Is The Reputed Hifg Quality of Stradivari Viulins a Myth?

[Shalmanese]

@Aldebaran

1. If you mean “people” chosen at random with no knowledge of music, about playing a violin etc… I would say this type of test is worthless before it starts.

If your original claim was “Why, the quality of a strad is so obvious, anyone can instantly hear one and know why they are so revered”, then yes, you should be grabbing people off the street. If it was “Yes, we admit that modern violins are very good, but, to a trained ear, the strad will always be a step above”, then yes, you should be choosing people with such a trained ear. I still don’t think you’ve made a distinction either way at present.

2. If you mean “people” who play the violin themselves, are familiar with the piece that is performed and play the work themselves, you have no lesser problem.

Suppose I’m in that jury.

1.The piece is baroque music. I find it horrible to listen to baroque music where the instruments are pitched wrongly. Yet people not familiar with the baroque pitch can find it horrible to listen to it and not everyone who plays baroque agrees with my preference.
Suppose the violinist belongs to that other group. Such a situation alone would already cloud my ability to focus on the behaviour of the instrument.

We choose some hugely popular and presumably unoffensive piece that also allows us to show of the range and subtely of the instruments. Being someone that plays no musical instrument whatsoever, take my suggestion completely as uninformed but I would submit that “Four Seasons” by Vivaldi would be a good piece.

2. In addition, the violinist plays the work in a way I can not identify with → my focus shall switch completely from the violin to this interpretation, out on finding “faults” and exaggerating every tone I dislike about this performance.
   It can be a marvel of an instrument I hear, the pitch of the instrument I feel as wrong and the player and his interpretation I feel as wrong pushes the instrument itself to the background and I exaggerate every tone I dislike, thereby also exaggerating what I can ever dislike about the quality of the instrument.
3. The pitching is correct in my view, but point 2 appears to be the case. Same situation occurs during my listening to the performance.
4. Pitching OK or no baroque music and no problems with the interpretation. Nevertheless I try (and you can’t prevent this) to find differences between the interpretation I hear and mine. This has (and you can’t prevent this to occur either) its influence on my ability to only listen to the instrument.

The violinist will be playing the same piece twice over, if you hate both pieces, then you should be prejudiced against the violins equally.

5. No baroque or good pitch and in addition the interpretation opens for me a new world of insight I never thought of before. Again my focus is not exclusively on the performance of the violin. The quality/beauty of the instrument is part of that whole concept I now enjoy greatly. I risk to exaggerate the quality of the instrument because of that.

In all these situations:
When you next hear the same piece played by the same violinist but with the other instrument, all your prejudices - good or bad - can only become more focussed on.
Your final conclusion is an overall impression of the whole performance, your judgement of the second one influenced by what you heard the first time. In addition no violinist is able to give exactly the same performance twice.

Wat is left to be counted as a judgement that is in no way the result of my (often unconcious) subjective appreciation?

Firstly, your blindfolded so you can’t see the beauty of the instrument. But, also, your arguing against a false positive result in which there is a strong preference shown but not because of the inherent qualities of the violin. A single performance isn’t enough to establish if the strad is superior, only to establish that both are equal if thats the way the preferences run. If it were the case that the audience strongly preferred one instrument, then further tests would be warranted with different performers and different orders and different songs. If it turned out that in every single case the strad was chosen, then a case could be made that the strad is indeed superior. If, however, in every performance, there was always a strong preference but it was split evenly between the strad and the modern, then it would indicate that factors other than the violin are what are really important to the performance and our test procedure isn’t senstive enough to determine which is superior. But it also means that your just as likely to hear a bad performance on a strad as a good performance on a modern so theres no point trying to find out anyway.

Suppose you set up the procedure with a newly composed piece of music = nobody in the audience ever saw or heard it before.
The first time you hear it, you can detect already its qualities and/or flaws. You are expected to bring what you hear in direct relation with the instrument played, yet it is impossible to do that because automatically your focus is on discovering the music. You can like or dislike it, like a piece of it or like the way the violinst give it an interpretation or dislike that. Further you start to muse about how you would approach it. All of this distracts you from only listening to the instrument.
The next time you hear this work you recognize what you hear and you are able to detect new details you didn’t noticed the first time. This does not prevent that you focus on the interpretation also and on the thought how you would do it = you are distracted from the sound of the violin as you were when you heard the work for the first time.
Nevertheless your overall ability to also listen to the role of the instrument in building the sound of what you hear is much better the second time you hear it.
No way you can judge both instruments equally in such a case.

