Fritz-blind test

[Don Noon]

But Don, the effect is perfectly audible. I'd think there should be something there. Sometimes they get better and sometimes they're a bit "deflated" by the end of Tchaikovsky's.

The violin or the violinist? There might or might not be something real that happens to the instrument sound, but it would take some more care to prove that it's in the instrument, not the player, recording equipment, or listener. I haven't seen that done yet. My tests indicate it's not a big effect, if any, and even if it was measurable or audible, I'm not sure how I could use that information to build a better fiddle. That might make me something of an infomercenary, but my main interest is building good fiddles, not basic research. There's plenty of opportunity to spend gobs of time on arching, graduations, wood properties, etc., which appear more useful to a builder.

[Flyboy]

The offer is meant to be tongue in cheek and I will happily not work on your chinese fiddle. You are correct that nobody here stated that they could pick them out. There are 16+ pages of discussion on a subject that comes up constantly in the profession. Always raising quite a stir, overwhelmingly with the same results, yet never actually proving the point well enough to put it to bed. The fact that it gets so much attention and effort means that either not enough people are convinced, which is beyond me, or that the people who think it's important to still conduct such tests are doing so for reasons other than to prove, again, what has already been shown. The intent of the Fritz test was to show weather skilled violin professionals can discern the difference between Strads and Moderns. To me, it's a convincing enough test. I'll say that I'm biased towards moderns. I'm also skeptical to a fault. So, my position is not as strong as others.

I've been to two such tests. Neither were as unbiased as this one and they were still inconclusive. I've also heard moderns played back to back with Strads in uncontrolled comparisons. I've preferred each at different times. I'm not saying that I've had anywhere as much exposure as others on the list or in the field. It was enough to convince me that the comparison is pretty much pointless. Comparing the best instruments, irrelevant of their history, offers more in my mind. Why put some arbitrary restriction or twist on the test just because of tradition?

Anyhow, just my ranting. I'm sure that when it's over, I'll repeat my vow to stay out of this kind of thread just to break my promise the next time.

--Joe

Actually, the Fritz experiment did show that a number of players were able to pick out ONE Strad. They just didn't care for it.

[NewNewbie]

I took the listening test, but instead of trying to tell which instrument was the 'Strad' I just picked which instrument sounded the best.

Oddly enough, I never have a hard time doing that.

My score by the way was 100% .... oddly enough .... again.

Whatever happened to "Leave Good Enough Alone"? :huh:

[stephen maloney]

Are these highly publicized results of the blind study earth-shattering or what?

Pretty hard to refute.

And consider the following, in the aftermath:

--There are unconfirmed rumours of heavy breathing and palpitations at Bein and Fushi...

--At Tarisio, xanax prescriptions are up 1000%.

--Burgess was seen partying at a bar last night with two very young (possibly underage) sheep.

Case closed. Le mythe de Crémone est mort.

[Evan Smith]

Oh well,,,,,,mabe I'd feel slightly better if the favorite was ricci's "Luiz Bellini"

[Addie]

Le mythe de Crémone est mort.

[Addie]

Uh-oh, look out!

[Michael.N.]

go aussies go, 300+ runs for clarke, but david burgess could do that blindfolded, according to his website

i dont know about the rest of you guys, but when i pick out an instrument, seeing it clearly is half the point, and if im going to enjoy listening to a stradivarius, i dont want the player to be blindfolded or wearing beer goggles

michael n its an unpublished study i did working for years selling and designing stereo speakers, believe it or not, playing perfectly in tune, and seperating good from bad sound do not always go hand in hand, in fact musicians are often so concerned about the music and intonation, that they have less concern about tone than some non players

That's not tone deaf Lyndon. Tone deaf (Amusia) is a specific condition that means one cannot distinguish pitch, it occurs in approximately 5% of the population. In musicians it will be considerably lower and probably zero in good Violinists. In fact I simply can't see how anyone can play the Violin well if they suffer from Amusia - but I could be wrong.

I think what you are referring to is an inability to distinguish timbre. I doubt even then that it's anywhere near to 75%, not even remotely close. I'd like to see some serious study on it if that's the case.

[Melvin Goldsmith]

The results of this test are interesting and maybe not surprising but I would warn against drawing too many conclusions from it or previous tests that gave similar outcomes.

My first objection to this test is that there is no 'control' to test the experiment itself.. If the idea was to see whether some of the finest makers of today can match Strad & co in a blind test ( however it was worded)there should have been one or two good cheap factory fiddles in there well set up with a cost of about £3000 half of that being on set up. My hypothesis is that such fiddles would have held their own in this test.... Imagine the reaction then...! A lot more people would be questioning the validity of this test for sure.

