Kodi Community Forum
VNSI4 alpha - Printable Version

+- Kodi Community Forum (https://forum.kodi.tv)
+-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33)
+--- Forum: PVR & Live TV Support (https://forum.kodi.tv/forumdisplay.php?fid=167)
+---- Forum: VDR (https://forum.kodi.tv/forumdisplay.php?fid=169)
+---- Thread: VNSI4 alpha (/showthread.php?tid=158684)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

RE: VNSI4 alpha - FernetMenta - 2013-03-27

@ezechiel1917, thanks. I will feed the parsers with this sample. I hope that I can reproduce the issue with your sample.

@Krautmaster, I don't think so, but you'll never know until you have tried Smile

RE: VNSI4 alpha - Krautmaster - 2013-03-28

Nvidia Problem?

http://pastebin.com/n2zT4AeQ

RE: VNSI4 alpha - Nachteule - 2013-03-28

Same here, your patch does not work Sad

I added my old debug message to the code (end of function ParsePESHeader) looking a little bit what's going on: pts and dts values are going crasy if issue occours, here one example

Code:
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f48a2, m_curPTS 23c886f48a2, m_curDTS 23c886f48a2, m_Wrap 0, m_NoOfWraps 286
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f56b2, m_curPTS 23c886f56b2, m_curDTS 23c886f56b2, m_Wrap 0, m_NoOfWraps 286
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f64c2, m_curPTS 23c886f64c2, m_curDTS 23c886f64c2, m_Wrap 0, m_NoOfWraps 286
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb2e571, m_curPTS 23defb2e571, m_curDTS 23defb2e571, m_Wrap 0, m_NoOfWraps 286
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID c0, pts 1efb2789f, m_curPTS 23defb2789f, m_curDTS 23defb2789f, m_Wrap 1, m_NoOfWraps 286
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f72d2, m_curPTS 23e886f72d2, m_curDTS 23e886f72d2, m_Wrap 0, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb2f381, m_curPTS 23fefb2f381, m_curDTS 23fefb2f381, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb32bc1, m_curPTS 23fefb32bc1, m_curDTS 23fefb30191, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f80e2, m_curPTS 23e886f80e2, m_curDTS 23e886f80e2, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb30fa1, m_curPTS 23fefb30fa1, m_curDTS 23fefb30fa1, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb31db1, m_curPTS 23fefb31db1, m_curDTS 23fefb31db1, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f8ef2, m_curPTS 23e886f8ef2, m_curDTS 23e886f8ef2, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb355f1, m_curPTS 23fefb355f1, m_curDTS 23fefb32bc1, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886f9d02, m_curPTS 23e886f9d02, m_curDTS 23e886f9d02, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID c0, pts 1efb2ab3f, m_curPTS 23fefb2ab3f, m_curDTS 23fefb2ab3f, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb339d1, m_curPTS 23fefb339d1, m_curDTS 23fefb339d1, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID bd, pts 0886fab12, m_curPTS 23e886fab12, m_curDTS 23e886fab12, m_Wrap 1, m_NoOfWraps 287
Mar 28 01:10:18 [10349] VNSI: ParsePESHeader: ID e0, pts 1efb347e1, m_curPTS 23fefb347e1, m_curDTS 23fefb347e1, m_Wrap 1, m_NoOfWraps 287
Mar

m_NoOfWraps grows extremely fast.

If I watch the instant recording, it looks like this

Code:
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID c0, pts 1efc541ff, m_curPTS 1efc541ff, m_curDTS 1efc541ff, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc5aed1, m_curPTS 1efc5aed1, m_curDTS 1efc5aed1, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc5bce1, m_curPTS 1efc5bce1, m_curDTS 1efc5bce1, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID c0, pts 1efc5749f, m_curPTS 1efc5749f, m_curDTS 1efc5749f, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc5f521, m_curPTS 1efc5f521, m_curDTS 1efc5caf1, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc5d901, m_curPTS 1efc5d901, m_curDTS 1efc5d901, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc5e711, m_curPTS 1efc5e711, m_curDTS 1efc5e711, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc61f51, m_curPTS 1efc61f51, m_curDTS 1efc5f521, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc60331, m_curPTS 1efc60331, m_curDTS 1efc60331, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID c0, pts 1efc5a73f, m_curPTS 1efc5a73f, m_curDTS 1efc5a73f, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc61141, m_curPTS 1efc61141, m_curDTS 1efc61141, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc64981, m_curPTS 1efc64981, m_curDTS 1efc61f51, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc62d61, m_curPTS 1efc62d61, m_curDTS 1efc62d61, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc63b71, m_curPTS 1efc63b71, m_curDTS 1efc63b71, m_Wrap 1, m_NoOfWraps 0
Mar 28 01:10:32 [10466] VNSI: ParsePESHeader: ID e0, pts 1efc673b1, m_curPTS 1efc673b1, m_curDTS 1efc64981, m_Wrap 1, m_NoOfWraps 0

Everything is ok, same channel almost the same time

Please notice that in the first log, stream id is bd, c0 and e0, in the second log only c0 and e0, bd does not exist. This is the only one difference between real LiveTV and recorded LiveTV I can see....
and the huge gap between pts of stream c0,e0 to bd. Seems that stream id confuses your bit 31,32 logic.

