Science of crowd noise?
November 2, 2010 1:03 PM Subscribe
Why does crowd noise sound like crowd noise?
I'm sure this is really obvious to everyone but me. In large sporting events -- let's say televised sporting events, just for specific context -- the standard crowd noise during regular action all sounds the same. The World Series sounds like the Super Bowl sounds like the World Cup. When a team scores, there's a big, unifed "YEAH!" But the rest of the time, there's this very specific sounding white noise.
Is there some kind of explanation from physics or acoustics that explains why one voice laid on top of another voice laid on top of another, when all of those voices are saying different things, has that 'crowd noise' sound? Is there a minimum number of voices at which it starts? It kind of exists in a crowded theatre, but it's not really the same effect.
I'm sure this is really obvious to everyone but me. In large sporting events -- let's say televised sporting events, just for specific context -- the standard crowd noise during regular action all sounds the same. The World Series sounds like the Super Bowl sounds like the World Cup. When a team scores, there's a big, unifed "YEAH!" But the rest of the time, there's this very specific sounding white noise.
Is there some kind of explanation from physics or acoustics that explains why one voice laid on top of another voice laid on top of another, when all of those voices are saying different things, has that 'crowd noise' sound? Is there a minimum number of voices at which it starts? It kind of exists in a crowded theatre, but it's not really the same effect.
Best answer: You've (almost) answered your own question.
White noise is an equal (random) distribution of frequencies. Get enough human voices going, who all have different resonances of their vocal tracts (male, female, adult, child) and there is enough random frequencies to start to resemble white noise. But the human voice doesn't cover the whole spectrum, nor is the amplitude of all the frequencies flat. So, on average, the crown noise will be similar to white noise, but have the distinct characteristic that you associate with crowds.
I don't know what the minimum number of people is required for this effect, but I bet it could be estimated.
posted by achmorrison at 1:13 PM on November 2, 2010 [1 favorite]
White noise is an equal (random) distribution of frequencies. Get enough human voices going, who all have different resonances of their vocal tracts (male, female, adult, child) and there is enough random frequencies to start to resemble white noise. But the human voice doesn't cover the whole spectrum, nor is the amplitude of all the frequencies flat. So, on average, the crown noise will be similar to white noise, but have the distinct characteristic that you associate with crowds.
I don't know what the minimum number of people is required for this effect, but I bet it could be estimated.
posted by achmorrison at 1:13 PM on November 2, 2010 [1 favorite]
Best answer: one additional point not made above is that for a single person speaking, the frequencies he or she produces are changing in time. at certain times like when they are saying a vowel like "oo" there will be a well-defined peak in the frequency spectrum, other times not so much. otherwise that person's vocalizations would just sound like sine waves or colored noise or something.
but once you add in multiple voices and assume that they are out-of-phase with each other, the frequency spectrum at any given time will become more flat, since you are averaging over many people's time-varying frequency spectra.
on the other hand, when everyone says "yeah" together, it's not accurate to assume that everyone is out of phase with each other. so you get a noticeably time-varying frequency spectrum yielding an intelligible word.
posted by alk at 1:27 PM on November 2, 2010
but once you add in multiple voices and assume that they are out-of-phase with each other, the frequency spectrum at any given time will become more flat, since you are averaging over many people's time-varying frequency spectra.
on the other hand, when everyone says "yeah" together, it's not accurate to assume that everyone is out of phase with each other. so you get a noticeably time-varying frequency spectrum yielding an intelligible word.
posted by alk at 1:27 PM on November 2, 2010
Anothing thing I was going to point out is that low sounds tend to travel farther, so that might explain why the sound tends to be low. The higher sounds less often make it to the microphones so the averaged out tone that you get tends to be lower than the "true" average would be if you could hear everything.
posted by If only I had a penguin... at 1:30 PM on November 2, 2010
posted by If only I had a penguin... at 1:30 PM on November 2, 2010
I'm just speculating here, but you'd probably get pretty close by shaping white or pink noise to two gaussians with the mean set at 130 and 210 for men and women respectively. Mess around with harmonic series and standard deviation and I bet you'd be in good shape.
posted by Dmenet at 1:32 PM on November 2, 2010
posted by Dmenet at 1:32 PM on November 2, 2010
The weirdest crowd noise I've ever heard was at the Million Mom March--while men (like me!) were welcome, they were far, far outnumbered. All the other large crowds I've been around have been mostly men, or at least a relatively even mix.
posted by MrMoonPie at 1:38 PM on November 2, 2010 [2 favorites]
posted by MrMoonPie at 1:38 PM on November 2, 2010 [2 favorites]
Remember on "Mork and Mindy" when Robin Williams would simulate a crowd sound by exhaling "haaaaaaaaaaaaaaaaaahhhhhhhhhhhh" and waving his hands in front of his mouth? It's surprisingly convincing, which should go some way toward explaining why actual crowds sound the way they do.
posted by kindall at 4:22 PM on November 2, 2010 [2 favorites]
posted by kindall at 4:22 PM on November 2, 2010 [2 favorites]
I don't have any actual data, but I know (from hearing PlayStation 2 football game sounds emanating from the other room...) that sound engineers have gotten really good at creating fake crowd noise and/or blending real crowd noise to the point where it's generic enough yet still real-sounding enough to work for multiple video-game situations. I seriously can't always tell anymore from the other room when it's the game and when it's actual football on TV.
posted by limeonaire at 8:21 PM on November 2, 2010
posted by limeonaire at 8:21 PM on November 2, 2010
I was at the University of Chicago Hospital when my dad was having surgery to remove prostate cancer. I was exhausted and I found the "white noise" of everyone in the waiting room talking was great for putting me to sleep. As long as I didn't concentrate enough to make out any single conversation, I was out like a light.
posted by IndigoRain at 10:48 PM on November 2, 2010
posted by IndigoRain at 10:48 PM on November 2, 2010
This thread is closed to new comments.
posted by AkzidenzGrotesk at 1:11 PM on November 2, 2010 [3 favorites]