This relates to the noise and music discussed in a previous thread
http://audiomastersforum.net/amforum/index.php/topic,6952.0.html
For testing, I used a song that was one voice and one acoustic guitar (Peak = -8dB, .RMS Avg = -28dB). (Real life radio play is of current hip-hop, R&B, and rap).
I generated three noise files in CoolEdit, one white, one pink, one brown. I adjusted their RMS Avg to equal the song's RMS Avg. I tried mixing the noise files, one at a time, with the song. In every case, most or all of the music was easily audible.
I then mixed the three noise files into one and adjusted its RMS Avg to equal the song. Again the vocals and guitar were easily enough heard.
I tried another noise-song mix with the combined noise amplified 6dB above the song, then another with the noise amplified 12dB. Even with the noise RMS Avg 12dB above the song, I did not seem to be missing a lot of the music (not that listening was in any way pleasant).
I know that the real world noise is not very close to what I used, but I did not have a source for more realistic noise. The factory noise does not have any where near as much high frequency and, while there are quite a few different sources, it's overall frequency distribution seems very uneven, with many peaks and holes. The mix is not constant. Still, I am hoping it is not totally unreasonable that some general statements might be made.
First, if anyone wants to argue that trying to listen to music under such noisy conditions is a lost cause, I completely agree. I don't understand the people who want to listen. Regardless, the circumstances exist. There are posted rules, one of which says "keep radio volumes low" but it is not enforced. Requests bring only very short term relief.
So, the basic question is:
if a PA system were installed and the volume setting was locked, what
objectively determined setting would be necessary for it to be reasonably said that "Yes, you can hear the music."
What might be a reasonable compromise that, while it might not satisfy anyone (a few want very loud, I want none), would give each side something useable?
Would the PA system have to put out as much sound as the background, something more, such as +3dB or +6dB, or might -6dB, as in my simple experiment, have some real legitimacy?
Lets start by assuming the background noise is less extreme than I expressed earlier, something less than immediately dangerous, such as between 75dBA and 80dBA.
Author