I wouldn’t conduct the test like this since it introduces another unknown variable.

Nobody in this thread has made the claim that no modern violin can be build (or is build) that is able to compete with the old ones. Yet I don’t see any evidence in the preferences of the most famous violinists of today.

Salaam. A

It may be quite hard for non-sceptics to grasp but there is a clear and consistent trend in human behaviour to genuinely believe something to be the case just because everyone else says it is. Especially for something like music, just by knowing your playing a strad can make the music sound better even though the sound waves coming out of the thing are exactly the same.

PS: A slightly simpler version of the single blind that avoids some of the problems proposed would be to arrange it so that 1/2 the audience to a concert are told that the violin being played is a strad while 1/2 are not told anything. Then, have them rate the performance and see how big a difference there was. While it doesn’t show how a strad might compare to a modern, at least it would demonstrate the power of pre-conception on our listening ability.

[Aldebaran]

@Shalmanese

I still don’t think you’ve made a distinction either way at present.

What distinction?
I think it is very obvious that people who don’t like to hear a violin to begin with, or have no clue if they will like the sound of a violin when entering such a test group, are excluded. The same for people with no experience for any sort of music or people who have only experience in listening to it.
You don’t become an expert on a violin and its possibilities if you don’t reach a certain level in playing the instrument yourself. Same counts for the work that is performed. You must be completely familiar with it to be able to judge how its reproduction sounds.

The violinist will be playing the same piece twice over, if you hate both pieces, then you should be prejudiced against the violins equally.

I think you must re-read my post on this… I has nothing to do with “not liking the work”. If I know I am going to be part of such a testgroup, I shall study the work. When you study a work, you become familiar with it, you make it your own and even if it is not completely your taste, you grow to like it.
I talked about the way the artist plays it and how that alone influences your judgement and how this judgement is even stronger - in all aspects- when you hear it a second time. All factors that influence your ability to listen to the violin independently from all these interfering factors. There is simply no way to do that.

Firstly, your blindfolded so you can’t see the beauty of the instrument.

When we talk about “beauty”, we talk about the quality of the tone. They are not showpieces although some can be very “beautiful” to look at.

But, also, your arguing against a false positive result in which there is a strong preference shown but not because of the inherent qualities of the violin.

I merely explain what happens when a violinist listens to a work he plays himself and that is performed by someone else. You can’t escape in no way what I describe (and I don’t even describe all of it).

A single performance isn’t enough to establish if the strad is superior, only to establish that both are equal if thats the way the preferences run.

With every work you hear the same process I describe starts all over, and with everyone in the audience on a personal level. Hence you can not rely on any of their individual comments as if there was no subjectiveness involved. Listening to music, any music, is one big amount of subjective processes leading to a global judgement of what you hear.

It may be quite hard for non-sceptics to grasp but there is a clear and consistent trend in human behaviour to genuinely believe something to be the case just because everyone else says it is.

Maybe, but I have experience with such violins.

Especially for something like music, just by knowing your playing a strad can make the music sound better even though the sound waves coming out of the thing are exactly the same.

Most certainly not. Such comment can only come from someone with no experience on a violin. Knowing that I play on a Guarnerius or an other Cremona instrument does/did nothing else to me then loving their indivdual sounds and possibilities. Not because of their labels. I already said they could have the label of a swinestable for my sake. I play on them because of what they are: superb instruments. I also turn to the same type of 4/4 instrument when I don’t take the Guarnerius (a 3/4, superb of quality or not, is not suitable for everything you want to do when you have reached a certain level of technique, not to speak of stature. SentientMeat maybe would argue that he is because of the size of his “plates” :)… ).

You can argue that all this has to do with me being a blank page when it comes to modern violins. It can very well be, but that factor does not explain why I didn’t like the Stradivarius I heard last year. Certainly not when comparing him to the Cremona instruments I play(ed) myself on for several years.

