If you try to slow the video, or advance slowly, it seems as the coning angle is all over the place on takeoff. First it may appear normal, but seems to vary, then right around 9 seconds, when just over the last building, you can see the rotor tilt left, to the camera, then back forward. Then the cone angle drops, and raises again, before the rotor drops a lot around 10 seconds, one frame seems like blades are crowded in the front, spreads out flat again and one blade departs. The two other puma departures had intact heads and blades. I would not rule out a gearbox seizure as some have speculated, but probably not my first guess. But since this event was takeoff rather than cruise, all control positions and airflow vastly different, so the subsequent chain events could be different. Still, usually rotor heads and the blade roots tend to stay grouped, even after powered impact. This blade is missing at the head altogether, so I would suspect a blade attachment, or at least a blade control mechanism that could send this to extreme pitch, that would produce an extremely high load in the flatwise direction rather than chordwise, before separation. Who has graphics of rotor/ control system? Assume they are same as the AS332.