Does anyone believe that this is done algorithmically? All the examples I checked seemed to be really poorly sync-ed, & as a DSP engineer I don't believe this task to be incredibly difficult...
would also be cooler to have a seating chart to click around to make sense of perspectives