1
$\begingroup$

I live in an arabic speaking country, and while I was sitting in a food court today, I noticed that the sound of the crowd was the same as the one I see in the stereotypical sound of the crowd. I have seen many threads discussing this on different forms, eg: Quora , reddit, Metafilter, and on this site, there is this tangentially related question. The problem with all of these is that, none of them is concrete, in sense of not having any calculative evidence.

Going down the reddit rabbithole and reading the metafilter thread gives me some idea on how a calculation should look like, but I don't feel so confident in my abilities to do it, could someone provide a fermi type calculation of this? Else, link to a source which does?

Improve this question
$\endgroup$
3
  • $\begingroup$ Are you claiming that all crowds sound the same, regardless of the language(s) being spoken? Presumably these crowds are of mixed ages and genders. $\endgroup$
    – PM 2Ring
    Commented May 20, 2023 at 0:26
  • $\begingroup$ It isn’t clear to me that this has anything to do with physics versus, say, human auditory perception. Why do you think that there is a physics-based answer? $\endgroup$
    – Ghoster
    Commented May 20, 2023 at 0:29
  • $\begingroup$ I beleive so, the reddit post suggests that it is also true regardless of place@PM2Ring $\endgroup$
    – Brian
    Commented May 20, 2023 at 5:56

1 Answer 1

Reset to default
2
$\begingroup$

For general background information: by the looks of it the following youtube speech acoustics playlist covers that background. I recommend absorbing that information first.

See also: comparison of color space and vowel space by Geoffrey Lindsay.

Quick summary:
In general any form of sound production produces not just a single frequency, but a series of harmonics. Example: if the fundamental frequency of a sound is 100 Hz then the harmonics are 200, 300, 400, 500, etc. Hz

The way human hearing processes sound is such that if the sound heard has that pattern of harmonics then our auditory recognition identifies the fundamental frequency.

(Moreover, if the lower frequencies are removed with a filter, then if that sound is played the perception is still the fundamental frequency. The brain processes the sound in such a way that the fundamental frequency is inferred from the pattern of harmonics.)

While it is the fundamental frequency of speech that we are consciously aware of, the harmonics of the sound of speech stack up quite high, and the brain is processing all of that information.

The information for vowel recognition is present in the range up to about 3000 Hz. To confirm: up to about three thousand Hz. That is quite a tall stack of harmonics.

While different languages use different vowel arrangement, the acoustic range of the vowel space is common to all languages.

(A language that would use a smaller vowel space would have to be spoken slower in order to still be intelligible, resulting in slower rate of information. Languages tend to all use full range of vowel space. Apparently there is an optimum in rate of information that can be conveyed, with languages all tending towards that optimum.)


The sound of a crowd of people, all talking individually, is a superposition of all of the speech, resulting in something comparable to white noise

Crowd-of-individually-speaking-voices is not quite white noise, human hearing can still suss out that it is speech sound from a crowd of people,

There are languages with very distinctive phonemes. For instance, languages with click consonants. Presence of click consonants will be recognizable.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.