> - My idea is to reduce as much as possible the bandwidth usage in the > network, so probably the ideal scenario would be to program the camera to > multicast its stream and have the clients connect directly to that one... am > I right?
Yes - provided, of course, that the clients are all on the same LAN (or else IP multicast routing is enabled between your server and client network(s)). > - If the camera or video server is not programmable for multicasting, is it > at all possible to open a single connection to the camera from one client and > then direct my clients to "tap" on the RTP video stream to that single client > without creating a duplicate UDP video stream for each? Sorry, I don’t understand this question. (However, you might (or might not) find our “testRelay” demo application useful; see “testProgs/testRelay.cpp”.) > - I had modified the proxyServer application to rebroadcast the camera > streams but that was not very useful since still each client had to open an > RTSP connection to the specific IP address of my server, so I guess that also > created duplicated streams in the network. I will follow your advise to try > and multicast with the "testH264VideoStreamer" app and see how it goes. I > think I can use the "input" part of proxyServer and combine it with > textH264VideoStreamer to be able to re-stream a live stream via multicast, am > I right? No, because the proxy server code works by creating multiple unicast connections - one for each front-end client. It does not work for multicast front-end clients. Ross Finlayson Live Networks, Inc. http://www.live555.com/
_______________________________________________ live-devel mailing list [email protected] http://lists.live555.com/mailman/listinfo/live-devel
