[Live-devel] I am new to LIVE555..

Shaswata Jash shaswata at alumnux.com
Wed Apr 30 08:53:20 PDT 2008


Hi Manas, 

 

My pointers are inlined-

 

Q1:I want to capture and encode both video and audio and stream it to
another end and receive RTP packets from the other end and  decode and play
it. So whether I have to run two RSTP programmes ie. both server and client
? . Can't I do this in a single programme.



Ans: Before deciding on whether both RTSP server & client need to be
implemented, you can once give a thought whether RTSP will fit into video
telephony scenario. SIP is certainly better choice. How are you going to
discover the IP address and port where the callee is listening for request
from a caller (in SIP term the 'Contact' header)? 

However, for the time being if you have decided to keep your callee and
caller with a static address (and that is reachable with respect to each
other), you can go with RTSP. In that case, the caller must act as RTSP
client and the callee must act as RTSP server. 

But note that both of the callee and caller must act as RTP client & server.


Q2: I went through the  test programme . I found that we can play only mpeg1
audio and video together. Can't I use mpeg4 and mp2 audio together and also
stream them together.



Ans: Yes, you can. You have to write your own program that will chose
appropriate RTPSink and sources. Though I may be wrong, but combination of
mpeg4 and mp2 is quite non-standard. For speech, there are much lightweight
codec, isn't? 


Q3: Is the live555 media server takes care of packet losses in the network
and synchronization between the sender and receiver when both are sending in
either directions. Or  I have to write my own code to do that.



Ans: To best of my knowledge, Live555 doesn't have any exclusive mechanism
to handle packet loss. The synchronization mechanism is essentially limited
to the fact that the server tries to throw the packet in the network at
appropriate time interval (or more technically according to PTS).

 


Q4: The  encoder  put the  encoded  video and  audio  data in two separate
file and the decoder can take video and audio data from two separate files.
If this is my case can I implement the live555 programe for my purpose
effortlessly as it can take the files as input and output..



Ans: The file approach will degrade performance for real time purpose. You
can develop your own source classes, which can directly take the encoded
data and may store in some in-memory data-structure. Then, the appropriate
sink will form RTP packets using the encoded data stored within that custom
source.

Reverse will be the case in decoding side.

 

Regards,

Shaswata

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.live555.com/pipermail/live-devel/attachments/20080430/840205a8/attachment-0001.html 


More information about the live-devel mailing list