[asterisk-dev] iLBC packet loss concealment (was: code-cleanup concerns)

[Steve Underwood]

Denis Smirnov wrote:

>On Mon, Apr 17, 2006 at 07:04:27PM +0800, Steve Underwood wrote:
>  
>
>>>transcode_via_sln, that used by default, is a VeryBadThing (tm)
>>>      
>>>
> SU> I think very bad is overstating it. However, its not right. The ITU 
> SU> specs define specific ways to go ulaw->alaw->ulaw and 
> SU> ulaw->G.726->alaw->G.726->ulaw, and those are not followed. If they 
> SU> were, tandem operation would be better. I'm not sure that is a huge 
> SU> issue, though. Most of the transcoding is for things like G.729 or 
> SU> Speex, where the only way is via 16 bit linear audio.
>
>Without this missfeature we can transcode, e.g, from 32k speex to 16k
>speex (if we define it as separate codecs) without translating to 8k
>slinear.
>  
>
That is a slightly different issue, as it involves rate changes. 
However, it really doesn't make sense to do this in practice, though. 
The two ends should negotiate to use a common variant of Speex. 
Transcoding between variants is silly. Even when there are branches in 
the path, such as with conferencing, it doesn't really seem like much of 
a gain.

Steve