Re: [codec] WG Review: Internet Wideband Audio Codec (codec)

[Ron <[e-mail]>]

On Thu, Jan 21, 2010 at 02:44:36PM +0100, [e-mail] wrote:
> In line as well : The first stage of the work has not been done yet : the
> detailed technical requirements have not been defined and agreed yet, the
> second stage of the work with other SDOs to analyse if already exiting codecs
> meet these requirements is not done and yet the Charter on the basis of which
> this WG could be launched is formulated as if the conclusion resulting from
> this 2 stages was known and obvious : no existing codec is suitable and so a
> new codec is needed
>  
> so, again : only these first 2 steps (requirement definition and standard
> analysis) are, at this stage, relevant to start a WG since the next steps
> depend on the conclusion of this work.
>  
> for instance , there could be some middle way between developping a new codec
> and reusing an existing standard : it could be much more efficient to simply
> extend/adapt an existing standard and, to achieve this, it would be better to
> rely on the SDO that have standardized this codec. 

Potential candidate codecs have been discussed since the very first BoF,
and surely before.  We have plenty of representatives from other SDOs
present in the group, so my question is, if such codecs exist, then surely
they should be put up, if not as contenders, then as baseline measures of
best practice in some matter or another.  Why hasn't that happened?

With the exception of a few "G.711 should be enough for anyone" jokes,
I haven't seen anyone present a "this can already do what you want" codec
that the group hasn't responded with "we can (already) do better than that".

If there are other codecs that we should have to measure up against, then
indeed I'd warmly welcome other SDOs to make their best suggestions soon,
for consideration during the next stage of work.  Anything you think puts
one of the existing candidates to shame in some aspect or another, would
certainly be a valuable and practical measuring stick we can use.

I'm pretty sure a lot of people have already done many of these measurements
for themselves.  We just need to whittle that down to the ones that give us
useful bars to raise.  Which ones are we missing still?

Cheers,
Ron