Kind of? He knew the day of the video and that it was in the US, and the data contained a 2 Doppler signals created by planes. He referenced the time gap between the two of them and applied that as a filter to all flight paths in the US. He further eliminated locations by specifying that there was silence between the two Doppler signals, and further again by calculating the Doppler signal shapes of the remaining flight paths and running a match test of each against the original audio. This left only a small search area, which he then shrunk even more by noting another audio signature.

He points out that theres many other methods than this that are easier to do for pinning down a location, but also showed that you often cannot actually completely eliminate this audio data either.