• 1
  • 141
  • 142
  • 143(current)
  • 144
  • 145
  • 168
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)
(2015-11-15, 22:08)popcornmix Wrote:
(2015-11-15, 21:35)Leopold Wrote: HD live TV seems much improved Smile

I'm using #1113 and I'm not seeing any green pixelation or stuttering with advanced deinterlace. My backend is on a separate Pi though.

Can you identify when this happened?
It was possibly with 1111b which increased the usb priority compared to deinterlace code, but it would be nice to confirm if that is definitely the case.

I haven't had much time to test this but I have gone back to an old build and still had perfect live tv, so I now believe that the improvement is down to recent changes in tvheadend. I am compiling tvheadend master branch. I don't think the usb priority would affect my setup since the tuner is on a separate pi? Although could the high ethernet traffic for HD live tv have the same effect since (IIRC) it uses the usb bus?
Leopold's Repository: Home of LibreELEC Dev Updater ...
(2015-11-18, 02:32)Milhouse Wrote:
(2015-11-18, 02:08)motorherz Wrote: Hi, I still have no sound with 1:1 3d.ISO files.
example: edge of tomorrow, framepacked 3d with DTS-HD MA

older versions (1106 for example) are able to play this image in 2d with sound, but not in 3d...

I'm using multichannel 5.1 PCM decoding to my preamp for all HD audio codecs

Has it always been like this, with 3D? What about the official OpenELEC 6.0 release - sound with 2D and 3D? If you can identify the build when the 3D audio stops, that would be good to know...


I tried some older releases, 1021 up to 1028b. the framepacking method is available since the 1023 (I think, this was the first release with this option). All these releases are not working with the framepacking method in this 3d.ISO. When I select 2d, it works fine.

But, today, I installed, the 1117 release. And it works!
I tried 1:1 3d.iso files with DTSHD-MA and TrueHD. It worked fine.
examples:
edge_of_tomorrow (3d.iso DTSHD-MA)
finding_nemo (3d.iso TrueHD)

What did you changed since 1116?
should I test some earlier releases for you to find out, when the sound stops? Or did you allready found out the reason for the silence?
(2015-11-18, 13:26)motorherz Wrote: What did you changed since 1116?

It's all in the release notes. There were some AE (Audio Engine) changes which may account for the improved behaviour.

(2015-11-18, 13:26)motorherz Wrote: should I test some earlier releases for you to find out, when the sound stops?

Knowing when the problem started would certainly help improve any understanding of why it had been broken, and avoid a repeat.

(2015-11-18, 13:26)motorherz Wrote: Or did you allready found out the reason for the silence?

Don't think so - I suspect it's just by pure chance/coincidence that a code change in #1117 fixed this issue - I doubt anyone has had a chance to look into your original report, let alone understand why it was broken in the first place.

If you are able to identify the first build that breaks audio then that might help narrow down the precise cause.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-11-18, 08:53)Leopold Wrote: I haven't had much time to test this but I have gone back to an old build and still had perfect live tv, so I now believe that the improvement is down to recent changes in tvheadend. I am compiling tvheadend master branch. I don't think the usb priority would affect my setup since the tuner is on a separate pi? Although could the high ethernet traffic for HD live tv have the same effect since (IIRC) it uses the usb bus?

Yes, ethernet traffic uses USB bus so could be affected by USB priority change.
However I wouldn't expect a difference between actual live TV and other interlaced files (e.g. recordings).
If problems only occur with live TV then it's probably not ethernet/usb related. If there are issues with both then it could be.

If a recent tvheadend server change has fixed an issue with artefacts in stream, then that is interesting to know (i.e. want to ensure it appears in stable builds).
If you are able to identify when the improvement occurred (and so what the commit was) it could be useful for others.
@Leopold - I've been uploading daily builds of tvheadend to the addons folder - there's about a months worth of builds, if that helps.

