On screen display not displaying correct time after seek action, and seek action not applying the correct time.

[RowanDabron]

IINA v1.3.4 Build 140
macOS
Monterey v12.2.1

I have the keybindings set so control+command+left or right arrow key seek backwards 3s or forwards 3s respectively, but the behaviour is rarely correct, it consistently skips 10s (as far as the video and on screen controller are concerned), and the on screen display shows a different time to that of the OSC.
Starting from OSD, OSC, and video all being in sync and correct, the first instance of seek skips the video and OSC 10 seconds, but the OSD skips 3s (in the correct direction), on the second instance the OSC and video skip 10s again and the OSD skips 10 seconds from it's previous incorrect time and so is still out of sync with the OSC and video.
It is possible to get the OSD to skip in the opposite direction to the intended behaviour, the OSC and video. This happens when I try to skip forward 3s twice and backwards 3s once (⌃⌘→ + ⌃⌘→ + ⌃⌘←), the OSC and video skips forward 10s twice, and backwards 10s once, but the OSD skips forwards 3s, forwards 10s, then forwards 5s.

Additionally the OSD time does not update as the video is playing, which may be intended behaviour since it is usually only on screen for a very short duration (I have set it to stay on longer to observe the above issue), but it would be nice if it could update continuously as the video plays, but that is a very minor nitpick for me.

I have attached some screenshots showing the OSD being out of sync with the OSC.

[HarukaMa]

You might want to see #4766 for the inexact seek part.

[RowanDabron]

Thank you very much, that explains that part.

Am I correct in thinking the out of sync OSD and OSC are still a valid issue or should this be closed?

[low-batt]

Please leave this issue open for now. I'm uncertain as to whether the OSD behavior is the same as issue #4779 or not. There is definitely some OSD behavior that needs to be addressed.

As you can see from the comments on issue #4766 that @HarukaMa referenced we are debating changing the default seeking behavior to be exact even though that is slower because the current behavior is unexpected and confusing.

Thank you for entering such a clear and detailed report.