system
December 21, 2022, 1:46pm
1
We’re proud to announce the new feature release of darktable, 4.2.0!
The github release is here: https://github.com/darktable-org/darktable/releases/tag/release-4.2.0 .
As always, please don’t use the autogenerated tarball provided by
github, but only our tar.xz file. The checksums are:
$ sha256sum darktable-4.2.0.tar.xz
18b0917fdfe9b09f66c279a681cc3bd52894a566852bbf04b2e179ecfdb11af9 darktable-4.2.0.tar.xz
$ sha256sum darktable-4.2.0_x86_64.dmg
89308cb7222ec0d0b539771dd90284e98b35bae4b53c8f3aa1b2feb861536b82 darktable-4.2.0_x86_64.dmg
$ sha256sum darktable-4.2.0_arm64.dmg
7cbc03cc0e9a4a1769e4321815ed99384de8c063990eb5d746ffbd7599e32193 darktable-4.2.0_arm64.dmg
$ sha256sum darktable-4.2.0-win64.exe
748e617d36f810890fcc7829fdf0e9999da2906b20293a2ac919808f80709e27 darktable-4.2.0-win64.exe
When updating from the stable 4.0.x series, please bear in
mind that your edits will be preserved during this process, but the new
library and configuration will no longer be usable with 4.0.x.
This is a companion discussion topic for the original entry at https://www.darktable.org/2022/12/darktable-4.2.0-released/
66 Likes
mikae1
December 21, 2022, 3:01pm
2
Congratulations on the release!
4 Likes
A very big bravo for this fabulous software, a very nice team work, Thank you, thank you, …
Greetings from Brussels,
3 Likes
EspE1
December 21, 2022, 3:51pm
4
WOW, I didn’t know it was Christmas Day already! Hurrah!!
Thank you very much to all developers and those providing various supporting actions with manuals, notes, web-pages, packing, translation, QA, answering questions here in forum and what not!
6 Likes
elstoc
December 21, 2022, 4:10pm
5
We changed our release dates to Summer and Winter Solstice, so you get it a bit before Christmas now.
13 Likes
Thanks everyone! This looks like a great release.
6 Likes
Thanks for the release. But I am unable to open Sony ILCE-7RM5 RAW files. I thought it was supported.
difrkaguilar
December 21, 2022, 6:03pm
8
An endless thank you to all darktable developers for create this great and addictive piece of software, this release comes with excellent and unique tools for develop our raw files.
6 Likes
Bbawt
December 21, 2022, 6:17pm
9
Wonderful, and many thanks to all the developers and volunteers!
Zero thanks to MS that still tries desperately to keep you from installing it
1 Like
bridgemen
December 21, 2022, 6:19pm
10
Had an upgrade on linux mint to 21.1 cinnamon today, and the new 4.2 .0 darktable release is working as it should on it , so tnx very much to all devs.
1 Like
priort
December 21, 2022, 6:43pm
11
Maybe not compressed raws Is that what you tried?? I am just guessing
bastibe
December 21, 2022, 6:53pm
12
It is absolutely amazing how every new version of Darktable has me giddy like my kids on Christmas Eve!
This has become a rare occurrence in my life as a jaded software developer. So thank you, for putting the joy back into software!
7 Likes
pphoto
December 21, 2022, 7:23pm
13
Thanks to all for this great piece of software
Thanks about the 4.2.0 for everyone involved! Releasing it before Christmas was a nice bonus, since now I have a couple of evenings time to play with the new version before I’ll be away visiting family
2 Likes
Donatzsky
December 21, 2022, 8:18pm
15
system:
Introduced balanced “OpenCL vs CPU” tiling. This makes it possible to use CPU tiling if there is not enough memory on the OpenCL card (which would require a lot of tiles to be handled on the card). In the end, the large number of tiles plus the overlapping area would make the use of the OpenCL code path slower (or much slower) than handling the image without tiling on the CPU.
Is this automatic? I looked through the settings, but couldn’t find anything.
g-man
December 21, 2022, 8:38pm
16
darktable-org:master ← jenshannoschwalm:balanced_tiling
opened 06:37AM - 24 Sep 22 UTC
**Rationale**: On many machines we find pretty fast multicore cpus and a lot of … system ram but a significantly gpu memory with somewhat resrticted number crunching power. There is no problem using OpenCL code in modules as long as that doesn't lead to excessive tiling with large overlapping areas. Such a situation was rare, now we have as examples D&S or laplacian highlights which might behave bad and it would have been better to use cpu code instead. One related issue is #11955
Suggested "solutions" like per-module settings (whatever the complexity in the UI might be) are not good at all as runtime defined parameters are not taken into account.
**proposed solution**: follows two assumptions:
1. we have a proper benchmarking result measuring raw processing powers cpu vs cl device. (The results we already have in opencl.c are by far not good enough as they don't reflect real-world performance.)
2. the "overall processed memory" weighed by the tested performance in 1) is a good estimation of the resulting workload.
This pr introduces a decision in the opencl pixelpipe code
1. If the module can't be processed without tiling
2. and we have a user defined device specific advantage over cpu
3. we estimate memory consumption cpu vs gpu
4. and thus decide if cpu code has an advantage over gpu
The per cl device **advantage** is found as the last entry in the cldevice specific conf key, the conf key is extended automatically and the advantage is set to zero disabling the check for now.
**How to** set up the advantage hint?
Measure performance (-d perf) in laplacian highlights after setting "diameter of reconstruction" to something high. While doing so make sure (by -d tiling) that the cl code doesn't tile. Check for execution times with opencl on and off. The advantage option can now be set to approx (cpu-time / gpu-time)
The code presented a) adds some complexity to tiling code b) does not lead to performance drops as long as the new parameter is untouched or properly set c) leads to performance gains in some modules especially if your graphics card is not a "power horse" and you have a fast cpu and lots of ram.
Edit: third set of opencl related changes for 4.2
1. i checked the existing benchmark code, pre-heating cpu & gpu gives somewhat better results but not of real significance. Enlarging size of data in the benchmark is not always possible for small cards.
It should be j in the string of opencl config values and it is currently disabled by default.
ya lossless. Looks like Darktable to does not support lossless compression and has no intention of adding it. I don’t want to loss information by shooting in compressed and I can’t just shoot in uncompressed raw, due to long term storage. Will have to look into converting to TIFF, or just moving to lightroom.
mikae1
December 21, 2022, 9:13pm
19
@The_year1500 , converting to lossless compressed DNG is an option.
priort
December 21, 2022, 9:25pm
20
DT relies on rawspeed so if and when this gets supported you will be able to use that. DNG should be fine if you want access to the DT software. I don’t think those compressed formats are straightforward and open source folks wont get the same access to documents for support as Adobe would…
Curious will ART open them??
So stoked!
