[Webkit-unassigned] [Bug 252181] New: 3D transform causes backdrop-filter effect to be applied incorrectly

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Mon Feb 13 12:03:19 PST 2023


https://bugs.webkit.org/show_bug.cgi?id=252181

            Bug ID: 252181
           Summary: 3D transform causes backdrop-filter effect to be
                    applied incorrectly
           Product: WebKit
           Version: Other
          Hardware: Unspecified
                OS: Linux
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: CSS
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: ana.tudor.lhnh at gmail.com

Created attachment 464973

  --> https://bugs.webkit.org/attachment.cgi?id=464973&action=review

Comparison between correct result in Chrome and the result in Epiphany.

How to reproduce: set both a 3D transform and a backdrop-filter on an element.

Live test: https://codepen.io/thebabydino/pen/oNMLELr

Live test with plane intersection: https://codepen.io/thebabydino/pen/vYzBPgd

I've only tested via Epiphany, but it's been confirmed it also happens in actual Safari - see replies with screenshots:

https://twitter.com/anatudor/status/1609985403251216387

https://mastodon.social/@anatudor/109621114959498459

Perhaps also relevant/ somewhat related to this other bug https://bugs.webkit.org/show_bug.cgi?id=249952

Another demo which looks broken because of this https://www.cssscript.com/demo/glassmorphism-credit-card/ - this demo in particular only uses `backdrop-filter` without a prefix so you're not going to see any blur effect at first, but if you add the prefix via DevTools, you're going to see the blurred backdrop area is incorrect during the animation/ if you also add a `transform: rotate(45deg)` on the `.card` element.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20230213/0457dc34/attachment.htm>


More information about the webkit-unassigned mailing list