[Finished] Improvements to Live TV playback rate matching - for testing (1 Viewer)

Status
Not open for further replies.

azzuro

Test Group
  • Team MediaPortal
  • May 10, 2007
    9,984
    5,663
    France - IDF
    Home Country
    France France
    @Owlsroost
    i can't build you last commit :
    Error 222 error C2039: 'OnBitRateChanged' : is not a member of 'ITSReaderCallback' D:\svnroot\1.Mediaportal\MediaPortal-1\DirectShowFilters\TsReader\source\TsReader.cpp 776 1 TsReader
     

    azzuro

    Test Group
  • Team MediaPortal
  • May 10, 2007
    9,984
    5,663
    France - IDF
    Home Country
    France France
    //Make compatible with MP1.11 and later versions if defined
    #define BITRATE_REPORT

    on tsreader.h must be reverted. :giggle:
     

    HTPCSourcer

    Retired Team Member
  • Premium Supporter
  • May 16, 2008
    11,418
    2,336
    Home Country
    Germany Germany
    Hi,

    Switching from UNC to RTSP (and using your files from posting #1) has cured my issue of short pauses at Live TV under Windows 10 - at the price of annoyingly long channel switch times :(. Except for this I am very happy with Live TV experience, which is otherwise running absolutely stable.

    Notwithstanding, a channel change time of 7-9 seconds, compared to some 3 seconds under UNC (when switching from encrypted channel A to encrypted channel B), is definitely a long time. Is there a chance to tune settings, e.g. reduce buffer times etc., to reduce the channel switching time?
     

    Owlsroost

    Retired Team Member
  • Premium Supporter
  • October 28, 2008
    5,540
    5,038
    Cambridge
    Home Country
    United Kingdom United Kingdom
    Is there a chance to tune settings, e.g. reduce buffer times etc., to reduce the channel switching time?

    Please try the attached files (TsReader.ax for the clients, all the others for the server) - installation instructions in the first post in the thread.

    You should get shorter channel change (zapping) times when using RTSP with these versions - please test and report back :)

    If you have problems with them, I need both the client and server logs.
     

    Attachments

    • EXP-Upgrade_555_MM_Owlsroost_16-05-2016_4.zip
      532.2 KB

    HTPCSourcer

    Retired Team Member
  • Premium Supporter
  • May 16, 2008
    11,418
    2,336
    Home Country
    Germany Germany
    Hi @Owlsroost ,

    Does your build in posting #1 include the recent tsreader logging bug fix ([MP1-4788] - TsReader logging thread is not closed on filter destruction) ?

    Regarding the zapping time, can you give some insight into what elements in RTSP use govern the channel switch time? Is there any furthr optimization potential or are we talking minor opportunities only?

    What I found interesting is that sometimes the channel change time is really fast. I have not systematically analyzed the situation and it might simply be due to switching over from one channel to the another on the same transponder, i.e. without the need to load another card. Anybody else tested the latest versions?
     

    Owlsroost

    Retired Team Member
  • Premium Supporter
  • October 28, 2008
    5,540
    5,038
    Cambridge
    Home Country
    United Kingdom United Kingdom
    Does your build in posting #1 include the recent tsreader logging bug fix ([MP1-4788] - TsReader logging thread is not closed on filter destruction) ?

    Yes, it does.

    Regarding the zapping time, can you give some insight into what elements in RTSP use govern the channel switch time? Is there any furthr optimization potential or are we talking minor opportunities only?

    Mostly it's mixture of less accurate 'seeking' in RTSP mode and TsReader having to wait for data from the new channel to arrive in real time (instead of being able to read through the timeshift buffer data faster than real time when it's searching for the new MPEG header information, as it does when directly reading the timeshift file).

    Yes, I think there is potential for further optimization, but I don't think it's ever going to be as fast as with UNC paths.

    What I found interesting is that sometimes the channel change time is really fast. I have not systematically analyzed the situation and it might simply be due to switching over from one channel to the another on the same transponder, i.e. without the need to load another card.

    I'm interested in log files for the fast and slow changes (I need both the client and sever logs, so I can relate what's happening at both ends of the system).
     
    Status
    Not open for further replies.

    Users who are viewing this thread

    Top Bottom