Original poster: Paul Nicholson <paul@xxxxxxxxxxxxxxxxxxx>
> [1988 Sutton/Spaniol paper]
> The Q values measured are of the order of 100-2000, limited by
> the resolution of the analyzer.
This is a very suspicious claim, and I would have to see the whole
paper to identify their errors. The suspicion arises because no
researchers, before or since, specialising in that field, have
reported anything like such high Q factors. These days it is
quite easy for amateurs to detect these resonances themselves
with sufficient S/N ratio such that a frequency modulated high Q
resonance would be most obvious. For example, the apparent
bandwidth would vary systematically with integration time,
and with short integration times (order 10 secs) the high Q
resonance would make a clear 'line' on a spectrogram, rather than
the broad band of noise that we actually get.