At this moment we’re gathering audio from both microphones (top and bottom one) and apply quite strong echo cancellation - I guess that’s where the “muffled” voice comes from. Echo cancellation parameters can still be tweaked (I settled on a set of params that’s pretty close to PulseAudio’s defaults so far), and of course we could only use a single microphone from the other end than the currently active speaker; however, I haven’t figured out a clean way to do that yet as there’s bunch of non-obvious obstacles on the way there. Let’s see if we can come up with a clean solution before we resort to ugly hacks though 
14 Likes