Expensive old violins sound better

Get your science fix here: research, quackery, activism and all the rest
Post Reply
User avatar
nekomatic
Dorkwood
Posts: 1376
Joined: Mon Nov 11, 2019 3:04 pm

Expensive old violins sound better

Post by nekomatic » Tue Nov 01, 2022 9:11 pm

Apparently

proper article summary
In this paper, the CTs [combination tones] generated by a set of violins of different quality and age have been investigated when playing a selected set of dyads. CTs were found in all of the violins, and the strongest of them occurred at a frequency below the lower note of the dyad. Its amplitude was strongly dependent on violin and dyad played and was greatest in two old Italian violins and decreased down to a minimum in a factory-made violin. All of these findings are well explained by the boosting action of A0, the main air resonance of the violin that correlates well with the strongest CT. A listening test, performed using selected dyads and violins, showed that the differences between dyads with and without CTs were correctly recognized by a group of professional and amateur musicians, suggesting a possible musical significance of the main CT.
Move-a… side, and let the mango through… let the mango through

IvanV
Stummy Beige
Posts: 2663
Joined: Mon May 17, 2021 11:12 am

Re: Expensive old violins sound better

Post by IvanV » Wed Nov 02, 2022 12:01 pm

nekomatic wrote:
Tue Nov 01, 2022 9:11 pm
Apparently

proper article summary
In this paper, the CTs [combination tones] generated by a set of violins of different quality and age have been investigated when playing a selected set of dyads. CTs were found in all of the violins, and the strongest of them occurred at a frequency below the lower note of the dyad. Its amplitude was strongly dependent on violin and dyad played and was greatest in two old Italian violins and decreased down to a minimum in a factory-made violin. All of these findings are well explained by the boosting action of A0, the main air resonance of the violin that correlates well with the strongest CT. A listening test, performed using selected dyads and violins, showed that the differences between dyads with and without CTs were correctly recognized by a group of professional and amateur musicians, suggesting a possible musical significance of the main CT.
What that article shows is that they sound different, in one specific way. There was no attempt to assess which was "better".

It was already well established than they sound different. This article adds some precision to precisely what is different about their sound, in one particular, narrow, and arguably relatively unimportant, aspect.

I say narrow and unimportant, because they were testing the sound of dyads, ie, playing two notes at once. Most of the time, you play just one note at once on a violin. So this piece of research showed that old violins sound different when being played in a way that you don't do most of the time.

Other research indicates that people prefer the sound of modern violins. There is also evidence that they are easier to play well. So modern violin makers are doing a good job. But not very well recognised for it, it would appear.

The leading classical violinist of the present day, Joshua Bell, owns a famous Stradivarius. But what that article fails to mention, and even his own website, is that he owns other violins, including modern ones. I have heard interviews with him saying that he doesn't only perform on his Strad, he chooses the most suitable violin for the repertoire. So from time to time, you might have heard him play a modern violin. But it is rarely mentioned, perhaps because these old violins are part of the mystique.

The practical reality of being a professional classical string player of any reputation is that you aren't taken seriously unless you own an old instrument. And paying half the price of a house - a typical violin for a many 2nd division professionals as found on lower desks of leading orchestras, etc - for an instrument in a badly-paid job is challenging. Unfortunately, it isn't rational.

hakwright
Buzzberry
Posts: 47
Joined: Mon Jan 20, 2020 12:58 pm

Re: Expensive old violins sound better

Post by hakwright » Wed Nov 02, 2022 12:40 pm

It's also worth saying that the violin player may have been fully aware of which instrument was which. It's possible to take steps to "blind" the player, to try and compensate for the fact that players will approach their playing differently if they think they're playing an old Italian violin, versus a cheap factory made violin. But this is not easy, since multiple senses are likely involved (sight, touch, smell).

So for this particular paper (I can't access the full text, so don't know the answer), it could be that the player recognised the "good" instruments, and their expectations led to a different style of playing, which could have affected the sounds produced and combination tones that resulted.

