The Pyramic architecture is a flexible 48-microphone array mounted on a powerful FPGA-SoC system. It allows to do massively multi-channel audio processing in real-time. This project implements an extension to the original system. In the original system, audio output is not allowed from the hardware processor system of the SoC, thus we cannot perform simultaneous capture and output. We implemented this through the design of an audio output controller, as well as a library (libpyramicio) that drives the capture and output from the software part and makes it easy for application developers to create algorithms that run on the Pyramic array.