018 05:38 PM, Robert Bieber wrote: > Oh yeah, I'm just talking about the price of the CPU, not the whole PC. > I'm planning on doing all out on RAM and solid state drives as well. I > guess GPU might be of some concern as well, how has GPU processing in > dark table on Linux been going lately? I know it exists, but I'm always > scared of anything to do with graphics drivers on Linux GPU might just be the upgrade with the biggest impact for RAW processing. To have that impact it will need not only the drivers to work, but also the corresponding OpenCL part... This topic comes up from time to time here, so, as already mentioned, it may make sense to search the archives. With Nvidia things may seem more stable, but it's not that one just installs proprietary drivers + OpenCL from the distribution's repositories and it "just works". You need proprietary Nvidia drivers for OpenCL to work. With AMD it is already possible to have a completely open source setup. amdgpu is in the kernel, and the OpenCL part can come from the open source ROCm. It will limit you to certain AMD GPUs and Linux kernel/X versions. To have a wider range, the open source kernel amdgpu can be combined with the OpenCL part from AMD's proprietary amdgpu-pro — that works well too. I have some random tests listed here: math.dartmouth.edu/~sarunas/darktable_bench.html -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] CPU Recommendations
On 02/04/2018 02:41 PM, Robert Bieber wrote: > Hi everyone, > > I think it's time to finally replace the computer I built five-ish years > ago. It was a good machine at the time, but since then my RAW files > have gone from 8MP to 16MP and now to 40, and running more expensive > iops, especially with masks and so on, is getting to be pretty sluggish. > > I haven't really kept up with PC hardware in the meantime, so I'm > curious what y'all would recommend for CPUs. How much is a decent > amount to spend for something high-end in the US market? I'm guessing > that dropping a grand on something/really/ high end is probably > unnecessary, but maybe I'm wrong. It may depend on where on Earth, but “grand” as in ~1000USD, will only get you half-way, at best, to a high end PC... You may want to check AMD Ryzen 5/7 + some higher-end GPUs, NVMe solid state storage, healthy amounts of RAM. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] openCL darktable - amddgpu-pro Radeon RX 460 and Ubuntu 16.04.03
On 2017-12-20 02:59, Peter Mc Donough wrote: > Hi, > > I use the amdgpu-pro driver with my standard Kubuntu 16.04.3, > installed with the --compute option, ie just openCL > > With the recent amdgpu-pro 17.50 I thought I could try a full install > of the driver. It works but openCL is not activated in darktable. > > If I install just openCL with, as suggested for the AMD Radeon > RX460, > > ./amdgpu-pro-install --opencl=legacy --headless > > it works. > > Did I miss something? Looking at the script, if it is invoked without options: ./amdgpu-pro-install it acts as ./amdgpu-pro-install --opencl=rocm So if you would still like to try the full install, then you should probably use ./amdgpu-pro-install --opencl=legacy This will install the “legacy” amdgpu-pro OpenCL, not the new open-source ROCm. Looks like the latter might be not working in your 16.04.3 case... -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] lensfun database update, lens ok camera ko
On 11/22/2017 04:37 PM, Brune Wayce wrote: > ... ... ... > I've made a file like your suggestion, with just an entry for the > DMC-G7, and now it works! > > But why in ~/.local/share/lensfun is accepted, and in /usr/share/lensfun > is not? That I don't know. Maybe a syntax error or darktable/lensfun not reading that location in your case. To find out, you may try starting darktable from the command line. You might see some error messages. Or start darktable with strace as suggested in previous message by Lorenzo: https://www.mail-archive.com/darktable-user@lists.darktable.org/msg04214.html -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Sharpening after resize
On 2017-11-14 09:18, Paul Deverson wrote: > Personally, I don’t sharpen in dt at all. I export unsharpened to a > tiff and then pre-sharpen using Nik Sharpener inside GIMP and then > output sharpen using Nik after resizing. > > Works for me! Yes, of course that's an option. When using other software I did output to TIFF and then batch-sharpened with Imagemagick[1]. Worked very well. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas 1. https://redskiesatnight.com/2005/04/06/sharpening-using-image-magick signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Is Darktable compatible with MAC OS X 12.13?
