I am very much new to the OSS. So may be my question is very simple or irrelevant.
Currently I am involved in a audio/video conferencing project which use OSS for sound capturing from micro phone. Then conferencing application send this captured audio from microphone via RTP to the other participants joined in the conference.
But the main problem is there is an echo because microphone picks the audio signal which was played back by speaker. To cancel out the echo I have planned to introduce the NLMS algo in my code. But this algo takes below mentioned inputs.
1> Captured audio data from microphone.
2> The audio data which was already played by the speaker.
I can easily get the audio data from microphone by reading /dev/dsp.But I have no idea how to get the audio data which was already played back by the speaker because the application which play audio is a different application and I won't have any access to its source code.
a)Is there any way to read the already played back audio data from the soundcard ??
b)Is there any other way in OSS to cancel out echo??