Hi Alessandro, currently there are two proofs of concept you can try: 1. Audio callback media port, see mailing list archive http://lists.pjsip.org/pipermail/pjsip_lists.pjsip.org/2014-March/017210.html and related messages. It is not perfect at all, but should be safe in C/C++. You provide callback routines in which you receive or provide audio frames. 2. File descriptor media port, see archive http://lists.pjsip.org/pipermail/pjsip_lists.pjsip.org/2014-March/017212.html and related messages. Here you can open pipes for receiving and transmitting the data. Anyway in C/C++ it is no problem to create own media port and add it to PJSUA conference bridge to have access to raw audio data. Currently we are trying with Benny to come up with a universal way to access raw audio data in C++ and higher-level languages. Cheers, - Vali Dne 17.3.2014 19:02, Alessandro Saccoia napsal(a): > Hi, > I'd like to use pjsip for our current project; I am trying to set up a SIP server that receives phone calls, forwards the decoded raw audio data to a pre-existing software module written in c++, and sends back to the caller some audio generated by the same c++ module. > I have successfully compiled the example applications, and I see that the functionality I need it's all there. The sipecho.c example showed me that it's possible to quickly build the answering part of the application, the playsine shows that it's possible to attach a traditional audio callback to the output channel, and simple_ua shows that I can get some useful information from the media stream of the caller. > Still, I am a bit struggling to understand how to put all the pieces together. > The goal is to have an application that whenever there is a call can instantiate and keep in the memory a structure with this interface (this is a fa?ade C interface) > > struct OurAudioProcessingLibrary { > void process(float** samplesIn, float** samplesOut, int numSamples); > }; > > but I haven't found a way until today. This signature can be thought as an audio processing black box, or an audio fx. I ask if somebody can put me in the right direction, or if someone is available to work for us during the development of this software. > Thank you > > Alessandro Saccoia > www.dinahmoe.com > _______________________________________________ > Visit our blog: http://blog.pjsip.org > > pjsip mailing list > pjsip at lists.pjsip.org > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org >
