Musical Coherence Limitations

I was just watching the Tabernacle Choir and realized they had some 360 people singing in unison adequate to preserve the quality of the words being sung. It occurred that there may be a limit to how many people can sing in unison and still be understood.

This seems to me to be similar to coherence in physics, such as in lasers, but at the level of the phonemes, rather than the carrier wave. But then, frequency coherence may also be important.

Reply to
Ricky
Loading thread data ...

søndag den 7. maj 2023 kl. 16.41.11 UTC+2 skrev Ricky:

here's close to 40000,

formatting link

Reply to
Lasse Langwadt Christensen

Probably helps that the Tabernacle is designed for great acoustics and its characteristics deeply analyzed:

formatting link

therefore they can get away with just ambient micing, the claim is they just have four Royer SF-2s on overhead duty for the whole choir:

formatting link

I'm guessing when they're on tour in some random space that doesn't work as well and they may have to close mic

"How many can be sung in unison and still understood" seems kind of an under-defined question...understood where? At what position in what space? Like if you're five feet away from a great singer you'll probably understand just fine regardless of how many backup singers there are. Are electronics allowed?

Once you start using mics and mixers and speakers/recording devices there's a lot you can do, you can just close mic your dozen best singers and have 500 more people that you mix proportionally lower in the mix like Britney Spears.

Reply to
bitrex

The phasing of the singers is not random - each is actively trying to match some kind of average of what's around them, so the model is a flock of PLLs trying to match the average of that flock.

In the time domain, unison requires that time sync error not exceed about twenty milliseconds.

For more data, dig into the computer music literature.

Joe Gwinn

Reply to
Joe Gwinn

I'll see your 60K and raise you 20K

formatting link

Reply to
John May

And sound propagates about 20 feet in that time. Round trip, 10 feet.

Reply to
John Larkin

I wonder how much they would improve with training and a leader? Choirs and bands are not unlike military organizations, where you obey orders unquestioningly (or at least quietly, lol).

Reply to
Ricky

Not sure what you mean by "phasing". If the best they can do is 20 ms, that's the wavelength of 50 Hz, so meaningless over the entire audible range. I assume you are referring to synching the other aspects of vocal music, other than phase?

Reply to
Ricky

I think any choir of appreciable size is going to have a conductor so each individual ideally isn't relying just on all their neighbors for timing.

Wrt phase:

"Mathematically, the phase synchronization is defined by the property that the phase difference is bounded within a finite range for an infinite duration of time [2]. Such strict definition, however, may not be applied to the choir singing, because chorus is inherently a transient process with occasional occurrence of switching among different notes. Even within a single note, the frequency ratio fluctuates to some extent. In a long span of time, the phase difference may eventually diverge to a large value. Hence, to apply the idea of phase synchronization to chorus, we need to relax the definition to adapt to a practical situation that the phase difference is bounded within a small enough range in a finite and stationary duration of musical note in a choir singing"

from:

<file:///C:/Users/Fujitsu%20Owner/Downloads/Synchronization_analysis_of_choir_singing.pdf>

The take away seems to be that with experienced choir singers phase difference between the singers tends to be bounded over some suitably short time interval.

Reply to
bitrex

I think whether some "idealized choir" of synchronized oscillators sounds musically pleasing, which is an issue of all the oscillators being largely on-time and on-pitch, is a somewhat different issue than preserving the "quality of the words being sung" which I think also requires preservation and alignment of the transients.

Reply to
bitrex

considering few of them are singers and many of them had more than a few pints, probably quite alot

Reply to
Lasse Langwadt Christensen

Nothing so mathematical as that. The twenty milliseconds is experimentally determined, and is the typical limit that if exceeded, unison is lost, which is quite audible.

Joe Gwinn

Reply to
Joe Gwinn

Sorry, I have no idea what you are talking about. You clearly have no idea what question I am asking. I give up.

Reply to
Ricky

Sounds the way scrambled eggs look.

Reply to
John S

Provided that they have the music in front of them and a conductor you can have very large choirs that sound fine from in front of them. Speed of light allows them to be well synchronised to each other.

Things fall apart when echoes from the walls or path differences become comparable with the length of a phoneme. Some big UK cathedrals are very difficult for readers because the first echo comes back at them strongly and with the right timing to trigger stuttering in anyone susceptible.

It is a general rule that apart from in stimulated emission an object cannot change brightness faster than the light travel time across it.

Reply to
Martin Brown

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.