A new approach is proposed that contemplates systems and methods to support canceling audio streams leaked from a speaker to a PSTN so that only the audio stream of a presenter or viewer speaking at a conference call is transmitted. Here, the audio streams being canceled includes the audio stream of an application being run or a material being presented online during a web conference. The voice audio stream of a presenter or viewer is then transmitted in addition to the computer audio stream, clean from any other audio stream of echoes or feedbacks captured by the hosting device during the conference call.