In conferencing applications, speech and audio quality can be degraded by background noise and microphone artifacts like popping. Also, the level (loudness) of speech can vary over time due to changes in the speaker s position. In this paper, we describe a system that combines a dynamics processor and a de-popper to improve the quality of speech in abovementioned cases. First, a short overview of dynamics processing and its application to speech level stabilization and pop removal is given. Several possible integrated architectures, combining these applications are compared using MATLAB simulations. Our real-time implementation of one of these architectures on the Texas Instruments TMS320C6713 DSK platform is discussed.