Darktable 2.2.5, installed from the .dmg referenced here http://www.darktable.org/install/#osx does work on macOS 12.13. I'm not using macOS much (it's only test workstation), but darktable installs and works fine as far applying some familiar modules and exporting. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Sharpening after resize
On 2017-11-13 20:34, Robert William Hutton wrote: > On 14/11/17 12:22, darkta...@911networks.com wrote: >> What are you doing/using for sharpening after export? My export sizes >> are around 900px jpegs for >> web with a quality of 70% for my 20Mpx and 24Mpx cameras. > > Isn't this what export styles are for? > > https://www.darktable.org/usermanual/ch02s03s12.html.php#d0e3405 That's what I am doing (have a set of styles which only include different settings for sharpening and apply one of them in “append” mode in export module). I assumed it is applied as a last step. But now, after reading the above reference, I'm really not sure, where in the pipeline is that style “appended”? Thanks, -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] what graphics card to get?
On 2017-01-17 09:50, Šarūnas wrote: > On 2017-01-16 11:23, Patrick Rudin wrote: >> schrieb Michael Born <michael.b...@aei.mpg.de>: >>> I'm running a RX480 with MESA OpenGL (because, I like open source) >>> and the OpenCL libs from the AMDGPU-PRO driver (see >>> http://www.gearsongallium.com/?p=2960 ) >>> Darktable works fine with this. >> >> Wow, very interesting. >> >> Could you please put these two files in the same directory: >> http://www.mirada.ch/bench.SRW >> http://www.mirada.ch/bench.SRW.xmp >> >> and post the full output of >> darktable-cli bench.SRW test.jpg --core -d perf -d opencl >> here? Haven't seen it noted here, but AMD's open source rocm-opencl[1] now seems to work fine as far as OpenCL provider for darktable. In my case it's Kubuntu 17.10, Xorg, AMD RX480, darktable 2.2.5. I was occasionally running benchmarks on the above provided bench.SRW with its .xmp settings, when I had access to some different hardware. Here are some random results: https://math.dartmouth.edu/~sarunas/darktable_bench.html Of course the numbers merely useful show that OpenCL works, as there are many other system variables to be accounted for. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas 1. https://github.com/RadeonOpenCompute/ROCm signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Which config for Darktable
On 06/24/2017 11:51 AM, Germano Massullo wrote: > Il 24/06/2017 15:02, Michael Below ha scritto: >> Hi, >> >> Am Fr 23 Jun 2017 16:44:33 CEST >> schrieb Mark Heieis <mhei...@alois.ca>: >> >>>>> Am 23.06.2017 um 17:01 schrieb Mark Heieis: >>>>>> To get AMD opencl on newer cards in Fedora 25, download current >>>>>> CentOS/RHEL amdgpu pro driver (17.10.). >>>>>> >>>>>> Unpack it and then under root run ./amggpu-pro-install --compute, >>>>>> which just installs opencl support >>>>> This is as simple as it should be! >>>> Sorry, no... I'm glad it works for you, but this is a hack that >>>> might break with the next update. It's a major advantage that the >>>> NVidia driver is available for a number of distributions, including >>>> Ubuntu and Debian... >>> I fail to see the problem and why it's a "no" and a "hack", as it is >>> just installing support libs as far as I can tell. It's been working >>> for me (RX480) since amdgpu-pro16.x using out-of-the-box Fedora 25. >>> Please explain how it would break, I'd like to understand. Worst >>> case, you just have to rerun ./amggpu-pro-install --compute >> It's not "as simple as it should be", since you are generally using >> Fedora package management to keep track of installed software, and you >> have to work around that here and install stuff by hand to get OpenCL >> on AMD, because of lacking support. And it's a "hack" since you are >> using part of a package built for RedHat. This may work in the current >> combination of RH version and Fedora version, but it is a matter of >> luck, not a matter of design. With the next change in either >> distribution the dice may fall another way. I wouldn't want to spend a >> couple hundred euros on that basis. >> >> But as I said, I'm glad it works for you. >> > Indeed as a Fedora packager I would never use such "hack around"s, but I > am glad that it works for him I haven't used amdgpu/amdgpu-pro on Fedora or any RPM systems. For Ubuntu, amdgpu-pro is not much more, but a set of regular .deb packages. One can use the provided script amdgpu-pro-install, which installs all those debs. But they can also be intalled individually, using dpkg. With open source amdgpu already in the kernel, there is no need to install all the debs. That's why there is a --compute option, which merely installs a subset, i.e. a few packages with OpenCL libraries, in a standard, APT way. No compilation is necessary; it works with all the kernels, including the latest 4.12 RC. It never occurred to me that this is a hack, i.e. installing some libs, but perhaps there are reasons I don't know about. