The prequel to the ‘A Quiet Place’ saga got me thinking.
spoiler alert!
There is a scene in which many humans march towards a safety point. Each individual human would have been relatively quiet, but because there are a lot of them (potentially hundreds), they end up being, as a whole, loud enough to alert the monsters so they get all killed.
This would suggest that many sources of noise which are near to each other and generate more or less the same amount of noise end up adding up so that the end result in dB is more or less the sum of the individual dB levels.
But then again, it’s fiction.
Back to reality, I work in a room full of different servers which have also very different levels of noise. I have noticed that from my standpoint, the noise of the quietest server seems to disappear whenever the loudest is running, so it kind of does blow my mind how our perception of noise works…
Wouldn’t a louder room raise the noise floor, too, so that any quieter signal couldn’t be extracted from the noisy background?
If we were to put a microphone and recording device in that room, could any amount of audio processing be able to extract the sound of the small server out from the background noise of all the bigger servers? Because if not, then that’s not just a auditory processing problem, but a genuine example of destruction of information.
That’s an excellent question, and like most excellent questions I think the answer is ‘it depends’ but I would bet for ‘yes’.
First, ‘noise’ is a part of the signal we don’t want or we don’t care about. In op’s example they talk about ‘two servers’ one quieter and another louder. The sounds of these sources is the part of the signal we care about so the sound of the louder server isn’t noise in any case. As long as the sound of the quieter server rises above the noise floor and the recording device have the dynamic range to record the sum of the sounds you should be able to ‘measure’ it (as in pointing to differences in the readings when you turn it on and off) and maybe cancel the other server to some extent by processing it.
Now to this point is a yes but ‘it depends’ because while the sound of the louder server isn’t noise it can rise the noise floor depending on the acoustics of the server room. Some server rooms have lots of reverb and echo, specially when they’re not very full, but in my experience most don’t, many put these plastic curtains to contain the AC and they dampen sound reflections pretty effectively. But with bad enough acoustics, a very loud ‘loud server’, and a quite quiet ‘quiet’ one the sound floor could rise over the quieter server sound.
Disclaimer: this is from the top of my head, I’m tired and not a hundred percent sure this is correct. Don’t take any important decisions based in my autistic rambling.