• 1
  • 88
  • 89
  • 90(current)
  • 91
  • 92
  • 244
Linux Radeon OSS with vdpau (howto)
Quote:If I also have the DVI connection to my monitor connected, no DTSHD or DolbyTrueHD, gets to the AVR. But all the other formats do passthrough.

That's a bug. I have seen that on the bugtracker. It seems only one EDID / handshake is used. Could you disable the other monitor via xrandr? Does dts-hd magically start working then?
Check that one: https://bugs.freedesktop.org/show_bug.cgi?id=72648

Concerning the 16..235 is not that easy, sadly - a bit read up here: http://forum.xbmc.org/showthread.php?tid=179163

Edit: Try overwriting the EDID for both monitors - see howto on how to do it for one monitor and make the same for the other
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
Good Evening,

I got a problem with my analog sound output.

For the devices named in the xbmc-logfile, the sound with "Device 1" (HDMI) is clean, but when using the "Device 2" (ALC898 Analog) the sound shows some small drops and is not clean.
As I do not want to use the HDMI sound all the time (TV shall be off while only hearing music) I need the analog output to be clean.

First installed all according to the first page, now installed Kernel 3.13_RC3 from the ubuntu ressources. I unfortunately cannot find any difference regarding the sound issue.

The same for removing "radeon.audio=1" from the grub command.

Here are the logs:

dmesg http://paste.ubuntu.com/6574000/
x-server http://paste.ubuntu.com/6574002/
xbmc http://paste.ubuntu.com/6574021/
vdpau http://paste.ubuntu.com/6574027/
dpkg -l http://paste.ubuntu.com/6574029/


"glxinfo | grep -i interop" does give an error: "Error: unable to open display"

Did anyone find this strange sound errors as well?

Mainboard is a http://www.asus.com/Motherboards/E2KM1IDELUXE/

Any idea what could be the problem?

Thanks in advance!
Cannot see anything in the logs. Despite your analog device only has one samplerate, 48khz.

and it's DISPLAY=:0 glxinfo |grep -i interop

Edit: Not sure, what you mean with clean. If you use a heaphone jack, better use the onboard plug. Btw. this Analog Sound issue is not related in any form to that howto.
Edit2: Make sure to disable eventually interfering Microphone ports.
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
I had to switch off the "auto-mute mode" in the alsamixer to get rid of the sound drops.

No everything is working very well.

Thanks a lot for the idea with the "interfering ports" and for this very good howto!
just wanted to report that with 3.13.RC3, wsnipex test ppa; and fernetmenta, my mpegs no longer crash, YAY!!!!
/Radeon 4550
/hadn't checked in a while so don't know what fixed it
Ubuntu 13.10 | 10TB | myth .27 - HVR2250 - HVR1250 | XBMC Gotham | Radeon 4550
Linux Mint 15 | XBMC Gotham | IntelHD4000
Openelec Custom | XBMC Gotham | Raspberry PI
Cyanogen 11 | XBMC Gotham | Samsung GSIII
The upstart script I posted in this howto has a big problem.

Upstart does not read /etc/security/limits.conf and also limit nice -1 -1 it ignores cause of the way we start xbmc.

The simplest solution I have found for now, is using slim login manager.

If someone else has a better idea, please state.
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
Puh, it was easier than I thought, thx to non existent documentation.

Add limit nice 21 21 to your /etc/init/xbmc.conf directly under the respawn, so that it looks like:

Code:
# xbmc-upstart
# starts XBMC on startup by using xinit.
# by default runs as xbmc, to change edit below.
env USER=xbmc

description     "XBMC-barebones-upstart-script"
author          "Matt Filetto"

start on (filesystem and stopped udevtrigger)
stop on runlevel [016]

# tell upstart to respawn the process if abnormal exit
respawn
respawn limit 10 5
limit nice 21 21

script
exec su -c "xinit /home/xbmc/startxbmc.sh :0" $USER
# the following two are to get an idea, if you want to user a window manager
#   exec su -c "xinit /usr/bin/fluxbox :0" $USER
end script
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
(2013-11-30, 21:27)metheos Wrote: My ASUS E35M1-I DELUXE required different settings in advancedsettings.xml to get the cpu and gpu temperatures correctly.

