Immich 3.4 Photo and Video Management Solution Adds JPEG XL Support

Immich 3.4 adds JPEG XL support, mobile tagging, OCR-powered automation, and security improvements to the self-hosted photo management platform.

Immich, the open-source, self-hosted photo and video management solution, has released version 3.4 with support for JPEG XL, a modern image format that compresses efficiently while maintaining image quality. The update adds JPEG XL decoding to the mobile app and allows the server to use JXL as a target image format.

Moreover, mobile users can now work with tags directly in the app, making it easier to organize photos without switching to the web interface. There is also an optional smoother timeline scrubbing experience, and image zooming now respects the original image’s pixel dimensions.

On the automation side, Immich 3.4 expands its workflow system with two new EXIF metadata filters: missing and present. These allow workflows to check whether a metadata field exists instead of relying on pattern matching.

Additionally, workflows can now run after OCR processing finishes. Combined with a new OCR text filter, this lets users automatically organize images based on detected text. For example, scanned receipts can be placed into a dedicated album when specific keywords are recognized.

On the security side, developers have tightened access controls across several areas, including shared links, partner timelines, duplicate detection, and album membership information.

Among the fixes, shared-link viewers can no longer access album member details, duplicate checks now require appropriate upload permissions, plugin host validation has been strengthened, and the Content Security Policy no longer includes unsafe-inline.

Other security changes improve OAuth cookie handling, prevent symbolic-link traversal, and sanitize variables used in email templates and EXIF-based storage paths.

Regarding performance, Immich 3.4 reduces unnecessary processing when loading albums and shared links, as the server no longer retrieves every asset in an album for several common operations, benefiting users managing large photo collections.

Database indexing has been improved for synchronization, while RKNN-based machine-learning models use less memory. Backup changes include parallel scanning with Restic and support for faster database backup compression using pigz, when installed.

Finally, the release addresses several mobile issues, including a bug that could move assets to the trash even after deletion was canceled, problems starting the app without an internet connection, and incorrect display of 10-bit HEIC and AVIF images on iOS.

For more information, see the changelog.

Bobby Borisov

Bobby Borisov

Bobby, an editor-in-chief at Linuxiac, is a Linux professional with over 20 years of experience. With a strong focus on Linux and open-source software, he has worked as a Senior Linux System Administrator, Software Developer, and DevOps Engineer for small and large multinational companies.

Leave a Reply

Your email address will not be published. Required fields are marked *