Page 1 of 3 123 LastLast
Results 1 to 10 of 28

Thread: Any way to improve skipping performance with .ts files?

  1. #1
    Join Date
    Aug 2011
    Location
    Virginia, USA
    Posts
    457

    Any way to improve skipping performance with .ts files?

    The A-110 is a little slow to react when trying to skip ahead/back and sometimes only seems to skip a few seconds instead of the 28 seconds it is supposed to. This is with .ts (ATSC) files. Is there any way to improve this? Will the new skipping patch for NPVR have any affect on this? Thanks.

  2. #2
    Join Date
    May 2006
    Location
    Canada
    Posts
    20,772
    My experience with ATSC is that it is as fast and accurate as the PC . Can you compress and send your logs?

    Here is a snippet from mine on a 1080i QAM file I pressed the 5 then Next to skip 300 seconds and was playing in under .7 seconds

    Code:
    2012-03-02 17:24:56.309	[DEBUG][3]	User pressed: D5
    2012-03-02 17:24:57.103	[DEBUG][3]	Raw PTS received:3451949
    2012-03-02 17:24:57.103	[DEBUG][3]	Reported PTS now:3430928  time: 76.2428444444444
    2012-03-02 17:24:57.412	[DEBUG][3]	User pressed: Right
    2012-03-02 17:24:57.506	[DEBUG][7]	requestedPositionInSeconds: 376
    2012-03-02 17:24:57.689	[DEBUG][3]	Raw PTS received:3471468
    2012-03-02 17:24:57.689	[DEBUG][3]	Reported PTS now:3450447  time: 76.6766
    2012-03-02 17:24:58.171	[DEBUG][3]	Raw PTS received:16932416
    2012-03-02 17:24:58.171	[DEBUG][3]	Reported PTS now:16911395  time: 375.808777777778


    Martin

  3. #3
    Join Date
    Jun 2007
    Location
    St. Paul, MN, USA
    Posts
    527
    Quote Originally Posted by cncb View Post
    The A-110 is a little slow to react when trying to skip ahead/back and sometimes only seems to skip a few seconds instead of the 28 seconds it is supposed to. This is with .ts (ATSC) files. Is there any way to improve this? Will the new skipping patch for NPVR have any affect on this? Thanks.
    Skipping and pausing with my A-110 were also sluggish. I put in the new skipping patches, and they do seem to have improved the situation significantly.

  4. #4
    Join Date
    Aug 2011
    Location
    Virginia, USA
    Posts
    457
    Here is the log from just a few skips forward followed by a few backward. The first couple were pretty quick (although it took a while for the OSD to pop up). The second or third skip backwards was a little slow.

    I will try the NPVR patch to see if it makes a difference.

  5. #5
    Join Date
    Nov 2003
    Location
    NextPVR HQ, Wellington, New Zealand
    Posts
    75,898
    The skipping patches shouldn't make any difference on an NMT. They're only about the PC.

  6. #6
    Join Date
    May 2006
    Location
    Canada
    Posts
    20,772
    I wanted the web.log too.

    Martin

  7. #7
    Join Date
    Aug 2011
    Location
    Virginia, USA
    Posts
    457

  8. #8
    Join Date
    May 2006
    Location
    Canada
    Posts
    20,772
    I don't see any major problem with accuracy at all in those requests all within a second of where NextPVR requested. The second request took more than two seconds and it does look like like NextPVR took some extra time calculating the position.

    In the summary below what you see is the the time between the keystroke being process, the streaming request from web.log, and then finally the response when the stream is playing in the correct locations.

    Code:
    2012-03-02 18:16:25.004	[DEBUG][3]	User pressed: Right
    2012-03-02 18:16:25.082	[DEBUG][7]	requestedPositionInSeconds: 32
    2012-03-02 18:16:25.097	[DEBUG][6]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:26.595	[DEBUG][3]	Reported PTS now:1449698  time: 32.2155111111111
    
    2012-03-02 18:16:27.593	[DEBUG][3]	User pressed: Right
    2012-03-02 18:16:27.609	[DEBUG][7]	requestedPositionInSeconds: 61
    2012-03-02 18:16:27.687	[DEBUG][7]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:28.607	[DEBUG][3]	Reported PTS now:2728976  time: 60.6439111111111
    
    2012-03-02 18:16:29.075	[DEBUG][3]	User pressed: Right
    2012-03-02 18:16:29.668	[DEBUG][7]	requestedPositionInSeconds: 89
    2012-03-02 18:16:29.715	[DEBUG][6]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:30.620	[DEBUG][3]	Reported PTS now:3978975  time: 88.4216666666667
    
    2012-03-02 18:16:31.072	[DEBUG][3]	User pressed: Right
    2012-03-02 18:16:32.148	[DEBUG][7]	requestedPositionInSeconds: 117
    2012-03-02 18:16:32.226	[DEBUG][7]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:33.662	[DEBUG][3]	Reported PTS now:5265760  time: 117.016888888889
    
    2012-03-02 18:16:35.908	[DEBUG][3]	User pressed: Left
    2012-03-02 18:16:36.922	[DEBUG][7]	requestedPositionInSeconds: 104
    2012-03-02 18:16:37.000	[DEBUG][4]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:37.718	[DEBUG][3]	Reported PTS now:4652397  time: 103.3866
    
    2012-03-02 18:16:38.966	[DEBUG][3]	User pressed: Left
    2012-03-02 18:16:40.713	[DEBUG][7]	requestedPositionInSeconds: 89
    2012-03-02 18:16:40.760	[DEBUG][7]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:41.805	[DEBUG][3]	Reported PTS now:3984230  time: 88.5384444444445
    
    2012-03-02 18:16:44.379	[DEBUG][3]	User pressed: Left
    2012-03-02 18:16:45.112	[DEBUG][7]	requestedPositionInSeconds: 75
    2012-03-02 18:16:45.190	[DEBUG][4]	Got Web Request (192.168.1.11): /stream
    2012-03-02 18:16:46.922	[DEBUG][3]	Reported PTS now:3368615  time: 74.8581111111111
    Martin

  9. #9
    Join Date
    Aug 2011
    Location
    Virginia, USA
    Posts
    457
    Quote Originally Posted by mvallevand View Post
    The second request took more than two seconds and it does look like like NextPVR took some extra time calculating the position.
    Does this mean that this is something that could possibly be fixed in the NPVR code?

    Also, what I like to do on the PC is pause before I skip back and forward keeping it paused until I find the desired position. But on the NMT it seems to start playback as soon as I skip while paused. Is this behavior something that could be changed?

  10. #10
    Join Date
    May 2006
    Location
    Canada
    Posts
    20,772
    Quote Originally Posted by cncb View Post
    Does this mean that this is something that could possibly be fixed in the NPVR code?
    Possible I guess but with all the variables it is hard to say if there is enough information for sub to track down why it takes so much longer to process your requests than it does mine. It might even be the file itself, as pointed out in the

    Also, what I like to do on the PC is pause before I skip back and forward keeping it paused until I find the desired position. But on the NMT it seems to start playback as soon as I skip while paused. Is this behavior something that could be changed?
    I read about it and this was a fairly recent change and I never implemented it because it's not something I would use. I just tried on the PC though and it seems that when I go Ctrl-Q followed by Ctrl-Right or FF/RW it starts playing just like on the NMT, and it is only with the mouse movement that stays paused. It might be a TSReader2 thing on the PC.

    Still, I will have a look to see if it is something I can even do on the client.

    Martin

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •