Re: World's Most Reverberant Room
Yes, B&K's best will get you close but not quite there. Perhaps there are some completely different microphone technologies that have very low noise. After all, unlike most audio applications, we don't need a lot of dynamic range for these measurements, just a very low noise floor. Superconductors and the tunneling effect (as used in the scanning tunneling microscope) come to mind, but I'm just making this up.
I think the trouble with narrow-band measurements is that, in general, the noise we're trying to measure is broad-band, so its power is reduced by the same ratio as the measurement system's noise and the signal-to-noise ratio remains constant. Were the signal we're trying to measure narrow-band, then band-limiting works. This is essentially what synchronous detection does.
Back in the late '70s I worked on displacement transducers with solid 100 A (10^-8 m) resolution that used linear variable differential transformers and synchronous detection, but they only had a bandwidth of a few Hz -- neither fast nor likely sensitive enough for this.
Another thing I wonder about is what is the thermal acoustic noise floor of air at room temperature (in a given bandwidth)? It must have one. Someone handier than I with that kind of physics probably can figure it out pretty easily.
Still, if I were charged with estimating the noise floor in the chamber I think I'd put a speaker out in the street (actually at more than one location) and produce a narrow band test signal that could be synchronously detected (or otherwise signal-averaged) in the chamber. This would give an estimate of transmission loss from the street to the chamber. Then do a regular noise measurement outside, subtract band-by-band, apply the weighting, and I'm done.
The big assumption here is that all the noise that matters is coming from "the street". Internally generated noise or noise from other sources, say transmitted through the earth, would be missed. I'm also assuming that the transmission medium is linear, but that's a pretty safe bet.
I believe some of you folks here do this sort of thing for a living. How would you do it?
--Frank
