Hi Red,
As a couple of others have already noted, the only way to get to grips with this problem is to gather some data for one or more music styles and characterise it. A basic empirically derived model can then be devised.
I spoke about this problem in another thread:
"Some time ago I was wondering the same thing. I used to consider the music spectrum to be have similar power per octave as pink noise (-3dB/octave). But I found that this approximation was too bass heavy. In the absence of any hard data except for a very old AES article that measured SPL in a club situation, I decided to measure it for myself.
I did some tests on the spectra of popular music, including rock, pop, electronic, dance music. I chose these styles and omitted classical and jazz as I tend to listen to the popular styles at higher volumes and the purpose of the tests were to determine the relative maximum SPL's required across the frequency spectrum.
The results showed an almost exact -1.5dB/octave drop as frequency increased from about 45Hz to 10kHz. In the octave below 60Hz the amount of power dropped 12dB relative to the octave above it. In the octave above 10kHz the power dropped by 10dB relative to the octave below it.
Part of the drop in the top octave may be due to the nature of the source material which was all MP3 encoded."
So, to answer your original question using this data, I figure the 1/2 power frequency to be about 150Hz. For three bands, about 70Hz and 200Hz.
Cheers,
Ralph
As a couple of others have already noted, the only way to get to grips with this problem is to gather some data for one or more music styles and characterise it. A basic empirically derived model can then be devised.
I spoke about this problem in another thread:
"Some time ago I was wondering the same thing. I used to consider the music spectrum to be have similar power per octave as pink noise (-3dB/octave). But I found that this approximation was too bass heavy. In the absence of any hard data except for a very old AES article that measured SPL in a club situation, I decided to measure it for myself.
I did some tests on the spectra of popular music, including rock, pop, electronic, dance music. I chose these styles and omitted classical and jazz as I tend to listen to the popular styles at higher volumes and the purpose of the tests were to determine the relative maximum SPL's required across the frequency spectrum.
The results showed an almost exact -1.5dB/octave drop as frequency increased from about 45Hz to 10kHz. In the octave below 60Hz the amount of power dropped 12dB relative to the octave above it. In the octave above 10kHz the power dropped by 10dB relative to the octave below it.
Part of the drop in the top octave may be due to the nature of the source material which was all MP3 encoded."
So, to answer your original question using this data, I figure the 1/2 power frequency to be about 150Hz. For three bands, about 70Hz and 200Hz.
Cheers,
Ralph