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Which config for Darktable
On 2017-06-23 08:57, Guillermo Rozas wrote: >>> Honest question: what is the sense of doing this? If one is using the >>> open source drive on "moral grounds", taking the OpenCL blob from it >>> and using it when convenient doesn't make it any less closed. I >>> confess that I don't know the status of the proprietary drivers for >>> AMD, are they fundamentally worse than the open ones? >> There are no functional/usable/not-broken free/open source opencl drivers. >> At least if you try to use darktable, and not just run some toy benchmarks. > > I get that. What I don't understand is what is the sense of using the > open source driver AND the OpenCL part of the closed one, instead of > directly using the closed source driver. Is more of a philosophical > question :) The sense is practical: AMDGPU-PRO (closed source driver) as a complete package is only available for Linux 4.4 and some distributions. amdgpu (opensource, no OpenCL) is part of Linux kernel, starting with >4.4(?). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Fedora Spin, was Does OpenCL require the AMDGPU-PRO driver? Yes, but does not require installation, just some libs
On 06/03/2017 05:41 PM, Pascal Obry wrote: > Le samedi 03 juin 2017 à 17:28 -0400, Šarūnas Burdulis a écrit : >> amdgpu is in the kernel, so no need >> fo module recompile (that's one reason I prefer it over nvidia…). > > But tthe amdgpu on the kernel does not have "image" support which is > required by darktable. So yes the module sit there, but can't be used > with darktable, [...] Hmm, not sure if we are not talking about the same things, but, at least for me, on Kubuntu 17.04, amdgpu + amdgpu-pro OpenCL libs work perfectly with darktable, i.e. OpenCL is available as reported by the darktable and a sample image processing takes few seconds as compared to tens of seconds with --disable-opencl. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Fedora Spin, was Does OpenCL require the AMDGPU-PRO driver? Yes, but does not require installation, just some libs
On 06/03/2017 04:46 PM, Peter Mc Donough wrote: > Am 02.06.2017 um 16:09 schrieb Peter Mc Donough: >> Am 02.06.2017 um 14:40 schrieb Šarūnas: >>> On 2017-06-02 06:22, Peter Mc Donough wrote: > >>> ... >>> AMDGPU-PRO installer now has a '--compute' option to install only the >>> OpenCL part: >>> >>> ./amdgpu-pro-install --help >>> Usage: amdgpu-pro-install [options...] >>> Options: >>>-h|--help display this help message >>>--px PX platform support >>>--compute OpenCL support only >>> >>> This works perfectly with amdgpu on [k]ubuntu, at least in 17.04. >> >> I' give it a try. >> > > Success with a fresh installation of Kubuntu 17.04. Darktable shows > openCL aktive, I left "home" untouched. > Let's see how it survives a kernel update? Yes, let's see. I “went” through a lot of kernel changes with this setup, from Ubuntu stock to mainline 4.10/11/12 and back (in order to get some weird Gigabyte BIOS to boot with the new Ryzen 7), and amdgpu-pro OpenCL still worked. amdgpu is in the kernel, so no need for module recompile (that's one reason I prefer it over nvidia…). -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] How to deal with underexposed image