I also does not explain why my relative, who has the Amati copy now and had previously already a very good instrument, didn’t have to think twice about it. Same with the other violin who is with someone who previously also had a very good instrument.
Do you think professionals in general change from very good to mediocre or even to “as good as they already have” just because someone provides them with a violin that has some “known” label in it?

A slightly simpler version of the single blind that avoids some of the problems proposed would be to arrange it so that 1/2 the audience to a concert are told that the violin being played is a strad while 1/2 are not told anything. Then, have them rate the performance and see how big a difference there was. While it doesn’t show how a strad might compare to a modern, at least it would demonstrate the power of pre-conception on our listening ability.

Not a good idea. I for one listen much more sceptical and critical when I know I listen to a stradivarius. In such a test this would most certainly influence my judgement and probably in a negative way.
Salaam. A

[SentientMeat]

Tusc: The problem is that SentientMeat states that there is no difference because he didn’t measure any difference.Us violinists on the other hand know from personal experience that there is a perceivable difference, hence SentientMeat must have failed to measure every factor.

I said there was no measurable difference in a plate after excitation compared to before it, ie. that regularly playing instruments does not affect the instrument in any way. I said categorially that there are differences in playability and tone between different instruments, and that discerning listeners could tell which intrument played which piece in when blindfold. Again, I’m not sure where we disagree: clearly it is possible to blindfold listeners or players and ask them questions about the piece of wood which amplifying the string vibrations. I’ll dig around for citations if necessary, but given an unfamiliar Stradivarius and an unfamiliar modern copy, their ability to spot the Strad was little (if at all) better than chance.

GM and Alde, if I blindfolded you, I would imagine you could still recognise your own instrument by playing it or perhaps even by listening to the tone when played by someone else - I think I could do that with my guitar. However, if someone handed you two unfamiliar violins and asked you to identify the Strad, could you? I don’t think I could tell a top-quality old Ramirez from a top-quality modern guitar, but I could explain why they were both good quality in terms of structural properties (most importantly the stiffness:mass ratio of the top plate).

[GorillaMan]

@SentientMeat

GM and Alde, if I blindfolded you, I would imagine you could still recognise your own instrument by playing it or perhaps even by listening to the tone when played by someone else - I think I could do that with my guitar. However, if someone handed you two unfamiliar violins and asked you to identify the Strad, could you?

I could identify my own violin purely by the smell of the wood …

But having never played a Strad, no, I cannot say with certainty that I would. But I’m confident that I could tell the different with a modern one. Less so if I had to tell the difference between a Strad and a Guaneri.

Suppose we were to compare the oscilloscope traces of a Strad playing each string, agaiants a nmodern violin…would the harmonics of each provide a key to the differences? A computer could analyze the differences, and report on a scientific basis. No need for human ears!

Unfortunately, the sounds produced by violins are astonishingly complex - the variance between the same note played differently on the same instrument would be enough of a challenge. And every different instrument will give a very different reading. Identifying which differences are the important ones would be a very difficult task. (And would involve human ears after all.)

On a side note: have these old violins been re-glued? I would imagine that wood pieces under continuous tension would eventually break the glued joints. I will have to ask a guy up the street 9he repairs violins and other string instruments). Can we assume that the old Strads (and Amatis, Guarnerius, etc.) have been rebuilt, possibly several times?

Yep, many have been - unless the job is botched, dissassembling any violin doesn’t affect its qualities. Apart from anything else, it’s only when they’re in pieces that we’ve been able to take accurate enough measurements to make physical copies. Which often don’t sound nearly as good as the original.

[SentientMeat]

Identifying which differences are the important ones would be a very difficult task.

A sound is just its component frequencies, and there is a minimum difference in amplitude at a given frequency which a human ear can detect that Colin Gough article explains this and gives good references. Like I sais, the acoustical impedance curves and sound pressure curves are the instrument.

[GorillaMan]

@SentientMeat

A sound is just its component frequencies, and there is a minimum difference in amplitude at a given frequency which a human ear can detect that Colin Gough article explains this and gives good references. Like I sais, the acoustical impedance curves and sound pressure curves are the instrument.

Are you saying that, using current technology, it is possible to identify individual instruments by their sounds? And does that bring us any closer to identifying the particular characteristics that makes us prefer one to another (which seemed to be the logic behind the suggestion I was responding to)?

