Various example implementations are directed to methods and apparatuses for facilitating conferenced communications. An example method can include receiving from a plurality of participants of a digital audio conference, audio signals from respective audio input circuits for each respective participant. A subset of the plurality of participants may be identified, as corresponding to loudest audio input compared to a remainder of the participants. At least one encoder pool may be generated by assigning each of the remainder of the participants to an encoder pool associated with a codec used by the respective audio input circuits. For each participant in the subset, digital audio packets of the loudest audio input may be separately encoded to remove feedback. Digital audio packets of the loudest audio input may be separately encoded using the at least one encoder pool, and the encoded data packets may be broadcast to the plurality of participants.