[vorbis-dev] Re: Vorbis RTP Internet Draft

Phil Kerr philkerr at elec.gla.ac.uk
Wed Jan 8 04:27:53 PST 2003



Hi Linus, 

This matter came up when I was discussing the submission process with
Colin Perkins and Steve Casner. 

Steve suggested having a unified MIME type and Colin suggested having
them separate, email snippet below: 

-------------------------------------

--> Stephen Casner writes:
>On Thu, 12 Dec 2002, Colin Perkins wrote:
>> --> Phil Kerr writes:
>> >WRT the MIME type registration there is a seperate I-D already in
the
>> >system, draft-walleij-ogg-mediatype-07.txt, which I can reference. 
Also
>> >there is another I-D covering the Ogg file format,
>> >draft-pfeiffer-ogg-fileformat-00.txt.
>>
>> My understanding is that those drafts relate to the ogg file format,
rather
>> than the vorbis codec when packetized in RTP. If this is the case,
you will
>> also need a MIME registration for audio/vorbis in the RTP payload
format.
>
>It is possible (even preferable) for both the file format and the RTP
>transport to share the same MIME subtype name, but it is likely that
>there would be differences in the parameters required for the two
>cases.  The "Encoding Considerations" section needs to specify
>transfer via RTP, in addition to what it might say for other modes.

[snip]
[me]

I agree with Colin that they should be seperate as there are other
projects, such as Ogg Tarkin, which could require a different MIME type.

[snip]
[colin]
In this case, my understanding is that the file format is a generic
container that can also contain video. I don't believe it's appropriate
to use the same MIME subtype name for that and for the RTP payload
format for just the audio.

Colin

<p>-------------------------------------

This was before the discussion on this list over the holidays regarding
the different MIME type registration for different media subtypes.

Going with this approach does enforce an explicit declaration of how the
RTP stream should be handled.

An Ogg stream should provide enough hinting information to a client
application as to handle a particular media, and having this as a base
MIME type should work well.  But once we move to the raw bytestreams for
Vorbis, or other media types, I think we may need to declare them
seperate.

Naturally this is all open for discussion and the drafts can be changed
once we have nailed this.

Cheers

Phil

<p>On Wed, 2003-01-08 at 10:22, Linus Walleij wrote: 
> tis 2003-01-07 klockan 18.18 skrev Phil Kerr:
> 
> > Below is the Vorbis RTP Internet Draft as sent to the AVT working group
> > of the IETF.
> 
> > Comments and feedback is still welcomed from the Vorbis community.
> 
> That guy at Nokia who first got my audio/vorbis draft was going to join
> the drafts into one, so that the media type audio/vorbis was defined
> inside the vorbis-over-RTP spec.
> 
> Also, when talking to the AVT group they made it very clear to me that
> audio/vorbis should be used as MIME-type for *BOTH* RTP payload and
> stream based formats such as Ogg over TCP. (Letting the access method
> determine the possible encapsulation format.)
> 
> There exist double MIME content types in the IANA for e.g. audio/mpeg
> (streams) and audio/mpa-robust but these are there for historical
> reasons. The AVT group prefer one unified spec for all transport
> methods.
> 
> The AVT group may have changed their opinion on this subject, but not
> that I'm aware of.
> 
> Linus
> 
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
> containing only the word 'unsubscribe' in the body.  No subject is needed.
> Unsubscribe messages sent to the list will be ignored/filtered.

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.



More information about the Vorbis-dev mailing list