Howard

User avatar
nekomatic
Dorkwood
Posts: 1376
Joined: Mon Nov 11, 2019 3:04 pm

Re: Expensive old violins sound better

Post by nekomatic » Wed Nov 02, 2022 2:05 pm

I am guilty of following the Guardian’s lead in posting this under a clickbait headline, for which apologies. Clearly ‘better’ of a violin’s sound is not an objective measure, but there is agreement that violins sound different. I don’t think the sample size was big enough for them to disentangle age from ‘quality’ (or its proxy, expense) so I don’t think the claims exclude the possibility that good modern violins are as good as Strads.

What I understand to be novel based on the article’s abstract (and they may of course be overselling this given they cite references that address a similar question) is that they’ve measured a specific physical phenomenon for the first time that appears to correlate with some dimension of perceived quality. I’d certainly agree that subjectivity of the performer could still be at work but at least it seems that what they were doing was playing specific intervals, presumably in the best tuning they could achieve on each instrument, rather than giving a musical performance that was then evaluated by subjective criteria.

Without having read the conclusions and discussion one idea that occurred to me is whether you could filter these combination tones out of a recording of a good instrument and thereby make it sound indistinguishable from a bad one.
Move-a… side, and let the mango through… let the mango through

IvanV
Stummy Beige
Posts: 2663
Joined: Mon May 17, 2021 11:12 am

Re: Expensive old violins sound better

Post by IvanV » Wed Nov 02, 2022 2:36 pm

nekomatic wrote:
Wed Nov 02, 2022 2:05 pm
Clearly ‘better’ of a violin’s sound is not an objective measure, but there is agreement that violins sound different.
A sound frequency analyser shows that sounds are different without needing to ask people. But we learn from the human experiment that the difference is large enough that amateurs (of some experience) can consistently hear the difference, as well as professionals with highly trained ears.

We can also carry out experiments on what people prefer. For example, there is very high correlation in what kinds of human appearance are preferred by heterosexuals of the opposite sex if you ask them. So we can argue - psychologists often do - that you can objectively measure how good-looking someone is, at least in a heterosexual context, and within a specific culture. Psychologists take advantage of this frequently. They use it in testing questions like, are good-looking people more likely to be offered a job. (Not necessarily, apparently.) Because you can get a sample of people to rate people's appearance, and be confident that if you asked another sample they'd give very similar answers.

In the case of listening to a violin, we can also ask people what they prefer. If there is high enough correlation in what people say, then we could say it is an objective difference. So, if that was true, we could say one sound is objectively "better". That is the kind of experiment that has been carried out in other cases, and with suitable levels of blinding of both performer and listener as to what they are listening to or playing. But there are likely to be complications. Different sounds might be more suited for different kinds of music. People of different levels of interest in, expertise in, experience of, music might have systematically different preferences.

Btw, in the experiment reported, these combination tones occurred when playing dyads - chords of two notes, played simultaneously on the same instrument. When you play a tune on the violin, you mostly just play one note at a time. Does this experiment have any relevance at all to the sound in that normal situation?

User avatar
Gfamily
Light of Blast
Posts: 5180
Joined: Mon Nov 11, 2019 1:00 pm
Location: NW England

Re: Expensive old violins sound better

Post by Gfamily » Wed Nov 02, 2022 3:34 pm

IvanV wrote:
Wed Nov 02, 2022 2:36 pm

Btw, in the experiment reported, these combination tones occurred when playing dyads - chords of two notes, played simultaneously on the same instrument. When you play a tune on the violin, you mostly just play one note at a time. Does this experiment have any relevance at all to the sound in that normal situation?
Maybe the quality of a violin shows up more clearly when playing dyads and triads.
Here's an excerpt from the Brahms Violin Concerto in D major where the notes are dyads and triads.
Screenshot_20221102-152756.jpg
Screenshot_20221102-152756.jpg (561.82 KiB) Viewed 2838 times
ETA correct the Concerto
My avatar was a scientific result that was later found to be 'mistaken' - I rarely claim to be 100% correct
ETA 5/8/20: I've been advised that the result was correct, it was the initial interpretation that needed to be withdrawn
Meta? I'd say so!

