[MMUSIC] Simulcast worries: What bitrate? (from the WEBRTC list)

[Harald Alvestrand <[e-mail]>]

Hi,

I had some worries about simulcast in the thread called "Towards
Simulcast" - Adam Roach asked me to post them here so that they can be
discussed in the right forum.

I became a little less worried after reading version -01 of the RID
draft, but I'm still worried. Uncomfortably many loose ends.

I'll raise a simple one in this message, and another message on my more
complex worry.

In -00, bitrate was undefined. In -01, bitrate (max-br) is defined as:

   o  max-br, for bit rate in bits per second.  The restriction applies
      to the media payload only, and does not include overhead
      introduced by other layers (e.g., RTP, UDP, IP, or Ethernet).  The
      exact means of keeping within this limit are left up to the
      implementation, and instantaneous excursions outside the limit are
      permissible.  For any given one-second sliding window, however,
      the total number of bits in the payload portion of RTP SHOULD NOT
      exceed the value specified in "max-br."

Question: Is this definition compatible with any of our other bitrate
definitions?
(I like it - it's more actionable than other bitrate definitions that
don't specify the measurement interval for conformance)

I'd be happier if we could say "this is the same bitrate as b=TEAS" or
something like that - if an implementation has to stick to multiple
definitions of bitrate for the same media flow, which will happen with
a=rid .. max-br and b= in the same message, it would be nice to be able
to use the same counters.


Harald