And as you’re so keen on that article, perhaps I should quote a paragraph for you:

In practice it is extremely difficult to distinguish between a particularly fine Stradivarius instrument and an indifferent modern copy on the basis of the measured response alone. The ear is a supreme detection device and the brain is a far more sophisticated analyser of complex sounds than any system yet developed to assess musical quality.

[SentientMeat]

Are you saying that, using current technology, it is possible to identify individual instruments by their sounds?

Yes - secure voice recognition software already does so in our everyday lives.

And does that bring us any closer to identifying the particular characteristics that makes us prefer one to another (which seemed to be the logic behind the suggestion I was responding to)?

Yes - the high stiffnes to mass ratio of the spruce plate which provides a smaller impedance mismatch at the bridge, allowing more efficient transfer of string energy to actual sound propagation (particularly at higher frequencies, when the forcing frequency is way above that of the low order plate resonances).

And of course I agree with Colin there. Again, I’m struggling to discern where you and I actually disagree. I am a musician and a vibraional physicist: I’m not sure what it is you’re trying to tell me which you think I either don’t know or don’t appreciate strongly enough.

[GorillaMan]

@SentientMeat

Yes - secure voice recognition software already does so in our everyday lives.

That’s a massive assumption - or has this software been successfully applied to identifying instruments?

And of course I agree with Colin there. Again, I’m struggling to discern where you and I actually disagree. I am a musician and a vibraional physicist: I’m not sure what it is you’re trying to tell me which you think I either don’t know or don’t appreciate strongly enough.

I think we’re disagreeing on three different points. First, whether our current knowledge of the physics of violins is adequate to answer the questions being posed. Secondly, whether any objective characteristics can exist that identify a ‘better’ violin. Thirdly, whether it is possible to ever completely separate myth from reality in order to compare different instruments.

[SentientMeat]

That’s a massive assumption - or has this software been successfully applied to identifying instruments?

Well, I tried one piece of software myself on guitars - it seemed work just as well as on human voice recognition (ie. it could recognise one guitar among many).

First, whether our current knowledge of the physics of violins is adequate to answer the questions being posed.

There are a lot of questions in this thread, most of which I’m trying to answer as factually as I can. Which specific ones were you thinking of?

Secondly, whether any objective characteristics can exist that identify a ‘better’ violin.

The acoustic impedance vs. frequency curves at the bridge explain a heck of a lot, and depend on the structural properties of the plates.

Thirdly, whether it is possible to ever completely separate myth from reality in order to compare different instruments.

Myths are shown to be myths by experiment. I’m happy to discuss any aspect of any experiment in order to banish myths.

[Tusculan]

Voice recognition is an entirely different thing than music recognition, even though there may be similar algorithms used. A bit of googling did lead me to older (1998) research claiming software having a 70% success rate in identification of musical instruments. link, see also a PDF paper link

SentientMeat, I don’t dispute your research. However, there are several reasons why I still don’t agree with the conclusions that some of the posters in this thread (but apparently not you) seem to draw from it.

1. That you didn’t manage to find differences in a wood plate doesn’t preclude that there are still changes that are so small that your instruments didn’t manage to pick it up, but that might be of perceptible influence to the ultimate sound (but I admit that I don’t find this too likely).

2. That you didn’t notice changes in a wood plate doesn’t rule out that a board in a violin, which is tailored and bent to provide good resonance, is much more susceptible to playing, either because it undergoes stronger forces because of the resonance, or because it may get in the right position for providing better resonance.

3. The wood itself is only one of a lot of factors that determine the sound and the violin’s quality. I’ll leave out the ‘external’ factors as the bow and strings; there still are things like board thickness (and local variances in thickness), the placement of the sound holes, the exact shape of the board, and the like. A violin doesn’t have a flat board like a guitar. And that isn’t just to make it prettier.

Granted, your research serves to cast doubt on the question whether the wood is involved in preserving the quality of a violin by regular playing, but it doesn’t provide sufficient evidence for further conclusions.

[GorillaMan]

@Tusculan

A bit of googling did lead me to older (1998) research claiming software having a 70% success rate in identification of musical instruments.

