Synchronized violin players reveal uniqueness of human networks | Hacker News Reader

Synchronized violin players reveal uniqueness of human networks

arstechnica.com

85 pointskawera16 comments

> Instead, Fridman et al. found that the players reacted by adjusting their playing, quickening or slowing their tempo to better synchronize with their fellow violinists

Not to take away from paper’s conclusion, but that is what professional orchestra musicians do. You are taught to “blend in”, regardless of whether the whole section is rushing or lagging.

A conductor can fix tempo issue, but can’t fix a un-synchronized section.

I’d like to see this experiment repeated on some soloists. (You do see soloists fighting the orchestra sometimes, to enforce their own interpretation, rather the conductors)

I tried a totally un-synchronized approach a while ago, basically relying on people's (total amateurs) inner clock and memory. This video shows what came out of it: https://rybakov.com/project/synchrony/

Do they have any reference (a metronome or clicktrack) or is it completely without reference? I think professional musicians would do better, but still be out of sync if they didn't have a tempo reference.

They all listened to a cover of the song a few times before singing – since the original John Lennon's 'Imagine' is way slower than one remembers. Still, the inability to be in sync without listening to each other, the inability to 'be as one' is kind of the point of this exercise.

that is a sign of a crap conductor. Soloist leads.

Constructive comment first: as a piece of music/art, I find this quite enjoyable to listen to!

> "In a state of frustration, they don't look for a 'middle,' but ignore one of the inputs. This is a critical phenomenon that is changing the dynamics of the network. Human networks are able to change their inner structure in order to reach a better solution than what's possible in existing models."

This may be true of trained musicians (who have to deal with this kind of thing in ensemble playing. I'm reminded of when I've volunteered in less experienced or youth orchestras), but I would be careful to extrapolate or expect similar results in other human systems.

My biggest gripe here is that I find it incredibly distracting when videos of string players do not sync the video to the sound they produced. You can easily notice it when a single violinist changes bow direction and the sound doesn't change. In a study and art piece about synchronization, syncing video to audio is pretty crucial!

> My biggest gripe here is that I find it incredibly distracting when videos of string players do not sync to the video to the sound they produced. You can easily notice it when a single violinist changes bow direction and the sound doesn't change. In a study and art piece about synchronization, syncing video to audio is pretty crucial!

Could this have to do with the structure of the experiment? Since they’re playing “together” (physically) but are hearing each other play on a delay I would imagine only one player in the video could sync with the audio?

> "Human networks behave differently than any other network we've ever measured," Fridman told The Jerusalem Post. "In a state of frustration, they don't look for a 'middle,' but ignore one of the inputs.

Huh. Maybe I'm reading too much into this, but it might imply a causal link between the overwhelming cacophony of modern media, and the breakdown of fact based discourse: faced with 'frustrated' (irreconcilable) viewpoints, it's the tendency of humans to simply ignore one of the inputs, and not seek the middle.

Perhaps. As a former orchestral musician I would say the behavior comes from the need for there to be a leader — if all that is happening is a cacophony of sound, no one has taken the reins as the tempo-keeper and everyone is either not paying attention or just confused. By joining with one of the other inputs, you create a majority or at least a plurality, and then the expectation is the other musicians will adjust to join the plurality. You might also increase volume or exaggerate movements to get everyone on board.

Hm. What would seeking middle be with QAnon or PizzaGate folk?

Steve Reich was there all along: https://en.wikipedia.org/wiki/Violin_Phase

(And a wonderful interpretation on Cello: https://www.youtube.com/watch?v=x7aj796c3A0)

And you don’t even need to build a network! Same results can be achieved by using multi threading on a single instance..

https://youtu.be/AnQdP03iYIo

If you've ever seen music for 18 musicians live, for me, it was an amazing experience. I think some of those techniques are used. I started hallucinating about 30 minutes into it.

Really cool work! They composed a simple melodic fragment and had numerous violinists play it with electric violins. They then had them use noise-cancelling headphones and selectively fed them only a couple collabrators' violin playing. They then blocked them from sight of the others, introduced delays, switched whose playing they could hear.

The implications here are pretty fascinating. Do humans conform to the average of others' playing, or do they pick one out (it's probably the latter)? Do they attempt to enforce "correctness" onto their own playing, or do they slow down/speed up to prioritize "unison" with others?

I know from playing in musical ensembles for years, each musician has their own approach. Some are confident and self-righteous, they really "sing out" and put their intonation and timing into the fore. Other musicians are blenders, they do very well solo but their main goal is to blend in perfectly with their neighbors. Then there are shy people, who try to hide in the background and blend in, and they don't want their voice heard.

The actual technical aspects of music are very precise, and yet we have this layer of artistry which is adapting to delays, improperly tuned things, or even improvisatory, stylistic flourishes. I love it! It's an artistic math, it's a mathematical manipulation of artistic objects.

A lot of human networks are more connected than square and triangle lattices. I wonder whether there is a way of taking a number of connected nodes, somehow extracting and looking at the interaction relevant to a sub-lattice, and generalising the result to the whole network. Am I being optimistic?

I wonder how the multiple graph ‘levels’ (which could interact with each other) would come into play. EXCHANGES, BROKERS, HEDGE_FUNDS, INTERNET_PIPES, PEOPLE.

What existing tools are there for analysing information propagation like in the paper?

Should it be analysed through the lens of uncooperative game theory of non-zero sum game. Or forming the core ...

Reply on news.ycombinator.com