[linux-dvb] Re: Which CRC-code is used in TS of DVB?

[Gregor Lawatscheck <gpel@mpex.net>]

At 03:11 27/05/2003, you wrote:
> >  > AFAIK there is no CRC in the TS but in the PES header? I haven't 
> checked
> >  > this, so I could be wrong.
> >
> > No CRC in either.
> >
>
> What is that? Quote from Holger Waechtler:
> [...]
> I think I've read CRC32 somewhere - is that correct?
> it's CRC32, the decoder model is described in ITU-T H.222.0, Annex A.
>[...]

I've now had a look at: http://www.mpucoder.com/DVD/pes-hdr.html - there 
can be some kind of CRC of the last PES, but not sure if that's CRC32 and 
whether any broadcaster bothers sending it along.

> >  > Now that sounds like it'll drop sections that didn't pass CRC32, 
> doesn't
> >  > it? I'm not aware CRC32 corrects anything. I personally just fear CRC32
> >
> > Correct.
> >
> OK, what passes the CRC-check, then?
IMHO a section (PES packet?)  where the checksum calculated from the packet 
contet itself matches the checksum in the next packet. There have got to be 
CRC checksums somewhere in the packets otherwise no point to calculate them?

> >  > errors, for example in archives etc. To me it's a checksum: 
> According to
> >  > http://www.cyberdyne-software.com/crc32.html CRC = "cyclic redundancy
> >  > check(sum)". What do you do when the checksum fails? "Invert[ing] 
> the wrong
> >  > bits" is certainly beyond me.
>
> The application's name is "CRC 32"!
I see no difference between CRC32 and CRC 32 other than cosmetic. 32 is the 
number of bits used for the checksum.

> I don't know the CRC32, and I'm currently looking for specs.
>
> See
>
> http://www.informatik.uni-leipzig.de/rnvs/lehre/rn1/DLL_AR02.pdf
>
> for description of cyclic redundancy CODE.
> > Confused with FECs.
>
> So the question: Is CRC32 in DVB-TS a cyclic redundancy checksum or
> code?


Throughout, this document speaks of "Fehlererkennungscode" - error 
detection rather than error correction. CRC32 also speaks of polynomially 
calculated checksums. I believe they are the same thing just the acronym 
was translated differently.

> If checksum the DVB-developers would be idiots - especially as
> data-packets are broadcasted by carussels where a client can't demand to
> resend the packet - so it has to be valid.
The way I understand it you calculate the checksum of the actual packet and 
then look at the CRC for it to see whether the packet is broken or not. 
That's all there is to it imho.

> > Yup, it is part of the demodulator hardware.
>
> With the full-featured cards only or also with low-budgets?
Both have FEC. I'm pretty sure because the code rate / FEC value sits in 
the channels.conf or is used with dvbstream etc.



--
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe linux-dvb" as subject.