The recordings based on motion detection have no audio. The audio packets are dropped in the hls_unified_thread.c Based on the documentation the motion buffer should store audio as well. (Am i getting things right?)