[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Allocation of a stream descriptor for MPEG-2

By all means do it the same on all MPEG-2 transports! To do this "right", you need to:
1. pick a stream_type value > 0x80 - look at the ATSC registry and find an unused one you like. Then send email to ATSC telling them what you did and providing a link to the IETF document for reference. 
2. register a format_identifier value with the SMPTE RA, such as "ULE1"
3. require transports carrying these streams put the ISO 13818-1 registration descriptor in the PMT ES_info descriptor loop with the value assigned from #2 4. update the IETF document with the above information so you have interoperability
Steps #2 and #3 may seem a bit redundant if you pick an unused value in #1. However, there is no formal registry and your "assigned" value can in fact be legally used by anyone else. All of the above steps would be required in an ATSC transport. DVB and ARIB may be more flexible - don't know. But in any event, this is how it should be done to ensure unambiguous signaling of the stream containing these packets.
ISO will not assign you a value below 0x80 from 13818-1, Table 2.29 since they did not define the elementary stream.
As an analogy, AC-3 audio (not an ISO codec) is customarily stream_type=0x81 (in ATSC and DVB both), and requires the registration descriptor with the registered value "AC-3" in the PMT ES_info loop.
No coordination is needed if you document this in the IETF document - ATSC and DVB should simply do it as recommended. If you do not document this in the IETF document, then someone will have to pay attention in DVB and ATSC both to ensure it is done the same way and documented somehow. Better that you do it once. 

Regards,

        Mike

At 3/11/2005 09:27 AM, Gorry Fairhurst wrote:


Thanks Mike,

This looks like the correct thing, which registry item should we be asking
to create?

If we have support for the same type in a DVB multiplex, do you know if
there is any merit in co-ordinating this process?

Best wishes,

Gorry

On 9/3/05 9:14 am, "Michael A Dolan" <md.1@tbt.com> wrote:

> Private range MPEG field values do not have an authoritative registry, but
> ATSC has informally provided an industry code point document:
>
> http://www.atsc.org/standards/Code_Point_Registry.pdf
>
> For format_identifier values (used in the registration descriptor), SMPTE
> is the formal ISO registrar:
>
> http://www.smpte-ra.org/mpeg.html
>
> Regards,
>
>          Mike
>
> At 3/9/2005 06:11 AM, Gorry Fairhurst wrote:
>
>> Dear all,
>>
>> The ipdvb WG of the IETF have developed a document that describes an Ultra
>> Lightweight Encapsulation (ULE) mechanism for the transport of IPv4 and IPv6 
>> Datagrams and other network protocol packets directly over the ISO MPEG-2
>> Transport Stream as TS Private Data.
>>
>> This document has now been submitted to the Internet Engineering Steering
>> Group (IESG) with a request for publication as an IETF Standards Track
>> document in the RFC series.
>>
>> https://datatracker.ietf.org/public/pidtracker.cgi?command=view_id&dTag=1163 
>> 4
>>
>> At the ipdvb meeting on 7th March, the WG noted that the MPEG-2
>> specification requires conformant TS Multiplexes to provide Program Specific 
>> Information (PSI) for each stream in the TS Multiplex. Other MPEG-2 based
>> transmission standards may also define Service Information (SI). This
>> information may allow Receivers and Re-multiplexors to locate a specific ULE 
>> Stream (i.e., the PID value that carries a ULE Stream).
>>
>> The WG therefore intends to apply for allocation of an appropriate code
>> point(s) (i.e. Descriptor / Identifier) from the relevant organisations that 
>> co-ordinate the use of MPEG-2.
>>
>> If you have experience of the procedure, or have any useful comments, or
>> would be willing to help develop text for an appropriate request, could you 
>> please let me know.
>>
>> Gorry Fairhurst
>> (ipdvb WG Chair)
>