Re: [codec] Multi-mode frame concatenation

Good idea. We were thinking of reserving the first two bytes "Op", which
is an invalid given the framing restrictions.

	Jean-Marc

On 11-07-29 02:37 PM, Jonathan Lennox wrote:
> On Jul 29, 2011, at 2:13 PM, Benjamin M. Schwartz wrote:
> 
>> On 07/29/2011 01:53 PM, Ralph Giles wrote:
>>> Do you mean using lower values of M when the config field indicates
>>> the frame has a longer duration?
>>
>> Nope.  I mean that once M is invalid, the other two bits are also
>> available for signaling.  I would not regard M=49, p=1, v=0 (byte==113) as
>> a multi-mode packet indicator, so it's still available for something else.
> 
> It would probably be a good idea to explicitly reserve some extension values for the external container, analogous to the way that H.264 reserves NAL Unit Types 24-31.  This ensures that the RTP payload format can both support a mode where it directly places Opus packets into an RTP payload, and also (if necessary) a mode where it add some additional framing or other transport-related data around them.
> 
> --
> Jonathan Lennox
> jonathan@vidyo.com
> 
> 
> _______________________________________________
> codec mailing list
> codec@ietf.org
> https://www.ietf.org/mailman/listinfo/codec
> 
> 