Thanks all, The scenario is odd... Its a intercom with some funky features. 3 ALSA inputs/outputs for headsets/microphones, all audio from all microphones should go to all earpieces. There is one special case input; A microphone is placed in the ambient noise (in an aircraft) The audio is inverted and added to the other 3 headset outputs - this should give Active noise cancellation; or some effect close to it. It is easier to do using discrete op-amps, but this is more fun :D Richard <div class="gmail_quote">On Aug 19, 2012 8:03 AM, "Emile Semmes" <<a href="mailto:emile.semmes@e6group.com">emile.semmes@e6group.com</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> On 8/18/2012 12:23 PM, Tim-Philipp Müller wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> On Sat, 2012-08-18 at 17:43 +0100, Richard wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Is there a component that is capable of combining/mixing multiple audio tracks in to one? I am reading audio from multiple ALSA devices and need to combine in to one stream and then output to multiple audio sinks. </blockquote> Perhaps the "adder" element is what you're looking for? If not, maybe the interleave element? Cheers -Tim _______________________________________________ gstreamer-devel mailing list <a href="mailto:gstreamer-devel@lists.freedesktop.org" target="_blank">gstreamer-devel@lists.freedesktop.org</a> <a href="http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel" target="_blank">http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel</a> </blockquote> Tim, You beat me to the punch. I second using adder too. Can you just have a pipeline from one alsasrc directly to the alsasink or do you need something like an output-selector for dynamic alsasink selection? Is this on the same device for immediate playback or can you save this to a file for later playback? There are a couple of ways to skin this cat. Note: these are prototype pipelines. I haven't tested them. Assuming you're using multichannel inputs, you could try using the interleave and deinterleave elements but that's kinda hackish alsasrc device="hw:1,0" ! deinterleave name=d1 \ alsasrc device="hw:1,1" ! deinterleave name=d2 \ interleave name=i \ d1.src0 ! i.sink0 \ d1.src1 ! i.sink1 \ d2.src0 ! i.sink2 \ d3.src1 ! i.sink3 ... and then the reverse to get to your alsasink. If you need to swap inputs to outputs, you'll have to throw in output-selectors. That's going to be a pretty complex pipeline, but it may do the trick. (This actually sounds like fun. Let me know if you need help doing this one) For later playback, if you can mux them into a quicktime container. Your pipeline would look something like: alsasrc device="hw:1,0" ! queue ! mux.audio_00 \ alsasrc device="hw:1,1" ! queue ! mux.audio_01 \ alsasrc device="hw:1,2" ! queue ! mux.audio_02 \ qtmux name=mux ! filesink You might have to do some tricks with qtmux so you can play it back. qtdemux was complaining about an empty atom during demuxing. Have fun, Emile _______________________________________________ gstreamer-devel mailing list <a href="mailto:gstreamer-devel@lists.freedesktop.org" target="_blank">gstreamer-devel@lists.freedesktop.org</a> <a href="http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel" target="_blank">http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel</a> </blockquote></div>