Does the meetme application translate all incoming and outgoing frames back and forth from PCM (signed linear)?
I.e. all users, connecting to a conference, call in using any (different) codec(s) but the mixing and the actual inner workings of the conference is all performed using the PCM (signed linear) codec?
Such a clever solution would then require a lot of transcoding (twice for each participant), right?
If there is something I have misunderstood, please enlighten me (by comment or source of knowledge [links]).