That’s the success rate at identifying pianos from trumpets, not Strads from Amantis. A bit like using voice recognition identifying a dog from a cat. Granted, it’s a few years old, but it reaffirms that the technology is nowhere near doing what we’re talking about.

[Tusculan]

Oh, I’m with you. I just wondered whether it was already possible to distinguish a saxophone from a clarinet, or a horn from a trombone. I hadn’t thought they were already that far. That makes me wonder, though, whether the software can only handle the easy cases (piano vs flute) or also the ones that even humans need experience for.

[SentientMeat]

That you didn’t manage to find differences in a wood plate doesn’t preclude that there are still changes that are so small that your instruments didn’t manage to pick it up, but that might be of perceptible influence to the ultimate sound (but I admit that I don’t find this too likely).

There is such a thing as a minimum perceptible difference in acoustics, and our instruments operate past this threshold such that they can pick up a difference in amplitude at a given frequency which a human could not.

That you didn’t notice changes in a wood plate doesn’t rule out that a board in a violin, which is tailored and bent to provide good resonance, is much more susceptible to playing, either because it undergoes stronger forces because of the resonance, or because it may get in the right position for providing better resonance.

Be very careful with this word ‘resonance’ - most musicians have a rather wooly and erroneous understanding of what it means. A resonance occurs at a frequency which provides maximum vibrational amplitude for a given force (think of the opera singer and the wineglass). The resonances of the wooden plate are determined before it is constructed, and the process of shaping and bending the plate into a violin modifies those properties only slightly. Put simply, the vibrational properties of the rectangular wood plate sitting on the workbench are not much different to when it is part of a working violin.

there still are things like board thickness (and local variances in thickness), the placement of the sound holes, the exact shape of the board, and the like.

Of course, but these things are well understood and, like I say, do not affect the vibrational properties that much. That’s why I started out in this thread by saying that some pieces of wood will sell for hundreds or thousands of dollars because we can be so sure that they will produce an instrument of exceptional quality.

Granted, your research serves to cast doubt on the question whether the wood is involved in preserving the quality of a violin by regular playing, but it doesn’t provide sufficient evidence for further conclusions.

The violin is the wood. The only thing that changes is the *player[/]i.

[Tusculan]

Sorry, but you’ve lost me. Do you seriously mean that a square wooden box has exactly the same properties as a violin? That the form of a violin is only for prettiness and doesn’t have any influence on the sound coming out of it?

I’ll grant you that I’m not an expert in acoustics; I just have my ears to go by. But I can’t believe that the form of the ‘box’ doesn’t have any influence, as you seem to imply.

[SentientMeat]

Do you seriously mean that a square wooden box has exactly the same properties as a violin? That the form of a violin is only for prettiness and doesn’t have any influence on the sound coming out of it?

Not exactly no: the shape and bend of the violin plate does constrain the displacement a little more ‘efficiently’ than a rectangular box would, but it’s not all that significant, really. “Just for prettiness” is a largely accurate description.

[Aldebaran]

I am with Tusculan to ask you where on earth you get the idea that the craftmanship involved in creating a violin comes down to making a “box”.

You have not only shaped wood
but also glue used to bring the pieces together
varnish
residu of dust etc… inside the violin
and on it (think only about the material - don’t know Enlgish name - used on the bow’s hair, leaving its residu whenever you play)
All factors which cause on their own already alterations in the sound the wood gives when brought into vibration (you hear already a clear differece when you cleaned your violin occasionally on the outside = clean its varnish from sticking elements like the residu left by the bow movements).

Leaving out all these factors:
Is the wood you use for your testing hundreds years old, is it during that period exposed everything the wood of a violin is exposed to when you cause the chain reaction for its vibration, producing the tone?
Did you have such wood and could you compare it with the reacions of “new” wood like is used in modern violins?

If not: How can you be 100% certain that being exposed to the forces causing the vibration of the wood, and this during centuries and typical in circumstances the wood of a violin is exposed to this, does not cause an alteration in the wood struture and influences thereby how it reacts on this exposure.
By measuring and theorizing the behaviour of new wood and merely theorizing on how such old wood supposedly behaves when undergoing the same experiments?
For the “let’s do the test, nothing easier then that” department I have an other difficulty you can try to overcome:
One violin “behaves” not the same in a certain environment then an other in that same environment. You ignored completely the factor “acoustics of the theater”.
If you want to do it all so scientific, you must be able to control every single detail and influence on everyting involved in your test.