sensors output:
Code:
radeon-pci-0008
Adapter: PCI adapter
temp1:        +53.0°C  (crit = +120.0°C, hyst = +90.0°C)

k10temp-pci-00c3
Adapter: PCI adapter
temp1:        +53.1°C  (high = +70.0°C)
                       (crit = +100.0°C, hyst = +97.0°C)

advancedsettings.xml commands:
Code:
<cputempcommand>sensors|sed -ne '1h;1!H;${g;s/radeon.*k10temp.*temp1: \+[-+]\([0-9]\+\).*/\1 C/p}'</cputempcommand>
<gputempcommand>sensors|sed -ne '1h;1!H;${g;s/radeon.*temp1: \+[-+]\([0-9]\+\).*/\1 C/p}'</gputempcommand>

couple of things i found -- syntax:
my sensors were found in reverse, so I had to change things around to read like this:
Code:
<gputempcommand>sensors|sed -ne '1h;1!H;${g;s/k10temp.*temp1: \+\([-+][0-9]\+\).*/\1 C/p}'</gputempcommand>
<cputempcommand>sensors|sed -ne '1h;1!H;${g;s/k10temp.*radeon.*temp1: \+\([-+][0-9]\+\).*/\1 C/p}'</cputempcommand>

you'll also notice the correction with \(+-)
so that it will correctly display the temperature if it drops below 0 degrees celsius
Linux Mint 18 LTS 64-bit - Kodi 17 Beta6
Odroid-C2 - Libreelec v7.90.009
Yeah - if my cpu temp would drop below 0 degrees ...
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
(2013-12-02, 09:24)fritsch Wrote: @metheos: No clue. I don't see anything in the logs. Never seen something like this in that combination.

I'm happy to report that I installed 3.13 rc3 and now my passthrough audio is working great for all supported formats.
Nice - but still curious, as normally AC3 and DTS are really comfortable formats, that every milk bottle can play. Make sure to adjust the init script, that will care for Underruns.
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
(2013-12-15, 18:51)dubsides Wrote: just wanted to report that with 3.13.RC3, wsnipex test ppa; and fernetmenta, my mpegs no longer crash, YAY!!!!
/Radeon 4550
/hadn't checked in a while so don't know what fixed it

Were you actually getting GPU Lockups though? I've looked through your dmesg output on the bug report and it looked like a different issue...

I'll try an updated stack later, regardless, see if it helps at all!
Here is a small recepice for debian. Over the week-end everything needed has been added. Its quite easys to get a working config on debian:

1) Install xserver-xorg-core from sid (1.14.5-1)
2) Install open source radeon driver from sid (7.2.0)
3) install mesa source from experimental (10.0.0.0-1)
4) Apply the following patch to enable gallium3d for mesa from bug <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=656719> : <http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=220;filename=mesa_9.2.2-1.1.debdiff;att=1;bug=656719>
5) rebuild from source
6) take the 3.13.-rc4 kernel code and upgrade you favorite kernel
7) get xbmc source from fernetmenta git tree
8) configure with ./configure --enable-external-libraries --disable-external-ffmpeg --disable-vaapi --disable-crystalhd --enable-pulse. the disable-external-ffmpeg is needed for any ffmpeg version later than 2.1 (I opened a bug already but no fix yet. 2.0.2 is ok)

EDITED: Note I use pulse because I still use analog audio amplifier and need to downmux anyway. If you want to use passtrough or ActiveAE, do not enable pulse.
EDITED 2: Now the deb-multimedia package should work out of the box as the bug that was causing XBMC to misbehave with ffmpeg 2.1.1. has been fixed in ffmpeg git and deb-multimedia packages.

Enjoy
don't tell people to enable pulse when we always advise against that.
Wsnipex is fully right here. We don't have a pulseaudio sink in xbmc yet. So choosing pulse here will a) disable ActiveAE b) disable passthrough completely. Futerhmore the pulse code was not tested nor futher developed (despite some song changing bugs I fixed 6 months ago).

Edit: hacks with pavucontrol I don't consider as sane workarounds.
First decide what functions / features you expect from a system. Then decide for the hardware. Don't waste your money on crap.
  • 1
  • 88
  • 89
  • 90(current)
  • 91
  • 92
  • 244

Logout Mark Read Team Forum Stats Members Help
Radeon OSS with vdpau (howto)26