Markku Leinio writes:
> At 12:44 24.6.2001 +0200, Marcus O.C. Metzler wrote:
> >Due to the limitations of the DVB-s card we cannot get to the original
> >TS coming from the transponder. The TS has to be recreated from the
> >audio and video buffer of the card. That leads to problems with muxing
> >and leaves us without a usable PCR. This makes it also difficult to
> >transcode the "TS" into a PS.
> >The TS stream from the Nova cards on the other hand is just as we get
> >it from the transponder with PCR and everything. It should be much
> >easier to make a usable PS from that.
>
> How about setting up a filter for PCR? I mean, in test_dvr.c there are set
> up two filters, one for audio (DMX_PES_AUDIO) and one for video
> (DMX_PES_VIDEO). There is, however, a constant named DMX_PES_PCR. Can that
> be used to get the PCR with the stream?
>
> Apparently n-tv sends PCR as a separate PID and not inside the video PID?
> Do you know the PID to be used? Video PID is 162 and audio is 96, at least
> with them I have got some kind of TS out of the /dev/ost/dvr.
>
> Strange thing is that every 12th (or so) TS packet is coming with the
> adaptation field set, but there is no information in the adaptation field,
> no flag is set. Why is that?
Well, how many times do we have to say it?
You cannot get the original TS packets from the AV7110.
Everything you get from the driver is a reconstruction made from the
available data which, in case of the currently watched program, is the
video ES and PTS and the audio PES. Anything else like the adaptation
field data is lost. Of course we still need (empty) adaptation fields
to adapt (hence the name) to PES packets lengths not divisible by 184.
Ralph
--
Info:
To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe linux-dvb" as
subject.