I've also just started uploading weekly builds of vdr-addon (this doesn't change much, if at all).
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
@popcornmix

I tried latest build.
without hvs_priority=0x04000000 I get a lot of green and other artefacts
with this setting there are still but rare artefacts

I would say no improvement to last builds.
The latest VDR backend #1116 from the addon folder of Milhouse does not work on my Pi2 running #1117 release.
VDR addon #1015 works fine. Usually no buffering when zapping and 2.5s (2-3s) channel switching.
There are no artefacts neither for SD nor for HD DVB-C channels.
(2015-11-18, 23:32)herrmeier01 Wrote: The latest VDR backend #1116 from the addon folder of Milhouse does not work on my Pi2 running #1117 release.
VDR addon #1015 works fine. Usually no buffering when zapping and 2.5s (2-3s) channel switching.
There are no artefacts neither for SD nor for HD DVB-C channels.

"does not work" doesn't tell me much.

I don't support the vdr-addon, I just provide it as a courtesy. If it's not working you might want to contact the maintainer, whoever that is... the last change to vdr is PR:4441.

Please provide logs, and detailed explanation of the problem etc., maybe it will give someone a clue what is wrong.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
Sorry, actually I just want to ask if some one successfully upgrade it. Maybe this only happen at my system. Unfortunately I can not make a log right now. I try to collect as soon as possible with a clean install on my spare SD
Well, I got a time slot for testing...

The problem is that VDR will not fully create its configuration file. After several restarts of the addon or reboot the system it only contain this:

/storage/.kodi/userdata/addon_data/service.multimedia.vdr-addon/config/setup.conf
AlwaysSortFoldersFirst = 1
SetSystemTime = 0
epgsearch.SVDRPPort = 6419
OSDSkin = sttng

channels.conf is ok, Suntek driver is installed, actually it should run

Here is the debug log after a reboot of the system:
http://pastebin.com/raw.php?i=iXL8704B

I add an other log which I could create after downgrade VDR addon, getting running and upgrade again. Now I can't start Live-TV channel anymore:
http://pastebin.com/raw.php?i=zthYZycS
@Milhouse

I figured out what caused my DVB-C stick to no longer work, it is this commit: http://git.linuxtv.org/cgit.cgi/linux.gi...824b14060f which is in a later commit changed to "goto err;" i.e. I changed it to simply read:
Code:
/* error bit set? */
/*
        if ((cmd->args[0] >> 6) & 0x01) {
            ret = -EREMOTEIO;
            goto err;
        }
*/

Had diffed vanilla kernel 4.3 drivers vs. 4.1.12, blacklisted dvb-usb-dvbsky.ko and loaded my changed
modules by hand, until that commit fell out. Funny enough, now even channel switching works nicely
again in last nights build. No more need to stop live TV to avoid Kodi hanging with "Buffering..."
message, and after some seconds doing nothing at all anymore.

I also tested my isosynchronous Empia USB DVB dongle again with hvs_priority and verified setting
in command line, no cigar. That one still has trouble with dropped USB data. On RPi/RPi2 only live TV
sticks with Cypress chips that do bulk transfers appear to be stable.
Katastrophentourist
(2015-11-19, 01:49)MarkT Wrote: I figured out what caused my DVB-C stick to no longer work, it is this commit: http://git.linuxtv.org/cgit.cgi/linux.gi...824b14060f which is in a later commit changed to "goto err;" i.e. I changed it to simply read:

As this is an upstream linux commit, you should report your findings there where the author of this commit may have suggestions.
https://www.kernel.org/pub/linux/docs/lk...-bugs.html
New OpenELEC Jarvis build #1118: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 4.3.0 #1 Thu Nov 19 01:13:28 GMT 2015 armv6l GNU/Linux

# vcgencmd version
Nov 18 2015 15:48:41
Copyright (c) 2012 Broadcom
version 0c9af7cde38e223c95c3404f468db8fc5004495c (clean) (release)

