current OSX Build

Software darktable
990

since nightly builds supports coreML here the complimentary non coreML build for older macOS

darktable-5.5.0+619~g8dbb8eebe8_arm64-coreml.dmg (macOS >= 11.3)

991

current master 5.5.0 AI enabled build + preliminary Canon R6mIII support (based on Add Canon EOS R6 Mark III support by JuanPabloZambrano Ā· Pull Request #745 Ā· LibRaw/LibRaw Ā· GitHub) + a couple of RF lensfun data from: Kameratrollet

for documentation of the AI masking have a look at [AI] AI object mask tool by andriiryzhkov Ā· Pull Request #20378 Ā· darktable-org/darktable Ā· GitHub

There are some impacts on macos support. A build which makes use of coreML requires at least macOS 14.4. (you can use the nightly builds available on darktable github)
For older macOS coreML isnā€™t available an so the AI processing will be done by CPU.
The intel build also canā€™t make use of coreML and AI dependency onnxruntime requires at least 10.15.

darktable-5.5.0+739~g959c8b4160_arm64.dmg (macOS >= 11.3 no coreML )
darktable-5.5.0+739~g959c8b4160_x86_64.dmg (macOS >= 10.15 no coreML)

you might run it with an in-memory library via terminal:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --library :memory:
or use a separate config directory:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --configdir ~/.config/darktable_test

also keep in mind that if you donā€™t disable writing sidecar files these can be messed up

both packages are unsigned, you might need to allow execution via terminal: " xattr -c <path/to/application.app> "

Release notes for 5.5.0 release: darktable/RELEASE_NOTES.md at master Ā· darktable-org/darktable Ā· GitHub

difference to released 5.4.1 see: Comparing release-5.4.1ā€¦master Ā· darktable-org/darktable Ā· GitHub

3 Likes
992

so we have to download coreML supported version from Releases Ā· darktable-org/darktable Ā· GitHub in the future?

993

yes, for coreml you need to use the official arm64 nightly builds. coreml isnā€™t supported for older macos than 14.4

994

current master 5.5.0 AI enabled build + preliminary Canon R6mIII support (based on Add Canon EOS R6 Mark III support by JuanPabloZambrano Ā· Pull Request #745 Ā· LibRaw/LibRaw Ā· GitHub) + a couple of RF lensfun data from: Kameratrollet

for documentation of the AI masking have a look at [AI] AI object mask tool by andriiryzhkov Ā· Pull Request #20378 Ā· darktable-org/darktable Ā· GitHub

There are some impacts on macos support. A build which makes use of coreML requires at least macOS 14.4. (you can use the nightly builds available on darktable github)
For older macOS coreML isnā€™t available an so the AI processing will be done by CPU.
The intel build also canā€™t make use of coreML and AI dependency onnxruntime requires at least 10.15.

darktable-5.5.0+806~ge10c1b455b_arm64.dmg (macOS >= 11.3 no coreML )
darktable-5.5.0+806~ge10c1b455b_x86_64.dmg (macOS >= 10.15 no coreML)

you might run it with an in-memory library via terminal:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --library :memory:
or use a separate config directory:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --configdir ~/.config/darktable_test

also keep in mind that if you donā€™t disable writing sidecar files these can be messed up

both packages are unsigned, you might need to allow execution via terminal: " xattr -c <path/to/application.app> "

Release notes for 5.5.0 release: darktable/RELEASE_NOTES.md at master Ā· darktable-org/darktable Ā· GitHub

difference to released 5.4.1 see: Comparing release-5.4.1ā€¦master Ā· darktable-org/darktable Ā· GitHub

2 Likes
995

current master 5.5.0 AI enabled build + preliminary Canon R6mIII support (based on Add Canon EOS R6 Mark III support by JuanPabloZambrano Ā· Pull Request #745 Ā· LibRaw/LibRaw Ā· GitHub) + a couple of RF lensfun data from: Kameratrollet

for documentation of the AI masking have a look at [AI] AI object mask tool by andriiryzhkov Ā· Pull Request #20378 Ā· darktable-org/darktable Ā· GitHub

There are some impacts on macos support. A build which makes use of coreML requires at least macOS 14.4. (you can use the nightly builds available on darktable github)
For older macOS coreML isnā€™t available an so the AI processing will be done by CPU.
The intel build also canā€™t make use of coreML and AI dependency onnxruntime requires at least 10.15.

darktable-5.5.0+879~gdb8be9b495_arm64.dmg (macOS >= 11.3 no coreML )
darktable-5.5.0+879~gdb8be9b495_x86_64.dmg (macOS >= 10.15 no coreML)

you might run it with an in-memory library via terminal:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --library :memory:
or use a separate config directory:
/Volumes/darktable/darktable.app/Contents/MacOS/darktable --configdir ~/.config/darktable_test

also keep in mind that if you donā€™t disable writing sidecar files these can be messed up

both packages are unsigned, you might need to allow execution via terminal: " xattr -c <path/to/application.app> "

Release notes for 5.5.0 release: darktable/RELEASE_NOTES.md at master Ā· darktable-org/darktable Ā· GitHub

difference to released 5.4.1 see: Comparing release-5.4.1ā€¦master Ā· darktable-org/darktable Ā· GitHub