Hello pjsip-team, I still try to implement the idea of spartial hearing into the pjsip conference. I tryed a lot and got it working manually by opening a wav_player port, getting the sound frames manually in a loop and putting the new, special stereo frames into a 2 channel wav_writer port. Therefore I already integrated a convolution engine. So far everything works fine. Nevertheless the next step is to integrate this feature into the pjsua application and running the convolution part automatically (with the master port/clock !?). I primarly worked with the pjmedia lib and I am not sure where to integrate the convolution part best in pjsua. I have to take care that all media ports are opend in stereo mode. I do this by using the parameter "--stereo" and tested the channel numbers every time the "pjmedia_conf_add_port" function is called. Which function do you recommend to implement the convolution part. The best case would be to get a pointer to the media port (and some details like size of the frame + channel number, to be sure). If I get a media port pointer + the size + knowing it is stereo, I can take the frame (left and right part included) and put it to my convolver which will overwrite the value of the actual frame => spartial hearing is integrated :-). I saw that sometimes links use the pjsua_conf_connect function but sometimes the pjsua_conf_add_port function. I'm not sure what the difference by using those function is or when to use which one of both. Thanks for support MatzeMuc86
