Re: Why is everything on E4 always out if lipsync ?


"Paul Schofield" <paul_AT_sonifex_DOT_co_DOT_uk> wrote in message
news:h_udnc7U1K4d2bXeRVnyjQ@xxxxxxxxxxxx
>
> "Chas Gill" <Chas.Gill@xxxxxxxxxxxxxxxxxxxxx> wrote in message
> news:dg8s2h$2eu$1@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>>
>> "Paul Schofield" <paul_AT_sonifex_DOT_co_DOT_uk> wrote in message
>> news:8u-dnepGrMJGcrreRVnysw@xxxxxxxxxxxx
>> >> >>> Why is everything on E4 always out if lipsync ?
>> >> >>>
>> >> >>> And when are they going to fix it. C4 isn't much better.
>> >> >>>
>> >> >>
>> >
>> >> I have noticed that when I watch satellite tv, sometimes when the lips
>> >> are
>> >> out of sync I just switch the box off, then back on and it fixes it.
>> >
>> > This means it's likely to be an STB buffering problem and is not the
>> > responsibility (and in fact beyond control) of the broadcaster. It may
>> > well
>> > be that the stat muxing and variable compression systems are not
>> > handled
>> > very well in some STB's so some channels will exhibit the problem more
>> > than
>> > others.
>> >
>> > FWIW we used to have a netgem iplayer with a VCR routed through the
> netgem
>> > scart sockets. After playing a video there was always a lip sync
>> > problem
>> > of
>> > about 2-3 seconds! Entering standby mode solved this.
>> >
>> > --
>> > Paul Schofield
>>
>> Before I go off on another rant (I may be demonstrating my ignorance
>> here)
>> could somebody point me to a good internet resource that explains exactly
>> how DTV works, i.e. why is it necessary to code audio & video separately.
>> (I presume that they are separate bitstreams otherwise how could they get
>> out of sync. - and if this is the case, what are the problems that
> conspire
>> to cause unsync.) I have a £200 Canopus ADVC110 which does an excellent
> job
>> of digitising analogue video for later editing on my PC - it can be left
> to
>> run for hours at a time and the resulting digitised file has perfect
>> sync.
>> As an amateur with an engineering background, but not in TV systems, I am
>> eager to learn why this can't be achieved with broadcast systems, with
> their
>> huge resources and budgets and, if it truly can't be achieved
> consistently,
>> why the hell have "they" bothered to introduce DTV before somebody has
>> figured out how to make it work properly?
>>
>> Chas
>>
>>
>
> Can't help you with the technical sites, however...
>
> I'm pretty sure that it's that audio and video are DECODED separately that
> causes the problems. The audio is coded as distinct frames so that the
> system needs to decode a frame (1152 samples of audio?) and then queue
> this
> up to be sent to the audio DAC, the time to decode a frame is more or less
> the same and relatively small. The video is made of all sorts of info that
> depends on the complexity of the signal and the resultant bitrate that is
> required. Some of this stuff is whole frame, some is predictive, some is
> just changed data, and then add error correction data to the mix. The time
> to decode a specific video frame is extremely variable and can be
> relatively
> large. This means the both video and audio data has to be buffered and the
> playout to the various DACs has to be synchronised to avoid lip-sync
> faults.
> It's clear that this doesn't always work, and it's difficult to pin down
> the
> problem. I'd guess it could be a number of things - that the lock to the
> stream may be faulty so that losing a couple of frames of video due to
> errors in the stream causes the video to step and so losing sync to the
> audio (switching from & to the channel should solve this); it could be
> that
> some channels have very long decode times so causing one of the buffers to
> run out of data; it could be that the STB processor cannot keep up so we
> lose sync that way; or even, as in the VCR case above, there is a bug in
> the
> calculation to determine the sync or buffer pointers are set or cleared
> erroneously
> to again lose lip-sync.
>
> On the closed system you describe everything starts at the correct, known
> point; there should be no imperfections in the data stream; lack of MIPS
> in
> the processor; or shortage of memory for buffering to cause loss of sync.
>
>
> HTH
>
> --
> Paul Schofield
>
>
>

Thanks, Paul, for an excellent description of what's going on. Makes me
feel a little better ;-)

Chas


.

Relevant Pages

  • Re: DTV audio / vidio sync
    ... stay in sync. ... make delay adjustments differentially, any delay of some part of the A/V ...
    (sci.engr.television.broadcast)
  • Re: Unsychronous recording from DV cameras
    ... What you need is correct timestamps, something DirectShow does not provide in the design and always make audio/video sync a gamble. ... The only time audio/video is accurately in sync is when the source send packet with audio video muxed. ...
    (microsoft.public.win32.programmer.directx.video)
  • Re: Audio/video is not in sync
    ... There are a number of shows on my cable system that the audio ... MCE viewing, I have noticed that if the channel doesn't tune in immediately, ... there may be some audio and video skew. ... If the CPU is more than 50% busy you may have sync issues. ...
    (microsoft.public.windows.mediacenter)
  • Re: Why is everything on E4 always out if lipsync ?
    ... >>> out of sync I just switch the box off, then back on and it fixes it. ... i.e. why is it necessary to code audio & video separately. ... system needs to decode a frame and then queue this ...
    (uk.tech.digital-tv)
  • Re: Audio sync issues
    ... According to this, all DVD-V audio is 48khz, or higher, 44.1 khz is not ... Sampling rate shouldnt affect sync. ... Seperate the audio streamfrom the video stream-called demuxing. ...
    (rec.video.dvd.tech)