On 2017-05-26 15:18, Piotr Ryszkiewicz wrote: > Hi all, > > I am trying to learn postprocessing photographs, still at the > beginning of learning curve. I'd like to ask more experienced users > what can I do with underexposed image as this one: > > https://drive.google.com/open?id=0B3cSQCYRbO3RZWZIZURKUEF6Rms > > I tried my best and was able to get output better than jpg out of > camera, but still I am not satisfied, as the result looks somehow > unnatural. Here is it: > > https://drive.google.com/open?id=0B3cSQCYRbO3RSkVtaXJPLURUUjQ > And here is xmp file: > https://drive.google.com/open?id=0B3cSQCYRbO3RSGpiM3YxV3pZV3c > > Would you be so kind and give me some suggestions how can I improve it > ? Or even better - could you process it yourself and post your xmp > file ? Here is my take: https://math.dartmouth.edu/owncloud/s/dt6IpL3RIHrYmSs The scene is back-lit, so most attempts to fill-in the shadows are likely to result in somewhat “unnatural” lighting. In such a situation the best probably would have been to use a fill-in flash, which might have also allowed to preserve details in the sky (reflection in the puddle reveals that there might have been some interesting sky). In my opinion, image is not underexposed — the sky part is overexposed. darktable 2.2.4; default base curve for the camera (no fusion). I assume we use calibrated and profiled displays, not only for darktable, but also whatever program is used for viewing the processed images is “color managed”. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Does OpenCL require the AMDGPU-PRO driver? Yes, but does not require installation, just some libs
On 2017-03-30 17:19, Peter Mc Donough wrote: > [...] > Just for getting the feel of it I run Xubuntu 16.04 LTS in virtualbox, > with an eye on installing it on a real machine. > There was a regular update on my virtual Xbuntu, kernel 4.4 to kernel > 4.8. So they are updating it. > > The following relates to Ubuntu, but if you have an answer;-) > Quesions would be: > Does the ampgpu-pro work with this kernel on a "real" Xubuntu machine. I > wouldn't bother if openCL were not available there. 'amdgpu', combined with a few files extracted from 'amdgpu-pro' package to enable OpenCL, does work with Linux 4.8. Your AMD chip needs to be among those supported by amdgpu and amdgpu-pro, of course. -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Does OpenCL require the AMDGPU-PRO driver? Yes, but does not require installation, just some libs
On 2017-03-28 01:45, Ulrich Pegelow wrote: > Am 27.03.2017 um 19:58 schrieb Pascal Obry: >> 01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. >> [AMD/ATI] Venus XTX [Radeon HD 8890M / R9 M275X/M375X] (rev 83) >> (prog-if 00 [VGA controller]) >> Subsystem: Dell Venus XTX [Radeon HD 8890M / R9 M275X/M375X] [...] > That's a Sea Island generation GPU from AMD. amdgpu-pro does not support > Sea Island and its predecessor Southern Island. The latter affects my > HD7950 (Tahiti). The following are listed as “compatible” on [1]: Radeon HD7700/7800/8500/8600 Radeon HD7700M/7800M/8500M/8530M/8600M/8700M/8790M/8800M Perhaps “compatible” != “supported” ? 1. http://support.amd.com/en-us/kb-articles/Pages/AMDGPU-PRO-Driver-for-Linux-Release-Notes.aspx -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Does OpenCL require the AMDGPU-PRO driver? Yes, but does not require installation, just some libs
On 2017-03-27 13:58, Pascal Obry wrote: >> Other info: >> >> kernel: 4.8.0-44-generic > > 4.9.0-2-amd64 Just tried with 4.9 and 4.10 from mainline kernel PPA: Found linux image: /boot/vmlinuz-4.10.6-041006-generic Found linux image: /boot/vmlinuz-4.9.18-040918-generic amdgpu + amdgpu-pro mix still works. >> lspci -v >> 02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. >> [AMD/ATI] Ellesmere [Radeon RX 470/480] (rev c7) (prog-if 00 [VGA >> controller]) >> Subsystem: XFX Pine Group Inc. Radeon RX 480 >> [...] >> Kernel driver in use: amdgpu >> Kernel modules: amdgpu > > Mine is: > > $ lspci -v > 01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] > Venus XTX [Radeon HD 8890M / R9 M275X/M375X] (rev 83) (prog-if 00 [VGA > controller]) > Subsystem: Dell Venus XTX [Radeon HD 8890M / R9 M275X/M375X] > Flags: bus master, fast devsel, latency 0, IRQ 142 > Memory at c000 (64-bit, prefetchable) [size=256M] > Memory at dfe0 (64-bit, non-prefetchable) [size=256K] > I/O ports at e000 [size=256] > Expansion ROM at 000c [disabled] [size=128K] > Capabilities: > Kernel driver in use: radeon > Kernel modules: radeon ^^ I'm afraid you would need to get 'Venus XTX Radeon HD 8890M' to work with amdgpu first… If it belongs to GCN 1st gen., then there is a chance, but for now compiling your own kernel seems necessary, e.g.: http://askubuntu.com/questions/861938/ubuntu-and-amds-gcn-1-0-gpus-with-amdgpu-pro-16-50-drivers -- Šarūnas Burdulis math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] what graphics card to get?
