Recently, the LIGO detections of the gravitational waves from black hole merger events have rightly caused a lot of media attention.
There is something I do not quite understand, however. In the media the gravitational wave signal is portrayed as stemming from the merger event itself. However, due to the fact that time (as measured by an observer far away) slows down close to the horizon of either black hole, this should be impossible. In fact, for a Schwarzschild black hole, you should never see anything "fall" into the black hole as time would progressively slow down to a standstill (and any photons be redshifted to infinite wavelength) in the limit where you cross the horizon. I expect the same to be the case even in the case of a more complicated geometry of two orbiting Kerr black holes.
The LIGO website (http://www.ligo.org/science/Publication-NINJA2/) explains that LIGO should detect three signals from the merger event related to what is called inspiral, merger and ringdown. But would each of these events not be obscured by the slowdown of time (again, as measured by a clock far away from the event)? In particular, as defined by the link, the ringdown signal is the waveform formed after the merger, but since the merger is in the infinite future of a distant observer, there should be no after.
What is actually detected - in the Earth's reference frame - in this supposedly cataclysmic event of a black hole merger?