I made a quick hack, bypassing the bit 31,32 logic if stream bd is there (just doing a return hdr_len), and voila, stuttering has gone Blush

RE: VNSI4 alpha - FernetMenta - 2013-03-28

Bravo Nachteule, you definitely deserve a reputation point for those findings Smile
VDR does not record teletext. I will only incluse video and audio for the wrap logic.

RE: VNSI4 alpha - FernetMenta - 2013-03-28

I have push a fix for this. Now pts wraps are only set by audio and video. There was an additional problem with this: the variables were static and multiple clients could have influenced each other.

RE: VNSI4 alpha - FernetMenta - 2013-03-28

(2013-03-28, 00:04)Krautmaster Wrote: Nvidia Problem?

http://pastebin.com/n2zT4AeQ

would be nice if you included some problem description. Maybe you hit the same problem I just fixed but hard to tell because you did not paste a DEBUG log Sad

RE: VNSI4 alpha - Nachteule - 2013-03-28

(2013-03-28, 11:00)FernetMenta Wrote: I have push a fix for this. Now pts wraps are only set by audio and video. There was an additional problem with this: the variables were static and multiple clients could have influenced each other.
Great job, issue is definitely gone here

RE: VNSI4 alpha - ezechiel1917 - 2013-03-28

(2013-03-28, 11:00)FernetMenta Wrote: I have push a fix for this. Now pts wraps are only set by audio and video. There was an additional problem with this: the variables were static and multiple clients could have influenced each other.

Hi, unfortunately my channel is still stuttering Sad

RE: VNSI4 alpha - ursli - 2013-03-29

I'm not sure if its the same problem but I had stuttering on channels with mono audio and optical output. Discussion there, turned out to be xbmc trying to transcode mono mp2 audio to ac3 causing the stutter, I had to turn off 5.1 support in the audio options to make it work -> http://forum.xbmc.org/showthread.php?tid=153627&pid=1314266#pid1314266

RE: VNSI4 alpha - ezechiel1917 - 2013-03-29

ursli:
thanks for your input, I've played numerous times with audio setup, but it doesn't seem to be related because all the time I watch recordings from that channel, they play just fine indepent on currently selected audio mode/output. Stuttering occurs only with Live TV, so it has to be vnsi demuxer related I guess.

RE: VNSI4 alpha - hfmls - 2013-03-29

i'm getting a 420kb plugin instead of 1000kb plugin vnsi vdr.
i'm using same process as before, anyone with same problem?

i did:
git clone --branch vnsi4 https://github.com/FernetMenta/xbmc-pvr-addons.git
cd /usr/src/xbmc-pvr-addons/addons/pvr.vdr.vnsi/vdr-plugin-vnsiserver
then
sudo make VDRDIR=/usr/include/vdr LIBDIR=/usr/lib/vdr/plugins

RE: VNSI4 alpha - ezechiel1917 - 2013-03-29

FernetMenta:
also noticed that latest commit brought audio stutter to another channel. reverting the commit fixes stuttering on that channel.
I can record another .ts sample if it helps (with all pids including teletext - that would be on windows if that's ok)

(2013-03-28, 02:32)Nachteule Wrote: I added my old debug message to the code (end of function ParsePESHeader) looking a little bit what's going on: pts and dts values are going crasy if issue occours, here one example

Is it possible you could share the code so I could include this debug in my compile and help FernetMenta identify stuttering of my channels?

hfmls:
I got 400k plugin when I didn't have proper vdr sources.

RE: VNSI4 alpha - hfmls - 2013-03-29

(2013-03-29, 04:06)ezechiel1917 Wrote: FernetMenta:
also noticed that latest commit brought audio stutter to another channel. reverting the commit fixes stuttering on that channel.
I can record another .ts sample if it helps (with all pids including teletext - that would be on windows if that's ok)

(2013-03-28, 02:32)Nachteule Wrote: I added my old debug message to the code (end of function ParsePESHeader) looking a little bit what's going on: pts and dts values are going crasy if issue occours, here one example

Is it possible you could share the code so I could include this debug in my compile and help FernetMenta identify stuttering of my channels?

hfmls:
I got 400k plugin when I didn't have proper vdr sources.

I see, can you tell me how to do it then please?
Thank you.

RE: VNSI4 alpha - ezechiel1917 - 2013-03-29

try Krautmaster's installation

(2013-03-22, 15:00)Krautmaster Wrote: add-apt-repository -y ppa:aap/vdr
apt-get update
sudo apt-get -y install dvb-apps
sudo apt-get -y install vdr vdr-plugin-dvbapi vdr-plugin-wirbelscan oscam fakeroot git-core
mkdir /home/xbmc/build
cd /home/xbmc/build
apt-get build-dep -y vdr
apt-get source vdr
mv vdr-* vdr
git clone --branch vnsi4 https://github.com/FernetMenta/xbmc-pvr-addons.git
rm -R vdr/PLUGINS/src/*
ln -s /home/xbmc/build/xbmc-pvr-addons/addons/pvr.vdr.vnsi/vdr-plugin-vnsiserver /home/xbmc/build/vdr/PLUGINS/src/vnsiserver
cd /home/xbmc/build/vdr/
make -j4 VDRDIR=/usr/include/vdr LIBDIR=/usr/lib/vdr/plugins
make install VDRDIR=/usr/include/vdr LIBDIR=/usr/lib/vdr/plugins


RE: VNSI4 alpha - FernetMenta - 2013-03-29

ezechiel1917, I was able to reproduce the issue with your sample. I will push a fix soon.