# lsb_release
OpenELEC (Milhouse) - Version: devel-20151119011232-#1118-g785752e [Build #1118]

# vcdbg log msg 2>&1 | grep DTOK
001708.841: Kernel trailer DTOK property says yes

# Kernel device tree status: Enabled

Based on tip of OpenELEC master (785752e5, changelog) and tip of XBMC master (46130fc9, changelog) with the following modifications: Build Highlights:
  1. New firmware
  2. Many OpenELEC package bumps
  3. Reverted attr/libcap/fakeroot commits due to build issues (clean build)
Build Details:
  1. Firmware (Nov 18):
    • arm_loader: Don't overwrite the /soc/ranges DT property
  2. OpenELEC:
    • Use pkg-config mechanism to synchronize compile flags and paths between VDR and VDR plugins instead of tons of patches. (PR:4441, 11 commits, 22 files changed)
    • xf86-video-nvidia: update to xf86-video-nvidia-352.63 (93949dcb)
    • xf86-video-nvidia-legacy: update to xf86-video-nvidia-legacy-304.131 (ba6d61c2)
    • projects/imx6/options: dont build RTL8192EU driver for imx6 project, this fixes #4445 (83fbb47f)
    • xf86-video-nvidia-legacy: remove old patch (b1cee411)
    • ffmpeg: update kodi patch (358440e9)
    • v4l-utils: update to v4l-utils-1.8.1 (2f60d413)
    • Pillow: update to Pillow-3.0.0 (49ad93c0)
    • simplejson: update to simplejson-3.8.1 (c590194d)
    • setxkbmap: update to setxkbmap-1.3.1 (ed0cadab)
    • libxshmfence: update to libxshmfence-1.2 (83598586)
    • ratpoison: update to ratpoison-1.4.8 (590f5ccd)
    • sqlite: update to sqlite-autoconf-3090200 (0e5eda28)
    • autoconf-archive: update to autoconf-archive-2015.09.25 (dfd78310)
    • boost: update to boost-1_59_0 (5cfc4d20)
    • fribidi: update to fribidi-0.19.7 (eabf888c)
    • glib: update to glib-2.46.2 (b67cebc2)
    • gmp: update to gmp-6.1.0 (06797d88)
    • intltool: update to intltool-0.51.0 (f9c68977)
    • swig: update to swig-3.0.7 (d072d4df)
    • libass: update to libass-0.13.0 (acb47431)
    • libbluray: update to libbluray-0.9.1 (8be039ea)
    • libgcrypt: update to libgcrypt-1.6.4 (2cf1a69e)
    • libressl: update to libressl-2.1.8 (00f19c2b)
    • elfutils: update to elfutils-0.164 (f574a01b)
    • m4: update to m4-1.4.17 (1ed818f5)
    • readline: update to readline-6.3 (56c1d955)
    • attr: add support for host-build (a5364c60)
    • libcap: add support for host-build (ce6a19e2)
    • fakeroot: update to fakeroot-1.20.2 (eee238d7)
    • libirman: update to libirman-0.5.0 (e147d69c)
    • attr: typo and fix PKG_DEPENDS_HOST (785752e5)
  3. XBMC:
    • [win32] Re-add the openssl hack, was never mean't to be removed without addin… (PR:8403, 1 commit, 1 file changed)
    • [win32] Windows has no GL any more so setting negate is obsolete (PR:8402, 1 commit, 1 file changed)
    • [win32] RenderingSystemDX::OnDeviceReset() - fire OnResetDevice event… (PR:8404, 1 commit, 1 file changed)
    • [PVR] Remove repeating timers from Recording lists (PR:8400, 1 commit, 1 file changed)
    • [PVR] Timerdialog fixes (PR:8389, 2 commits, 2 files changed)
    • [PVR] Next recording recording widget fix (PR:8386, 1 commit, 2 files changed)
    • AE Pulse: Ignore volume changes on corked sinks (PR:8396, 1 commit, 1 file changed)
    • Fix inversion of export library dialogs, fix #16317, fix #16358 (PR:8356, 2 commits, 2 files changed)
  4. Additional commits/pull requests/changes not yet merged upstream:
    • Reverted: [env] a5364c60: attr: add support for host-build (Reason: doesn't build)
    • Reverted: [env] ce6a19e2: libcap: add support for host-build (Reason: doesn't build)
    • Reverted: [env] eee238d7: fakeroot: update to fakeroot-1.20.2 (Reason: doesn't build)
    • Reverted: [env] 785752e5: attr: typo and fix PKG_DEPENDS_HOST (Reason: doesn't build)
    • Reverted: [pkg] PR:8404: [win32] RenderingSystemDX::OnDeviceReset() - fire OnResetDevice event? (Reason: needs VideoPlayer rebase)
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-11-18, 00:18)Milhouse Wrote:
(2015-11-17, 18:58)Ollo Wrote: However I have an issue that AC3 and DTS are not detected on my receiver using SPDIF which worked at least partly on OE 6.0. Means DTS was detected and AC3 not.