On 2017-01-16 11:23, Patrick Rudin wrote: > schrieb Michael Born <michael.b...@aei.mpg.de>: >> I'm running a RX480 with MESA OpenGL (because, I like open source) >> and the OpenCL libs from the AMDGPU-PRO driver (see >> http://www.gearsongallium.com/?p=2960 ) >> Darktable works fine with this. > > Wow, very interesting. > > Could you please put these two files in the same directory: > http://www.mirada.ch/bench.SRW > http://www.mirada.ch/bench.SRW.xmp > > and post the full output of > darktable-cli bench.SRW test.jpg --core -d perf -d opencl > here? Using OpenCL libraries extracted from the latest amdgpu-pro-16.50 recognizes all the 8GiB of memory in RX 480. There is no fallback to CPU. -- Šarūnas Burdulis http://math.dartmouth.edu/~sarunas [opencl_init] opencl related configuration options: [opencl_init] [opencl_init] opencl: 1 [opencl_init] opencl_library: '' [opencl_init] opencl_memory_requirement: 768 [opencl_init] opencl_memory_headroom: 300 [opencl_init] opencl_device_priority: '*/!0,*/*/*' [opencl_init] opencl_size_roundup: 16 [opencl_init] opencl_async_pixelpipe: 0 [opencl_init] opencl_synch_cache: 0 [opencl_init] opencl_number_event_handles: 25 [opencl_init] opencl_micro_nap: 1000 [opencl_init] opencl_use_pinned_memory: 0 [opencl_init] opencl_use_cpu_devices: 0 [opencl_init] opencl_avoid_atomics: 0 [opencl_init] opencl_enable_markesteijn: 1 [opencl_init] [opencl_init] found opencl runtime library 'libOpenCL' [opencl_init] opencl library 'libOpenCL' found on your system and loaded [opencl_init] found 1 platform [opencl_init] found 2 devices [opencl_init] device 0 `Ellesmere' supports image sizes of 16384 x 16384 [opencl_init] device 0 `Ellesmere' allows GPU memory allocations of up to 4048MB [opencl_init] device 0: Ellesmere GLOBAL_MEM_SIZE: 7526MB MAX_WORK_GROUP_SIZE: 256 MAX_WORK_ITEM_DIMENSIONS: 3 MAX_WORK_ITEM_SIZES: [ 256 256 256 ] DRIVER_VERSION: 2236.5 DEVICE_VERSION: OpenCL 1.2 AMD-APP (2236.5) [opencl_init] compiling program `demosaic_ppg.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/demosaic_ppg.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/demosaic_ppg.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `atrous.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/atrous.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/atrous.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `basic.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/basic.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/basic.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `blendop.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/blendop.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/blendop.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `highpass.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/highpass.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/highpass.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `nlmeans.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/nlmeans.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/nlmeans.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `gaussian.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/gaussian.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/gaussian.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `sharpen.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/sharpen.cl.bin' [ope
Re: [darktable-user] what graphics card to get?
