Human-perceived loudness !== Noise level

One of the issues I'm facing at the moment is trying to figure out the easiest way to overcome the issue of human-perceived loudness !== noise level (because the human audible range doesn't encompass the entire frequency spectrum).

Human-perceived loudness can be measured on a decibelmeter but we need to figure out how we want to do this:
- do we want to find a microphone that can only detect noise between a certain frequency range (so that it aligns with the A weighting curve) or
- do we want to do some analysis on the noise that we get (like weighting the high frequency noise more) from a decent resolution microphone so that we can capture "all sounds"?

I think the latter makes the application more versatile; we'd be able to recycle the equipment we use for other things; for example, I think it would generally be quite interesting to see (whether there are any) differences in white-noise, with the number of machines in the space!

@nelsonic - please could I ask what your thoughts are on this?