If you can identify the first build here that changes the DTS/AC3 SPDIF behaviour, then it will help a lot.

I did some testing using a installation from scratch without changing configuration except audio output. The TV (HDMI passtrough) and HifiBerry are connected to the receiver via SPDIF/TOSLINK.
  • OE6.0
    • ALSA passtrough: AC3 detected but no sound; DTS is OK
    • HDMI passtrough: AC3 is ok, DTS unknown as blocked by the TV
  • #1005 - as OE6
  • #1006
    • ALSA passtrough: No AC3; No DTS just Stereo PCM
    • HDMI passtrough: No AC3; No DTS just Stereo PCM

All the later builds I tested #1009, #1015, #1117 behave the same as #1006.

Thanks
New OpenELEC Jarvis build #1119: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 4.3.0 #1 Thu Nov 19 21:02:48 GMT 2015 armv6l GNU/Linux

# vcgencmd version
Nov 18 2015 15:48:41
Copyright (c) 2012 Broadcom
version 0c9af7cde38e223c95c3404f468db8fc5004495c (clean) (release)

# lsb_release
OpenELEC (Milhouse) - Version: devel-20151119210156-#1119-g223b12d [Build #1119]

# vcdbg log msg 2>&1 | grep DTOK
001691.359: Kernel trailer DTOK property says yes

# Kernel device tree status: Enabled

Based on tip of OpenELEC master (223b12df, changelog) and tip of XBMC master (64de219d, changelog) with the following modifications: Build Highlights:
  1. Bump even more packages
  2. Fix fakeroot/libcap/attr build issue
  3. Bumped connman-1.30 - please report any unusual network issues (failure to acquire lease etc.)
Build Details:
  1. OpenELEC:
    • Revert "fakeroot: update to fakeroot-1.20.2" (6a1b7409)
    • fakeroot: add buildnote (1b55e538)
    • libass: dont depend on enca anymore (4946e939)
    • remove package 'enca' (4a968f78)
    • pkg-config: update to pkg-config-0.29 (57e5f0e7)
    • lzo: update to lzo-2.09 (d0127bfe)
    • glew: update to glew-1.13.0 (eb478a6c)
    • libjpeg-turbo: update to libjpeg-turbo-1.4.2 (43d03468)
    • yasm: update to yasm-1.3.0 (16bfdd91)
    • bluez: update to bluez-5.36 (a1d59dcf)
    • ethtool: update to ethtool-4.2 (7690fe7e)
    • wireless-regdb: update to wireless-regdb-2015.10.22 (f5ac826b)
    • dialog: update to dialog-1.2-20150920 (f0f2397e)
    • hdparm: update to hdparm-9.48 (813deb3e)
    • nano: update to nano-2.4.3 (23c3197f)
    • jasper: update security patches (eebfc3d9)
    • iptables: build with shared modules, this fixes #4450 (7acf1983)
    • libcap: dont build libcap:host with attr support (98b12a0d)
    • fakeroot: update to fakeroot-1.20.2 (223b12df)
  2. XBMC:
    • [addons] dont include broken versions in available updates (PR:8373, 3 commits, 6 files changed)
  3. pvr.mythtv:
    • 3.3.9 (PR:45, 2 commits, 7 files changed)
  4. Additional commits/pull requests/changes not yet merged upstream:
    • Added: [env] PR:4455: connman: update to connman-1.30
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
  • 1
  • 141
  • 142
  • 143(current)
  • 144
  • 145
  • 168

Logout Mark Read Team Forum Stats Members Help
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)10