feat: jxl browser support detection

[nicosemp]

Description

Besides Safari, other browsers support (or are adding support) to JPEG XL images. One example is Zen Browser, which inherited the flag from Firefox's upstream and enabled it by default.
This means that adding support for more and more browsers one by one can be problematic.

At the same time, there is an easy and fast way to detect whether a browser supports JPEG XL images. It involves creating an Image object, assigning a minimal image/jxl payload to its src and monitoring the async loading with an event listener.

Fixes # (issue)

How Has This Been Tested?

• Run this test code from any browser's console:

const img = new Image();
img.addEventListener('load', () => alert('yes'));
img.addEventListener('error', () => alert('no'));
img.src = 'data:image/jxl;base64,/woIAAAMABKIAgC4AF3lEgA=';

It will show an alert with:

• yes if supported, expected in these browsers as of 2026-01-28:
  • Safari
  • Zen Browser
  • Firefox, with the image.jxl.enabled set to true in about:config. Otherwise no
• no if not supported

Screenshots (if appropriate)

Checklist:

• I have performed a self-review of my own code
• I have made corresponding changes to the documentation if applicable
• I have no unrelated changes in the PR.
• I have confirmed that any new dependencies are strictly necessary.
• I have written tests for new code (if applicable)
  • I'm not sure how a test could check this change, but I'm happy to add any specific test if required
• I have followed naming conventions/patterns in the surrounding code
• All code in src/services/ uses repositories implementations for database calls, filesystem operations, etc.
• All code in src/repositories/ is pretty basic/simple and does not have any immich specific logic (that belongs in src/services/)

Please describe to which degree, if any, an LLM was used in creating this pull request.

No LLM.

[mertalev]

I think the approach of testing support is reasonable, but the implementation doesn't seem particularly robust.

[meesfrensel]

Calling this function twice or more would be a bug, right? Is it possible to prevent that? I don't know if it's valid TS syntax, but maybe something like (() => { /* logic */ })() could work. Otherwise, I think it makes sense to add an assertion at the start of the function that asserts !supportedImageMimeTypes.contains('image/jxl') to at least surface potential bugs asap.

[nicosemp]

Calling this function twice or more would be a bug, right?

@meesfrensel if this whole file is executed multiple times for any reason, then supportedImageMimeTypes is also redefined at every execution, without image/jxl. The function runs after the first definition so I don't think it could ever be executed on an already populated Set.

I could add a guard anyway like so, but I feel it would be redundant:

// before the function
let isJxlCheckDone = false;

// inside the function, at the beginning
if (isJxlCheckDone ) return;
isJxlCheckDone = true;

[nicosemp]

@mertalev do you need any other action from me?
I have the HEIC version almost ready, but I'll rebase after this is merged before creating the PR.

[mertalev]

Nope! This is just waiting for us to be done with hot fixes for this release.