Jump to content


Photo

Openpli-5 (still next master)


  • Please log in to reply
1177 replies to this topic

Re: Openpli-5 (still next master) #221 Taapat

  • PLi® Core member
  • 2,345 posts

+121
Excellent

Posted 4 April 2016 - 21:01

As I wrote, this is not my file.

This file as an example for me offered one user.
He reported that he and his friends are several mkv file with such problems, he just tried to find little as possible.
Unfortunately, he can not say on which gstreamer version starts problems with the following files.
He is only checked that changing library to version 1.4.5 file is played without problems.



Re: Openpli-5 (still next master) #222 Taapat

  • PLi® Core member
  • 2,345 posts

+121
Excellent

Posted 4 April 2016 - 22:54

I fix the problems in this mkv file by creating patch which revert changes from: https://bugzilla.gno...g.cgi?id=745192

 

https://github.com/T...24309f3a306c41b

 

I understand that this may cause problems with other files, but so far I have not seen such.



Re: Openpli-5 (still next master) #223 littlesat

  • PLi® Core member
  • 57,469 posts

+708
Excellent

Posted 5 April 2016 - 07:59

I suggest this should be forwarded to the gstreamer team....


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: Openpli-5 (still next master) #224 malakudi

  • Senior Member
  • 1,449 posts

+69
Good

Posted 5 April 2016 - 08:09

@Taapat: as I said, this is a problematic file and the producing application should be fixed, not gstreamer. ffmpeg also complains about this file.



Re: Openpli-5 (still next master) #225 peti

  • Senior Member
  • 115 posts

+1
Neutral

Posted 5 April 2016 - 10:03

I still do not have luck on my Vu+ Uno. Started from scratch, but the image is not booting. It's stuck at the bootscreen, or in some cases there's even no bootscreen. Any idea? If I try to upgrade the last working image (from about 2 weeks ago) with opkg, the box freezes while configuring the packages and doesn't boot anymore.


Edited by peti, 5 April 2016 - 10:04.

T90 | 28.2E-23.5E-19.2E-13E-9E-4.8E-0.8W-4W

Zgemma H7S | OpenPLI develop branch


Re: Openpli-5 (still next master) #226 blzr

  • PLi® Core member
  • 2,270 posts

+118
Excellent

Posted 5 April 2016 - 11:00

yup, I have the same issue with my et9500, on the first boot after flash it stuck on the bootlogo,
and on subsequent boots even the bootlogo does not appear, it's either a black screen, or a frozen cursor in the upper left corner

on vuzero everything works fine, so seems that only  'ancient' boxes/kernels are affected...

I'm eventually going to try to see in the hyperterminal what's going on underneath, need to prepare a cable to connect (no DB9 connector in et9500)


True sarcasm doesn't need green font...

Re: Openpli-5 (still next master) #227 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 5 April 2016 - 11:05

@chris
 
If you can instal sometimes on your Duo2 the Plugin TV feature and then the Youtube function in the Plugin test?
 
 
I have tried everything, but these simply become no Videos played.
 
Greeting Akki


Re: Openpli-5 (still next master) #228 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 11:41

@Taapat

 

I tested the file self and it does play on all players of my pc

ubuntu 14.04 (x86-64)

-standard video player (which is using gstreamer)

-vlc player

 

with firefox browser streaming there it's also using totem player (which is using gstreamer)

 

file plays 100 % nice.

 

on a stb we well always use internal codecs for video (no raw video)  internal or gstreamer codecs for audio.

 

On a pc all is always converted to raw/video raw audio by gstreamer. or  in vlc with there codecs.

 

@malakudi ,

 

ok the media is made in a certain way , but the question is : is the media wrong or is the interpretation off gstreamer wrong ? with other words Is there a standard iso norm for creation off a mkv container ?

 

Looks that gstreamer inserted a patch for certain mkv container media so that it plays but at the same time brakes other mkv containers who always worked very well.

 

What is right or what is wrong ? that's rather the question here.


Edited by christophecvr, 5 April 2016 - 11:43.


Re: Openpli-5 (still next master) #229 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 11:59

 

@chris
 
If you can instal sometimes on your Duo2 the Plugin TV feature and then the Youtube function in the Plugin test?
 
 
I have tried everything, but these simply become no Videos played.
 
Greeting Akki

 

I do not have that plugin ,

 

But If it's third party I gues it's made for older images , and will not work.

 

The youtube plugin self does work perfect on duo2



Re: Openpli-5 (still next master) #230 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 12:09

I fix the problems in this mkv file by creating patch which revert changes from: https://bugzilla.gno...g.cgi?id=745192

 

https://github.com/T...24309f3a306c41b

 

I understand that this may cause problems with other files, but so far I have not seen such.

Perhaps You could reopen the bug , and post a link to this type off media.

Tell what You did to let it work

 

The media does not have a pts but well a dts time stamp , .. normally it should work.



Re: Openpli-5 (still next master) #231 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 12:21

As extra about that media : the audio is ac3 not dts but it's using a dts time stamp that looks not to be right.

 

I gues they converted media with dts audio with a wrong set and or configured audio converter.



Re: Openpli-5 (still next master) #232 Akki

  • Senior Member
  • 204 posts

+3
Neutral

Posted 5 April 2016 - 12:42

If no problem to themselves is to be installed sometimes fast the Plugin,
only wanted to know whether there are also the problems with you.
 
