Avalpa Community

cut tables on casting
Page 1 of 1

Author:  rubengc [ Tue Nov 19, 2013 2:02 pm ]
Post subject:  cut tables on casting

Hi everyone!

I got some problems to cast DVB-C correctly, for example:

I have a NIT with some descriptors, but when i cast them i can see (with whireshark) only a malformed NITs (the information is fragmented), but works fine in my receptor tdc

This happens with the other tables too (SDT, EIT)

For cast, i use this commands:
tscbrmuxer (wihtout null.ts)
tstdt (only in the first transport stream)

Why i cast malformed or uncomplete packets?

Author:  aventuri [ Wed Nov 20, 2013 3:55 pm ]
Post subject:  Re: cut tables on casting

sorry but we can't help you if you don't copy here the complete script you are using to create and transmit the TS stream.

the tools are those, but who knows how they are tied together. then are you using fifos or files between them?

what finally puzzles me is they you say the decoder at the end of the chain are doing fine.. i'm wondering if you are using the right tool also for your diagnosis.

wireshark is indeed a powerful tool, and you should have 7 TS packets into an ethernet frame but that doesn't mean you could have here a section spanning more then one TS packet.

how long is the NIT section, for example?



Author:  rubengc [ Fri Nov 22, 2013 12:59 pm ]
Post subject:  Re: cut tables on casting

Thanks andrea for reply

I attach my py files

For sending i use a php script that genrate this commands:
tscbrmuxer b:1400 nit1.ts b:1400 nit2.ts b:1500 sdt1.ts b:1500 sdt2.ts b:1500 sdt3.ts b:1500 sdt4.ts b:1500 sdt5.ts b:1500 sdt6.ts b:1500 sdt7.ts b:1500 sdt8.ts b:1500 sdt9.ts b:1500 sdt10.ts b:1500 sdt11.ts b:1500 sdt12.ts b:1500 sdt13.ts b:1500 sdt14.ts b:1500 sdt15.ts b:1500 sdt16.ts b:2000 tdt.ts b:2000 tot.ts b:2000 antena3eit1.ts b:2000 antena3eit2.ts b:2000 antena3eit3.ts b:2000 antena3eit4.ts b:2000 antena3eit5.ts ...(more eits) > fifo.ts &
sleep 2
tstdt fifo.ts > fifotdt.ts &
sleep 2
tsstamp fifotdt.ts 156800 > fifostamp.ts & (the bitrate is from a tscbtmuxer with all eits)
sleep 2
tsudpsend fifostamp.ts 45500 156800

I try to use a null.ts with 4kk of bitrate, but this dont improve

edit: ah, this for DVB-C

Author:  aventuri [ Sat Nov 23, 2013 6:42 pm ]
Post subject:  Re: cut tables on casting

i see the problem.

you SHOULD have all the TS inputs into the tscbrmuxer with DIFFERENT PIDs.

so if you have sdt1.ts sdt2.ts ... sdt15.ts ALL with PID 17, they are going to CLASH each other at lest for the continuity counter and eventually with the section too, if the sections span more than a TS packet (188).

the same is true for those EITs (PID18) and NIT (PID 16) and all the other "well known" PIDs..

the solution is to append all the EITs into a single long file (for example allEits.ts) and THEN use that allEit.ts as parameter for tscbmuxer.

do the same for NITs (cat nit*.ts > allNit.ts) and all SDTs ("cat sdt*.ts > allSdt.ts) and then feed them once into tscbrmuxer, something like:

tscbrmuxer b:2800 allNit.ts b:10000 allSdt.ts b:1500 allEit.ts ... > fifo.ts &

Author:  rubengc [ Tue Nov 26, 2013 12:05 pm ]
Post subject:  Re: cut tables on casting

Thanks you very much!!!

I don't know anything about this

I tried with nit and sdt, now i will rewrite my code for generate a new name structure of eits for make the same (thats probably can solve my problem with present/following events)

Thanks thanks thanks!!!!

Author:  aventuri [ Thu Nov 28, 2013 6:48 pm ]
Post subject:  Re: cut tables on casting


we really can't help you more, as it seems you are not trying to understand how things are supposed to work..

in my point of view, you need a proper consultancy service. on the other hand, if you are just playing with it for fun, that's a great way to spend your time! :-)



Page 1 of 1 All times are UTC + 1 hour [ DST ]
Powered by phpBB® Forum Software © phpBB Group