Jerome,
We are discussing audio experiments with conference calling. The first
idea is to present several arriving mono sound channels as a stereo mix.
Stereo can make it easier to recognize who's talking.
The mixing-to-stereo can be done assuming a default positioning for each
participant, but I like the idea of being able to arrange who's where.
And being able to conveniently control the volume of each participant -
very useful.
I have a couple of tools I'd like to try out in this context. There's a
little GUI for driving "the mix"; and there's an implementation of the
audio mixer.
The computational mixer in its simplest form wants compatible WAVs to
mix (e.g. 16 bit at a given sample rate), and it needs to be hooked up
where it can feed the near-real-time audio output. For example I have it
preparing the data just before it's needed by a PortAudio callback.
Here's what I am looking for at this point:
- Are there any Issues that would get in the way of trying this?
- Where in the code will I find the right audio streams to mix? The
right audio output to feed? I'm looking for keywords to search on in the
source.
- How do you test conference calling?
Kent
Jérôme WAGNER wrote:
Hello Kent,
That is very interesting. I have already heard of some 'spatialization'
libraries for sound but have never really seen one work.
This sounds like a very interesting feature for conference indeed.
Right now, the sounds are played in MONO mode because they come as MONO
streams from the network, be it should be possible to play them as stereo
and spatialize them as you said.
What kind of insight would you need to start working on such a
library/feature ?
Jerome
-----Message d'origine-----
De : [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] De la part de Kent
Langseth
Envoyé : jeudi 12 octobre 2006 13:53
À : wengophone-devel@lists.openwengo.com
Objet : [Wengophone-devel] Re: Conference call audio mixing
The network topology is relevant insofar as it delivers low-latency
audio streams from multiple sources. Mainly my interest is in providing
the best local experience given these audio streams. So far I've found
the most dramatic enhancement involves arranging sound sources in the
user's hearing space.
Envision a conference call with several people. Somewhere in the
software the digital audio streams have to be put together for local
audio output. That's where I want to work.
Kent
Jérôme WAGNER wrote:
Hello Kent,
That's a very interesting idea. WengoPhoneNG is a good start base to do
something like that cross platform.
Could you describe a bit more what angle you take on that subject ?
Are you more interested in the network topology aspects of the p2p
conferencing or in the signal theory aspects of mixing numerous channels
with spatialization of the sound ?
Jerome
-----Message d'origine-----
De : [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] De la part de Kent
Langseth
Envoyé : mercredi 11 octobre 2006 20:28
À : wengophone-devel@lists.openwengo.com
Objet : [Wengophone-devel] Conference call audio mixing
Hello,
I am interested in implementing audio mixing algorithms for p2p
conference calls, generating stereo locally. Can anyone tell me whether
this will be reasonable within WengophoneNG? What should I look at?
Thank you.
Kent
_______________________________________________
Wengophone-devel mailing list
Wengophone-devel@lists.openwengo.com
http://dev.openwengo.com/mailman/listinfo/wengophone-devel
_______________________________________________
Wengophone-devel mailing list
Wengophone-devel@lists.openwengo.com
http://dev.openwengo.com/mailman/listinfo/wengophone-devel
_______________________________________________
Wengophone-devel mailing list
Wengophone-devel@lists.openwengo.com
http://dev.openwengo.com/mailman/listinfo/wengophone-devel