So far I didn’t see anything that comes even close to guarantee this.

Salaam. A

[GorillaMan]

@SentientMeat

Not exactly no: the shape and bend of the violin plate does constrain the displacement a little more ‘efficiently’ than a rectangular box would, but it’s not all that significant, really. “Just for prettiness” is a largely accurate description.

My bolding - if you’re talking about the front and back of the body, there’s no bending involved. I’m disconcerted that you might be making such an error about violin construction.

[Aldebaran]

@SentientMeat

Not exactly no: the shape and bend of the violin plate does constrain the displacement a little more ‘efficiently’ than a rectangular box would, but it’s not all that significant, really. “Just for prettiness” is a largely accurate description.

Please, you can’t be serious about this.
Why on earth do you think that the old masters learnd their job from childhood on, why took it so many years to learn it and why do you think not every violinbuilder can build top quality violins?
It sounds as if you think a computer-led machine produce a top quality violin if you only put the right information in the computer to have the robots do the job.
Salaam. A

[SentientMeat]

GM:: if you’re talking about the front and back of the body, there’s no bending involved. I’m disconcerted that you might be making such an error about violin construction.

OK, the arching of the plate if you prefer - it still doesn’t affect its vibrational properties much, but sorry for the sloppy terminology.

Alde:All factors which cause on their own already alterations in the sound the wood gives when brought into vibration (you hear already a clear differece when you cleaned your violin occasionally on the outside = clean its varnish from sticking elements like the residu left by the bow movements).

The type of glue, varnish and dust residue does not affect the vibrational properties of the wood: again, I would offer an experiment in which you were asked to tell the difference in a blindfold test.

Is the wood you use for your testing hundreds years old, is it during that period exposed everything the wood of a violin is exposed to when you cause the chain reaction for its vibration, producing the tone?

Like I’ve said perhaps eight times now, so long as the storage conditions are within certain temperature and humidity thresholds, nothing changes the wood’s vibrational properties (except aging via drying out perhaps, and even that is not clearly apparent by any means).

Did you have such wood and could you compare it with the reacions of “new” wood like is used in modern violins?

Yes - the wood in modern violins is similar (but never identical) in its vibrational properties to that in Strad manufacture.

How can you be 100% certain that being exposed to the forces causing the vibration of the wood, and this during centuries and typical in circumstances the wood of a violin is exposed to this, does not cause an alteration in the wood struture and influences thereby how it reacts on this exposure.

The forces do not matter, only the storage conditions.

One violin “behaves” not the same in a certain environment then an other in that same environment. You ignored completely the factor “acoustics of the theater”.
If you want to do it all so scientific, you must be able to control every single detail and influence on everyting involved in your test.

The theatre is not the violin. The experiments hold the player, musical sequence, string, bow and environment constant as far as they reasonably can, such that the violin is the variable, which is why people can identify specific violins in blind tests, all else being constant. We must ask what factors exceed the minimum perceptible difference threshold.

Please, you can’t be serious about this.
Why on earth do you think that the old masters learnd their job from childhood on, why took it so many years to learn it and why do you think not every violinbuilder can build top quality violins?

Because not all violin builders can identify the pieces of wood with the appropriate qualities. Their crafsmanship is not in doubt, but there is a lot of unnecessary ornamentation in violin making, acoustically speaking, such as its traditional and beautiful shape.

It sounds as if you think a computer-led machine produce a top quality violin if you only put the right information in the computer to have the robots do the job.

Again, that’s to do with craftsmanship (which is a little different to acoustical expertise in choosing pieces of spruce), but given the appropriate materials I don;t see why it need be such a dark art.

[GorillaMan]

@SentientMeat

OK, the arching of the plate if you prefer - it still doesn’t affect its vibrational properties much, but sorry for the sloppy terminology.

Sorry for still not understanding your point properly - are you saying that any two pieces of spruce, carved into the same shape, will create a violin with the same sound and behaviour? Or that the shape they are carved into does not matter? In either case, what is it that makes every violin sound different?