- Will the difference between a 90 dB S/N component and a 120 dB S/N component be audible in a typical listening room with a 30 dB background noise level?
This is a misconception. I actually published a paper on this a few years back based on peer review journal article and research at AES. Our hearing sensitively is highly non-linear. 60 dB at 20 Hz will be silent to us where as -5 dBSPL is audible at 3 kHz! So to know if room noise is an audible barrier, you have to perform a translation and then superposition it on top of our hearing sensitivity. This work was done based on survey of listening rooms and here are the results as I post in my article:
As you see, room noise drops exponentially with increasing frequencies. This is because it is so much easier to block high frequency noise. Low frequency noise on the other the hand, easily travels through walls and even concrete! Your meter will be happy to pick up noise from freeway from miles away, yet your ear will not.
As you see in the above graph, we can build rooms that are completely silent even though they could have bass noise 30 dB.
Using this research, we see that if we want to have reference quality maximum loudness (measured at 120 or so dB in live unamplified concerts), we need about 20 bits of dynamic range or 120 dB. Put inversely, ideally you want your system to have a noise floor of -120 dB.
Now if you are listening in a living room, then such constraints can be lowered fair bit. But why? You can buy a $100 DAC that performs this well, once again proving that achieving full transparency in electronics doesn't cost much money. Just requires careful shopping and not following marketing statements of fidelity but reliable measurements.