On 2017-01-16 13:43, Martin Schoepf wrote: > Hi everybody, > > I attached the output of my AMD 480 8gb. > > I previously had a NVDIA GTX 770, and I would have kept it, if it > wouldn't have broke. > > I'm running Kubuntu 16.10 with the amd-pro driver. > Intel(R) Core(TM) i7-4771 CPU @ 3.50GHz, 16gb ram > > I was happy with both gpus, but if you can live with ubuntu and the > amd-gpu pro driver I think you get more for less bugs at AMD. Attached is the output from using RX 480 with open-source amdgpu and libOpenCL libraries from amdgpu-pro (as per http://www.gearsongallium.com/?p=2960), Ubuntu 16.10 + plasma-desktop. Martin, how do you use amdgpu-pro with 16.10? Kernel from 16.04? -- Šarūnas Burdulis http://math.dartmouth.edu/~sarunas [opencl_init] opencl related configuration options: [opencl_init] [opencl_init] opencl: 1 [opencl_init] opencl_library: '' [opencl_init] opencl_memory_requirement: 500 [opencl_init] opencl_memory_headroom: 0 [opencl_init] opencl_device_priority: '*/!0,*/*/*' [opencl_init] opencl_size_roundup: 16 [opencl_init] opencl_async_pixelpipe: 0 [opencl_init] opencl_synch_cache: 0 [opencl_init] opencl_number_event_handles: 25 [opencl_init] opencl_micro_nap: 1000 [opencl_init] opencl_use_pinned_memory: 0 [opencl_init] opencl_use_cpu_devices: 0 [opencl_init] opencl_avoid_atomics: 1 [opencl_init] opencl_enable_markesteijn: 1 [opencl_init] [opencl_init] found opencl runtime library 'libOpenCL' [opencl_init] opencl library 'libOpenCL' found on your system and loaded [opencl_init] found 1 platform [opencl_init] found 2 devices [opencl_init] device 0 `Ellesmere' supports image sizes of 16384 x 16384 [opencl_init] device 0 `Ellesmere' allows GPU memory allocations of up to 154MB [opencl_init] device 0: Ellesmere GLOBAL_MEM_SIZE: 617MB MAX_WORK_GROUP_SIZE: 256 MAX_WORK_ITEM_DIMENSIONS: 3 MAX_WORK_ITEM_SIZES: [ 256 256 256 ] DRIVER_VERSION: 2036.3 (VM) DEVICE_VERSION: OpenCL 1.2 AMD-APP (2036.3) [opencl_init] compiling program `demosaic_ppg.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/demosaic_ppg.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/demosaic_ppg.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `atrous.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/atrous.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/atrous.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `basic.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/basic.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/basic.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `blendop.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/blendop.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/blendop.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `highpass.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/highpass.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/highpass.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `nlmeans.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/nlmeans.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/nlmeans.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `gaussian.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/gaussian.cl.bin' [opencl_load_program] successfully loaded program from `/usr/share/darktable/kernels/gaussian.cl' [opencl_build_program] successfully built program [opencl_build_program] BUILD STATUS: 0 BUILD LOG: [opencl_init] compiling program `sharpen.cl' .. [opencl_load_program] loaded cached binary program from file `/home/sarunas/.cache/darktable/cached_kernels_for_Ellesmere/sharpen.cl.bin' [opencl_load_program] successfully loaded
Re: [darktable-user] Locked Video LUT
On 2017-01-15 04:29, Maurizio wrote: > Updated DisplayGUI but problem is still here... > Maurizio The error message you are seeing is probably from dispwin, which is called by the DisplayCAL GUI and needs to be updated. dispwin is part of the Argyll suite of tools. Šarūnas > In data sabato 14 gennaio 2017 19:57:01 CET, David Vincent-Jones ha scritto: >> On openSUSE 42.2 I was able to install DisplayCAL 3.2.3 from the repos. >> and this fixed my problem. >> >> David >> >> On 01/14/2017 01:38 PM, Šarūnas Burdulis wrote: >>> On 01/14/2017 12:57 PM, Maurizio Paglia wrote: >>>> Same problem for me on the same distribution Maurizio >>>> >>>> Il 14 gen 2017 18:49, "David Vincent-Jones" <david...@gmail.com >>>> <mailto:david...@gmail.com>> ha scritto: >>>> >>>> After installing the Nvidia openCL package I now find that DisplayCAL >>>> is unable to install my display calibration. I am getting the message >>>> 'Argyll CMS:Dispwin : Error: - We don't have access to the VideoLUT >>>> for clearing'. >>>> >>>> Any ideas how to get around this problem? >>> >>> Same problem on Ubuntu 16.10. >>> >>> I ended up downloading updated software from >>> >>> http://www.argyllcms.com >>> >>> Version 1.9.2 works fine. >>> >>> Šarūnas >>> math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Activate OpenCL for AMD FirePro D300 in Ubuntu
On 11/20/2016 12:21 PM, Frank J. wrote: > Hallo, > I installed Ubuntu 16.04 LTS on an "ashtray" [1]. > This hardware has 2 powerfull GPUs. > But DT did'nt find a device for "OpenCL". > > Which drivers must I install to get OpenCL-Support for Darktable? > "radeon" ist installed now. > "amdgpu-pro" is not installable, it has some conflicts. amdgpu-pro. It, including OpenCL for darktable, worked on 16.04. Perhaps conflicts can be resolved with 'apt-get -f install'? Šarūnas https://math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] Re: Color checker.