User avatar
nekomatic
Dorkwood
Posts: 1376
Joined: Mon Nov 11, 2019 3:04 pm

Re: Expensive old violins sound better

Post by nekomatic » Wed Nov 02, 2022 3:46 pm

IvanV wrote:
Wed Nov 02, 2022 2:36 pm
A sound frequency analyser shows that sounds are different without needing to ask people.
Sure, but auditory perception is such that some easily measurable differences are inaudible while some much smaller ones are obvious, some distortions that objectively impair a signal make it sound better, and so on.
Btw, in the experiment reported, these combination tones occurred when playing dyads - chords of two notes, played simultaneously on the same instrument. When you play a tune on the violin, you mostly just play one note at a time. Does this experiment have any relevance at all to the sound in that normal situation?
I am the only person in my household who doesn’t play the violin, so I have a fair idea how violining works. I am guessing that playing the dyads is found to provoke the combination tones at clearly measurable levels, but that doesn’t mean they are irrelevant to other scenarios. There will certainly be times while playing a monophonic line that two notes sound at the same time, and those articulations might be particularly important to the perception of ‘quality’, for example.
Move-a… side, and let the mango through… let the mango through

User avatar
nekomatic
Dorkwood
Posts: 1376
Joined: Mon Nov 11, 2019 3:04 pm

Re: Expensive old violins sound better

Post by nekomatic » Wed Nov 02, 2022 3:56 pm

Anyway, if you were going to set up a comparison of violins based on subjective evaluation of performances by a skilled violinist what would you get them to play? Probably something like this.
Move-a… side, and let the mango through… let the mango through

IvanV
Stummy Beige
Posts: 2663
Joined: Mon May 17, 2021 11:12 am

Re: Expensive old violins sound better

Post by IvanV » Wed Nov 02, 2022 4:45 pm

nekomatic wrote:
Wed Nov 02, 2022 3:56 pm
Anyway, if you were going to set up a comparison of violins based on subjective evaluation of performances by a skilled violinist what would you get them to play? Probably something like this.
And here we get back to that issue that maybe the most suitable violin may vary according to the repertoire, as apparently Joshua Bell thought when he gave an interview I listened to.

On the one hand, the kind of solo situations that occur in the solo line in a 19th century concerto, and in Bach partitas, are rather different from each other, and from what most violins are asked to do most of the time.

But then again, those virtuoso solo situations are probably when it matters most just what the violin sounds like. If you are Joshua Bell, that's the kind of thing you play most of the time.

User avatar
bolo
Dorkwood
Posts: 1016
Joined: Mon Nov 11, 2019 1:17 pm
Location: Washington DC

Re: Expensive old violins sound better

Post by bolo » Wed Nov 02, 2022 5:07 pm

Joshua Bell once played his Strad in a Washington Metro station, to see how many people would stop to listen:

https://www.washingtonpost.com/lifestyl ... story.html

User avatar
Grumble
Light of Blast
Posts: 4746
Joined: Mon Nov 11, 2019 1:03 pm

Re: Expensive old violins sound better

Post by Grumble » Wed Nov 02, 2022 6:31 pm

bolo wrote:
Wed Nov 02, 2022 5:07 pm
Joshua Bell once played his Strad in a Washington Metro station, to see how many people would stop to listen:

https://www.washingtonpost.com/lifestyl ... story.html
I’m willing to bet if you limited the audience to those who walked past him that day and offered them free tickets to a performance by him, very few would know who he was and not many would turn up.
where once I used to scintillate
now I sin till ten past three

User avatar
basementer
Dorkwood
Posts: 1504
Joined: Mon Nov 11, 2019 1:03 pm
Location: 8024, Aotearoa
Contact:

Re: Expensive old violins sound better

Post by basementer » Wed Nov 02, 2022 11:44 pm

IvanV wrote:
Wed Nov 02, 2022 2:36 pm
Btw, in the experiment reported, these combination tones occurred when playing dyads - chords of two notes, played simultaneously on the same instrument. When you play a tune on the violin, you mostly just play one note at a time. Does this experiment have any relevance at all to the sound in that normal situation?
It's by no means an advanced technique, one of the test pieces currently on Trinity's grade 1 violin syllabus uses it. It might be more common than you've noticed.
Money is just a substitute for luck anyway. - Tom Siddell

User avatar
EACLucifer
Stummy Beige
Posts: 4177
Joined: Fri Dec 13, 2019 7:49 am
Location: In Sumerian Haze

Re: Expensive old violins sound better

Post by EACLucifer » Thu Nov 03, 2022 10:56 am

IvanV wrote:
Wed Nov 02, 2022 12:01 pm
What that article shows is that they sound different, in one specific way. There was no attempt to assess which was "better".
This is a subject that could easily run to an essay, probably rather more than that.