I suspend sometimes the Plugin as a download, would be nice if you, nevertheless, sometimes could test.

 

Attached Files



Re: Openpli-5 (still next master) #233 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 14:36

@Taapat

 

I just did I try out since I just found out that the audio track in the container is ac3 . (ac3 is suposed to use pts timestamps not dts and mkvdemux does require sync between video/audio)

 

Kowever the file posted above does work perfect with a (tempo brutal not correct work-aroud I did)

 

see :

http://forums.openpl...ndpost&p=540955



Re: Openpli-5 (still next master) #234 malakudi

  • Senior Member
  • 1,449 posts

+69
Good

Posted 5 April 2016 - 15:50

 

@malakudi ,

 

ok the media is made in a certain way , but the question is : is the media wrong or is the interpretation off gstreamer wrong ? with other words Is there a standard iso norm for creation off a mkv container ?

 

Looks that gstreamer inserted a patch for certain mkv container media so that it plays but at the same time brakes other mkv containers who always worked very well.

 

What is right or what is wrong ? that's rather the question here.

 

Right is that the file should have correct timestamps, and it doesn't. ffmpeg complains about it as well, although it plays it. Original file if remuxed with ffmpeg matroska muxer, plays back fine. So the problem is on the application generating such file, not in gstreamer. If I were the gstreamer developer and you opened a ticket about this, that's what I would reply. I might be wrong though, this is just my interpretation of the situation. I think the problem is in the "WonderShare Matroska Muxer" and not in gstreamer.


Edited by malakudi, 5 April 2016 - 15:51.


Re: Openpli-5 (still next master) #235 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 16:05

 

 

@malakudi ,

 

ok the media is made in a certain way , but the question is : is the media wrong or is the interpretation off gstreamer wrong ? with other words Is there a standard iso norm for creation off a mkv container ?

 

Looks that gstreamer inserted a patch for certain mkv container media so that it plays but at the same time brakes other mkv containers who always worked very well.

 

What is right or what is wrong ? that's rather the question here.

 

Right is that the file should have correct timestamps, and it doesn't. ffmpeg complains about it as well, although it plays it. Original file if remuxed with ffmpeg matroska muxer, plays back fine. So the problem is on the application generating such file, not in gstreamer. If I were the gstreamer developer and you opened a ticket about this, that's what I would reply. I might be wrong though, this is just my interpretation of the situation. I think the problem is in the "WonderShare Matroska Muxer" and not in gstreamer.

 

I just found out. gstreamer handle it wel off in a correct way. But our videoink just blocked the video playing due to a missing timestamp into media. Well iff audio is avbl and it is , it plays which also was the case. Use whatever timestamp avbl priority is given to timestamp off pts however else that from dts is used.

 

By me master-next, gst-1.9.0.1 on dm8000 and duo2.

 

It all works nice for all mkv containers nowhere any regression by just using last (experimental mediasink)

 

https://github.com/c...8bf7b498f43c774



Re: Openpli-5 (still next master) #236 kalkan99

  • Member
  • 6 posts

0
Neutral

Posted 5 April 2016 - 16:08

dear friends,
I scanned 12 pages relevant to the subject.
I possible device duo 2
Where can I find a link to the image in its final form?
Thank you in advance for your friends to share.



Re: Openpli-5 (still next master) #237 christophecvr

  • Senior Member
  • 3,131 posts

+140
Excellent

Posted 5 April 2016 - 16:18

dear friends,
I scanned 12 pages relevant to the subject.
I possible device duo 2
Where can I find a link to the image in its final form?
Thank you in advance for your friends to share.

Currently all about gst-1.0 is in development phase .

 

New try-out experimental images will come made by MastaG they are non official but good , and persons who like to test do test them .

 

But to have a stable images You will have to be more patient.



Re: Openpli-5 (still next master) #238 kalkan99

  • Member
  • 6 posts

0
Neutral

Posted 5 April 2016 - 16:23

Thank you for your answer.



Re: Openpli-5 (still next master) #239 peti

  • Senior Member
  • 115 posts

+1
Neutral

Posted 5 April 2016 - 18:52

I still do not have luck on my Vu+ Uno. Started from scratch, but the image is not booting. It's stuck at the bootscreen, or in some cases there's even no bootscreen. Any idea? If I try to upgrade the last working image (from about 2 weeks ago) with opkg, the box freezes while configuring the packages and doesn't boot anymore.

 

I collected some logs, can you please check? Attached also a working one (about a month old image) for reference

Attached Files


T90 | 28.2E-23.5E-19.2E-13E-9E-4.8E-0.8W-4W

Zgemma H7S | OpenPLI develop branch


Re: Openpli-5 (still next master) #240 blzr

  • PLi® Core member
  • 2,270 posts

+118
Excellent

Posted 5 April 2016 - 23:36

@peti

I found this: https://github.com/o...b6503d0e3f72331 -

reportedly (upgraded) glibc 2.23 doesn't like bcm7405/7413 boxes for whatever reason (brodcom drivers fault?)

 

to check I built master-next image with only glibc rolled back to 2.22 and it boots and works fine...

 

not sure if we can expect any glibc fix/backport upstream, maybe older boxes simply will no longer be able to keep up with OE
 

and btw I made a serial cable in the meantime, but it turned out that hyperterminal is pretty useless for investigating this issue, logs for broken and working image were the same in my case

(just ended with 'Starting program at 0x805fdb90' and that's all)


True sarcasm doesn't need green font...


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users