On 07/30/2016 05:35 PM, Michael wrote: > How do we use these things with dark table? I like very much the color balance, which I see in Olympus out-of-camera JPEGs, and I wanted to have a darktable style as a starting point for processing RAW. I used a similar GretagMacbeth color pattern to create a style using "color zones" module. That is, I took a picture of the color checker in direct sunlight and tried various settings on RAW to match the JPG. Color zones module manipulation produced the best result, though the match still wasn't perfect. The resulting style works very well on some scenes, but less so on others. Then I discovered IT8 target from Wolf Faust and this: http://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/ The latter is of course a much more systematic approach. It uses "color checker lut" module (as of today one has to use darktable compiled from git master) and a darktable-chart utility. Using Olympus RAW (ORF) and out-of-camera JPG, the resulting style provides amazingly similar color rendition. For most of the scenes anyway. I used both Olympus' "natural" and "vivid" JPG modes. Still for some scenes, I still prefer my initial, empirical, "color zones" style... In the end, it is good to have these style available and try them on different scenes. Šarūnas https://math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
Re: [darktable-user] dt under Ubuntu 16.04
On 04/28/2016 12:46 PM, Alex wrote: [...] >> on my side it is: > $ apt policy darktable > darktable: > Installiert: 2.0.3-1 > Installationskandidat: 2.0.3-1 > Versionstabelle: > *** 2.0.3-1 500 > 500 http://de.archive.ubuntu.com/ubuntu xenial/universe amd64 > Packages > 100 /var/lib/dpkg/status > > but still no go Did you try to look for extra libexiv libraries and what is libdarktable.so using?: To find libexiv versions: locate libexiv To check what libdarktable.so is using: ldd /usr/lib/darktable/libdarktable.so | grep libexiv -- Šarūnas Burdulis http://math.dartmouth.edu/~sarunas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] dt under Ubuntu 16.04
On 04/27/2016 02:53 PM, Alex wrote: > 2016-04-27 20:39 GMT+02:00 Alex <zeita...@gmail.com>: [...] >>>>> I tried: >>>> sudo apt-get purge --auto-remove darktable >>>> sudo apt-get purge darktable >>>> sudo apt-get remove darktable >>>> and then >>>> sudo apt-get install darktable >>>> but still fails to start with the message: >>>> >>>> darktable: symbol lookup error: >>>> /usr/bin/../lib/darktable/libdarktable.so: undefined symbol: >>>> _ZN5Exiv213XmpProperties10registerNsERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEES8_ >>>> >>> Have you verified that /usr/bin/../lib/darktable/libdarktable.so is >>> actually provided by darktable package? >>> Also, are you sure you don't just happen to have self-compiled exiv2 >>> somewhere? >>> >>> Sounds like mixing C++ ABI versions. C++ FTW. >>> >>> Roman. >>> >> There is /usr/lib/darktable/libdarktable.so (Just noticed it's /../, not /.../) To find libexiv versions: locate libexiv To check what libdarktable.so is using: ldd /usr/lib/darktable/libdarktable.so | grep libexiv -- Šarūnas Burdulis http://math.dartmouth.edu/~sarunas darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] dt under Ubuntu 16.04
On 04/26/2016 03:57 PM, Alex wrote: > I tried uninstalling and installing from both ppa and official repo. I > am getting: > darktable: symbol lookup error: > /usr/bin/../lib/darktable/libdarktable.so: undefined symbol: > _ZN5Exiv213XmpProperties10registerNsERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEES8_ > and dt is not starting. libdarktable.so from darktable in Xenial is /usr/lib/darktable/libdarktable.so. Judging by the error message, in your case the library is loaded from some other install, and it looks like it might be not finding libexiv. You may want to cleanly uninstall and remove all previous darktable installs and then reinstall. It should pull in libexiv if it is missing, as that is one of the package dependencies. darktable 2.0.3 from the official Xenial repository works fine here on 16.04 (albeit on Kubuntu). ~# apt-cache policy darktable darktable: Installed: 2.0.3-1 Candidate: 2.0.3-1 Version table: *** 2.0.3-1 500 500 http://archive.ubuntu.com/ubuntu xenial/universe amd64 Packages 100 /var/lib/dpkg/status -- Šarūnas Burdulis http://math.dartmouth.edu/~sarunas signature.asc Description: OpenPGP digital signature
