Mildly off topic but indulge me, or ignore me, all good!
I once went for a job interview, the company specialised in seismology instrumentation, sat in their offices near Aldermaston they could see the frequency of the waves lapping on the shore of the UK and earthquakes round the globe ... vibration is an interesting subject, I didn't accept their offer but did go on to work for two more UK manufacturers of vibration monitoring equipment. One of the tools of the trade is an FFT, however you soon learn that while big spectral spikes can tell you a lot, there are other, better ways, of digging out smaller signals from noise. For example ways to find harmonics hidden in the "grass". Stuff you can't see but is there in the data. The audio sweeps REW does are great for comparing roughly in room responses and identifying low frequency modes, but the idea they will tell you definitively if there is any audible change is complete and utter hogwash. It is reasonably hard to get repeatable results from such things, while changing nothing. I don't want to do down REW, it is a great, free, tool. It has its uses. More than I know of I don't doubt. But a spectrum is not the be all and end all of signal analysis, whether vibration (which is broadly the same band) or audio.