I've started to chase this through Team Kodi contacts. We have direct chat access to several people, one of whom has been copied into the ticket. These things take time and (us) trying to escalate them is often counterproductive.
Intel Alder Lake 2160p @ 23.976 Hz passthrough HD Audio dropouts (i7-1270p/N100)
-
sky42 -
December 18, 2023 at 1:49 PM -
Thread is Unresolved
-
-
Hence why you include the nice workaround into libreelec for now...
But it's awesome to hear that there is contact with the Intel team and that a proper fix will hopefully arrive sooner rather than later :). -
seeing is believing
-
how's it going? Does Intel even remember what an iGPU is/does?
-
No update so far on the Intel case.
-
No update so far on the Intel case.
Btw, immense thanks on all the work you've done so far on this ticket. Your build/script really make LibreEelec usable on modern Intel hardware (at least for those who like to play 4K), and LibreElec perfectly fits what i wanted to do with my Intel Nuc.
I hope we'll get a more permanent fix somehow in the near future, but at least you've given some way to get it all to work :). -
I have a Beelink S12 with N100 and have the same audio dropouts mentioned in this thread. I tried the script in post 46, fired up a movie, and the audio was completely gone and the video was sparkly. So I rebooted, fired up a movie, and the audio drop outs are still there.
So I ran the script again, and it says that the bit is not set. Does it not persist through reboots? -
-
modetest |grep connected
185 184 connected HDMI-A-1 1600x900 45 184
194 0 disconnected HDMI-A-2 0x0 0 193
198 0 disconnected DP-1 0x0 0 197
Hashbrown:~ #
So, I poked through modetest, I can see that 185 max bpc is set to 10, however I still get the audio dropouts. Dont know why it says 1600x900, as it is 2160p. -
I've started to chase this through Team Kodi contacts. We have direct chat access to several people, one of whom has been copied into the ticket. These things take time and (us) trying to escalate them is often counterproductive.
So, any updates?
-
IMO it's worth trying the new Intel Xe driver.
CONFIG_DRM_XE=m
i915.force_probe=!xxxx xe.force_probe=xxxx (disable i915 and enable Xe driver, xxxx=Intel PCI graphics ID)
I'm running this on TGL and kernel 6.9. No idea if it fixes anything, I don't use audio passthrough.
-
Thanks for the info.
It will be some time until I am free to test that.
-
Anyone know if there are any updates on this? Is Intel still trying to figure it out?
-
https://gitlab.freedesktop.org/drm/i915/kernel/-/issues/10199 <= no updates other my nagging, but we'll keep doing that
-
Just checked my latest LE12 build from today with kernel 6.6.34 and the problem is still there.
-
I managed to build a LE12 with kernel 6.9 from master that you can download here https://sky42.libreelec.tv/release/12.0/G…020/kernel-6.9/.
This is with as less as possible changes. Just 6.9.5 and a systemd patch. No encryption or lvm yet.
smp thanks for the XE driver tip. I did just do that and it does not solve the problem.
All of you that want to test the XE driver with kernel 6.9
- find your PCI id with lspci -nn |grep -Ei 'VGA|DISPLAY'
- for me that is 00:02.0 VGA compatible controller [0300]: Intel Corporation Alder Lake-N [UHD Graphics] [8086:46d1]
- set kernel boot parameter like i915.force_probe=!46d1 xe.force_probe=46d1
- disable 10 bit workaround touch /storage/.dont-activate-10-bit-max (the script will fail because it can not handle xe)
- reboot
- check with lspci -vs 00:02.0 and look for Kernel driver in use: xe -
JFYI
- issue update https://gitlab.freedesktop.org/drm/i915/kerne…99#note_2469257
- Xe driver has the same problem and so far i could not test 10 bit with Xe (dont know how to set 10 bit) -
Jeez... the Intel guy can't reproduce the issue. How come.
Maybe it's a HDMI handshake thing that only happens with AVRs? Oh well, at least we have the script.Thinking about it: How the Intel guy expects to test HD audio passtru without an AVR?
-