How is the work on transcoder for Video?

2 messages Options
Embed this post
Permalink
szhairui

How is the work on transcoder for Video?

Reply Threaded More More options
Print post
Permalink
On the roadmap, we can see the transcoder for any -> mpeg2 is on working now.

Now I am start my work on writing a backend module after reading some documents
from upnp.org and the sample backend source files, especially the lolcats_storage.py.

Because my backend is major for video files playback, I am eager to know the progress on
video format transcoder. Can someone show me some hints?

In my opinion, the best way for transcoding video files is using ffmepg. I guess you are trying the same way too. :)

Will the first version of video trancoder support operations from Control Point such as
REW, FF, seek ? or just the play and stop?


Frank Scholz

Re: How is the work on transcoder for Video?

Reply Threaded More More options
Print post
Permalink
Hi szhairui,

> On the roadmap, we can see the transcoder for any -> mpeg2 is on working now.
it is not in svn trunk yet. But all pre-work for the 1st generation of
the video-transcoder is done. It just needs to be included into the
transcoder.py code.

> Because my backend is major for video files playback,
what are you working on? Is it something we can add to
the backend list?

> I am eager to know the progress on video format transcoder.
> Can someone show me some hints?
>
> In my opinion, the best way for transcoding video files is using ffmepg. I
> guess you are trying the same way too. :)
we have two options.

1) a GStreamer based one - more or less the same like
   the audio-transcoder, and the necessary pipeline elements
   are afaik in all major distributions by now
2) an external process (mencoder,...) based on the
   ResourceItem and the ExternalProcessPipeline from the TestStore

> Will the first version of video trancoder support operations from Control
> Point such as
> REW, FF, seek ? or just the play and stop?
the first version will only have play,stop and pause.

Byte-based and especially the DLNA time-based movement in the
transcoded stream will be in the second version. But I'm sure
that the way it is designed now these additional features won't
be too difficult to add.

Regards,
Frank
_______________________________________________
coherence-dev mailing list
[hidden email]
http://lists.beebits.net/cgi-bin/mailman/listinfo/coherence-dev