However, I'll try to give a precis of my thoughts on this as a former builder of - albeit very different - string instruments.

In this case, better is going to be defined roughly as "sounds like a Strad", because that's what violin builders are usually aspiring to. While that may sound foolish and self-limiting - and in some ways is - what isn't foolish or self limiting is building instruments that are meant to sound as they did when the repertoire that is to be played was composed. If we want to play music that sounds as the composer intended it to sound, then the instruments really have to sound that way. Now there are some very notable instances where that is not done - an example would be the replacement of the ophicleide and serpent in the last movement of Sinfonie Fantastique with a pair of tubas, which sound good, but very different to what Berlioz had in mind when he wrote it.

I once discussed with a talented violinist the idea of making a "Dark Violin", with rosewood replacing ebony for the soundboard and mahogany or even limba replacing maple for the neck, back and ribs, and possibly a cedar top. That would produce an instrument with a very different sound, likely quite dark and warm and mournful. However, we expect a violin to sound bright and cutting, so it would be unlikely to see anything other than niche usage.

IvanV
Stummy Beige
Posts: 2663
Joined: Mon May 17, 2021 11:12 am

Re: Expensive old violins sound better

Post by IvanV » Thu Nov 03, 2022 12:56 pm

EACLucifer wrote:
Thu Nov 03, 2022 10:56 am
In this case, better is going to be defined roughly as "sounds like a Strad", because that's what violin builders are usually aspiring to. While that may sound foolish and self-limiting - and in some ways is - what isn't foolish or self limiting is building instruments that are meant to sound as they did when the repertoire that is to be played was composed. If we want to play music that sounds as the composer intended it to sound, then the instruments really have to sound that way. Now there are some very notable instances where that is not done - an example would be the replacement of the ophicleide and serpent in the last movement of Sinfonie Fantastique with a pair of tubas, which sound good, but very different to what Berlioz had in mind when he wrote it.
I can understand that is what a violin maker might aspire to. But ultimately their aim is to satisfy their customers. And the experiment demonstrates that their customers might have a different preference.

In the experiment I referred to, people were simply asked, what do you think sounds better? The subjects included a mix of amateur and professional violinists. So they were informed listeners, not the general public. They were presented both with several pairwise comparisons to select from, and also a list experiment where they chose their favourite from a list. This is good experimental design, because people are sometimes inconsistent between pairwise comparisons and list choices. For example, preference experiments from a list can be affected by the order of presentation in the list. In the pairwise comparisons, each pair included a Strad, but the listeners didn't know that. So the final conclusion that they didn't prefer the Strad was quite strong: the Strad was rejected in every pairwise comparison with a good modern violin.

I can understand that we might be conditioned to accept a Strad-like sound as what we expect "good" to sound like for a violin. But seemingly most informed listeners in that experiment routinely rejected the Strad. So we might be able to conclude that "just like a Strad" was not their preference at all. And the modern violin makers might be affected by that, when trying to satisfy their customers.

Now there might be some difficulties with the experiment. Bell, for example, says that it takes a good month for him to get used to playing an old violin that he hasn't played before, and to be able to do with it what it is capable of doing. Whereas modern violins are much easier to pick up and play close to their best ability. So maybe in the experiment, the player was not able to bring the best out of the Strad they were given to play. And that might be reason for its consistent rejection.

