[prev in list] [next in list] [prev in thread] [next in thread] 

List:       semsdev
Subject:    [Semsdev] a question about conference servers
From:       rco () iptel ! org (Raphael Coeffic)
Date:       2005-11-10 17:06:49
Message-ID: 200511101656.20108.rco () iptel ! org
[Download RAW message or body]

On Friday 04 November 2005 07:20, Jorge C wrote:
> Hello everybody,
>
> I have a question about the conference server. I wanted how the mixer part
> is done. Specifically, how it could be implemented. For example, should the
> input streams (and so the input samples) from different participants be
> interleaved to build the output stream? or should they be added in linear
> way? are there other alternatives?

It depends on what you want. If you want to mix some streams into one, then 
you have to add the samples (signed format only!) and use some kind of gain 
control to avoid the addition to overflow the output. Then, you should not 
forget that every participant don't want to hear its own stream, so you have 
to substract it (or make a different addition for each participant).
Alternatives are, let the client do the mixing and do only source selection 
and interstream synchronisation.

You can search for that on google... there are a lot of papers on that 
subject.

regards
Raphael.

[prev in list] [next in list] [prev in thread] [next in thread]