openmeetings-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "seba.wagner@gmail.com" <seba.wag...@gmail.com>
Subject Re: GSoC project ideas wanted
Date Mon, 11 Feb 2013 22:12:31 GMT
Hi Ed,

thanks for your ideas.

About idea no 1:
That is an interesting idea, however it won't be possible that you provide
a "free to choose" bandwidth for each user.
The background is: Every stream that any client consumes has to be created
somewhere.
So what could be realized is that every stream that is broadcasted from one
user via webcam to Red5/OpenMeetings will be re-transcoded into multiple
streams (high, middle, low) bandwidth.

So there might be some limitations to that:
 - "high" quality will never be better then the original material. We can't
make a picture better then the original. So all re-transcoding will only
make the original to lower quality, never to higher.
 - Re-transcoding has to happen on the server side (and number of streams
are limited, we can't provide a stream on the required bandwidth
"on-demand" for each user, or only with very big effort)
 - it will require real-time transcoding on server side which is possible
with FFMPEG and some integration into Red5. But we would need a very
specialized student that is keen and very motiviated as there is hardly any
documentation on that available in the internet.
What a project makes a success is if all participant know the potential
outcome and the tools and methods that are needed to realize that. I would
be happy to put this project on our list but it will be difficult to find
somebody with the needed skills.

Sebastian



2013/2/12 BBS Technik <dormitilla@gmx.de>

> Hi all,
>
> I think, one of the gratest liminations for satisfactory video
> conferencing with om is the limited bandwidth of internet connections of
> the clients .
> Therefore I would like to suggest the following ideas for a GSoC project :
>
> 1. The image size of the videos transferred from the om server to the
> clients should be adapted to the video window size set in the recipient
> client.
> Thus the recipient client itself could influence the transferred amount of
> data to it.
> Then all the participants achieve the best possible result for them.
>
> 2. A second proposal concerns that the screensharing  bandwidth
> requirements has an great impact on the overall quality of the video
> conference.
> Here, in a project the existing function of sreen sharing could be
> expanded and enhanced.
> For example, the possibility for the transfer on only one application
> window, regardless of its size.
> Or the possibility of shared browsing with a locally installed browser.
> Moreover, certainly an improvement of the used compression method would be
> a very good project topic.
>
> I would be  happy if the subject of bandwidth consumption would plays a
> role in the selected GSoC project .
>
> Best regards
>
> Ed
>
>
>
>
> -------- Original-Nachricht --------
> > Datum: Tue, 12 Feb 2013 08:44:54 +1300
> > Von: "seba.wagner@gmail.com" <seba.wagner@gmail.com>
> > An: dev <dev@openmeetings.apache.org>
> > CC: user@openmeetings.apache.org
> > Betreff: GSoC project ideas wanted
>
> > Google Summer of Code is about to start soon!
> > Google sponsors every student with 4500USD. Plus 500 for the Apache
> > Foundation.
> >
> > We are searching for ideas what porential students can do.
> > Ideas from Non-Developers are welcome too!
> >
> > We will add the ideas to JIRA then with a special label so students can
> > find it.
> >
> > Sebastian
>



-- 
Sebastian Wagner
https://twitter.com/#!/dead_lock
http://www.webbase-design.de
http://www.wagner-sebastian.com
seba.wagner@gmail.com

Mime
View raw message