User avatar
EACLucifer
Stummy Beige
Posts: 4177
Joined: Fri Dec 13, 2019 7:49 am
Location: In Sumerian Haze

Re: Expensive old violins sound better

Post by EACLucifer » Thu Nov 03, 2022 8:27 pm

IvanV wrote:
Thu Nov 03, 2022 12:56 pm
EACLucifer wrote:
Thu Nov 03, 2022 10:56 am
In this case, better is going to be defined roughly as "sounds like a Strad", because that's what violin builders are usually aspiring to. While that may sound foolish and self-limiting - and in some ways is - what isn't foolish or self limiting is building instruments that are meant to sound as they did when the repertoire that is to be played was composed. If we want to play music that sounds as the composer intended it to sound, then the instruments really have to sound that way. Now there are some very notable instances where that is not done - an example would be the replacement of the ophicleide and serpent in the last movement of Sinfonie Fantastique with a pair of tubas, which sound good, but very different to what Berlioz had in mind when he wrote it.
I can understand that is what a violin maker might aspire to. But ultimately their aim is to satisfy their customers. And the experiment demonstrates that their customers might have a different preference.

In the experiment I referred to, people were simply asked, what do you think sounds better? The subjects included a mix of amateur and professional violinists. So they were informed listeners, not the general public. They were presented both with several pairwise comparisons to select from, and also a list experiment where they chose their favourite from a list. This is good experimental design, because people are sometimes inconsistent between pairwise comparisons and list choices. For example, preference experiments from a list can be affected by the order of presentation in the list. In the pairwise comparisons, each pair included a Strad, but the listeners didn't know that. So the final conclusion that they didn't prefer the Strad was quite strong: the Strad was rejected in every pairwise comparison with a good modern violin.

I can understand that we might be conditioned to accept a Strad-like sound as what we expect "good" to sound like for a violin. But seemingly most informed listeners in that experiment routinely rejected the Strad. So we might be able to conclude that "just like a Strad" was not their preference at all. And the modern violin makers might be affected by that, when trying to satisfy their customers.

Now there might be some difficulties with the experiment. Bell, for example, says that it takes a good month for him to get used to playing an old violin that he hasn't played before, and to be able to do with it what it is capable of doing. Whereas modern violins are much easier to pick up and play close to their best ability. So maybe in the experiment, the player was not able to bring the best out of the Strad they were given to play. And that might be reason for its consistent rejection.
I really wish I was sufficiently compos mentis to dig into this properly right now, but sadly I'm not. It's a huge subject, touching on our perception of quality and sound, and how we express composers intentions. Both are significant subjects, so certainly, there's material for an essay or more on this, I'm just not up to it right now. These are all good points, but we're still just scratching the surface here in philosophy of music terms.

User avatar
Sciolus
Dorkwood
Posts: 1313
Joined: Mon Nov 11, 2019 6:42 pm

Re: Expensive old violins sound better

Post by Sciolus » Thu Nov 03, 2022 8:30 pm

IvanV wrote:
Thu Nov 03, 2022 12:56 pm
I can understand that we might be conditioned to accept a Strad-like sound as what we expect "good" to sound like for a violin. But seemingly most informed listeners in that experiment routinely rejected the Strad. So we might be able to conclude that "just like a Strad" was not their preference at all. And the modern violin makers might be affected by that, when trying to satisfy their customers.
Conjecture: That might be down to familiarity. There are far more good modern violins than there are Strads, and while Strads may get more exposure (being played by the biggest performers -- though not always), I'm sure they get fewer earholes overall. So most people are used to the sound of a modern violin and have learned to like it.

User avatar
bjn
Stummy Beige
Posts: 2915
Joined: Wed Sep 25, 2019 4:58 pm
Location: London

Re: Expensive old violins sound better

Post by bjn » Sat Nov 05, 2022 10:53 pm

IvanV wrote:
Wed Nov 02, 2022 12:01 pm
The practical reality of being a professional classical string player of any reputation is that you aren't taken seriously unless you own an old instrument. And paying half the price of a house - a typical violin for a many 2nd division professionals as found on lower desks of leading orchestras, etc - for an instrument in a badly-paid job is challenging. Unfortunately, it isn't rational.
I have a professional violinist friend who formed a limited company to purchase an old violin, though not a Stradivarius. She had a bunch of friends and family to chuck in cash, in total quite a hefty six figures. She gets to play it professionally, they get to own an asset that is beating the stock market. Though a somewhat illiquid asset.

Post Reply