My second objection to the test is that regardless of price no player would buy an instrument given the limited time testing for each violin... and the paper documents players saying they need more time..It's a bit like only being allowed to test drive a car around the dealers lot. A players initial reaction to an instrument is not always the same when they have played it for two weeks. That is why it is quite normal for musicians to try a violin for at least that period of time before they decide to financially invest in it or not. The test is only really a test of initial reactions...violins are a bit like partners in love...often what initially seems very obviously attractive soon can feel shallow, tarty and unattractive and vis versa. Some players are real violin Jockeys and can assess a fiddle quite fast but I've never seen one who'd put their money down immediately if making a major investment in a principal instrument to play on via the criteria of this test

My third objection to this test is that as the authors say it was not easy to obtain the Strads & del Gesu for the experiment and they made do with what they could get. On the other hand unless advised otherwise the modern violins might have been hand picked & the best and most fettled the makers could offer? It is not a surprise to me that in the conditions of the experiment with a relatively short time to asses each of the many violins that moderns and olds came out with no big distinction between the two...I would also be interested in the psychological effect of the experiment on the players.... coming from visual disorientation etc It is not normal for players to be blinded and asked to judge many violins. I hate to say it but the blindfold is a well used tool of disorientation in interogation. The players themselves may have felt some pressure in case they cannot tell a Strad from a modern...It is not a natural environment at all..

My fourth objection to this test is that it involved moderns selected from some very fine makers against what appears to be a random sample of Strads & del Gesu. I don't have any problem with the idea that good modern violins can match some old Cremonese..In fact my work is based on doing that. Often some of the old Cremonese don't play well....maybe a bit tired! On the other hand. I have personally heard/ Played?assessed with virtuoso player a few Strads and Del Gesu that seemed to have supernatural qualities that I have never experienced elsewhere...It would be impossible to get a full set of instruments like this into the test but the result would be different. I strongly doubt that the best of moderns can yet match the best old Cremonese.

There is a disturbing cult like tendency amongst modern makers to give each other plaudits and awards, to think we knew how the old guys did it and we are as good as them. …..It's no time to be patting ourselves on the back and thinking we are as as good as Strad.

I'll say it again. A $1500 factory fiddle selected for looking right and with 1 $1500 makeover would have given a good account of itself in this experiment....

[James M. Jones]

I think what you are referring to is an inability to distinguish timbre. I doubt even then that it's anywhere near to 75%, not even remotely close. I'd like to see some serious study on it if that's the case.

Any one who can recognize a voice is not tone deaf,different levels of sensitivity to timbre, yes, but deaf ,no.

Boy, I wish I could remember the NPR program ....I was driving across south Dakota ...they were talking about perfect pitch ....and said that in fact most had it ,but couldn't use it . As an example they suggested to "think" about "listing to your favorite song" or something to that effect, The audio image in the mind would be on pitch.The problem became when people tried to use it,the signal memory was to weak to be useful.

[AtlVcl]

The amazing diversity of opinion in this thread leads me to the conclusion that something must be...

on the Fritz...

[Anders Buen]

My best guess is that the best Strad in the test is the Titian, and that the del Gesu in there is a good one related to any of the judges of the player competition. At least the Titian I've heard is supposed to be one of the best Strads when it is well set up. At least that was Sam Z's words in a Strad article on it. It is remarkably even sounding.

[Melvin Goldsmith]

My best guess is that the best Strad in the test is the Titian, and that the del Gesu in there is a good one related to any of the judges of the player competition. At least the Titian I've heard is supposed to be one of the best Strads when it is well set up. At least that was Sam Z's words in a Strad article on it. It is remarkably even sounding.

Anders, I also guessed it is the Titian.

[Doc Eastman]

From the PNAS conflict of interest policy: "Such conflicts of interest include relationships with corporations whose products or services are related to the subject matter of the article. These relations include employment, substantive ownership of stock,1 membership on a standing advisory council or committee, service on the board of directors, or public association with the company or its products.."

Since so many people in a field of interest have real or perceived conflicts of interest, the approach in medical research is for everyone involved to file their conflicts of interest in writing, so they are available to people who review the data, publishers, and people who read the study results. If someone has a vested interest (stock holder, or could otherwise materially benefit from the results), they should be excluded from any critical vote or result. In clinical trials they usually leave the room during critical discussions and voting, as their very presence may bias other participants who might not want to offend the vested person. So if a maker placed a violin in the trial and then participated in the testing, their conclusion must be excluded from the data analysis. IMHO they should not have been present during the test. In the test under discussion if probably didn't make a difference, since most gave the wrong answer or couldn't tell.

[martin swan]

The point of the test was that the violins weren't identified for the players. So I don't see how any conflict of interest could affect the test.

However, I don't know of any such possible conflict of interest. Some people have implied that one of the violins was a Joseph Curtin - is there any evidence of this, or is it just more flailing?

It would take a brave maker to put a violin up for blind testing against the Titian! Hats off to him if he did ....

[Addie]

I can see that having a Curtin in the test might be advantageous to him (good advertising), however, I think most objections would amount to sour grapes.

I hope this test encourages further and more rigorous tests in the future. I can think of a lot worse uses for taxpayer funding!

[Torbjörn Zethelius]

It would take a brave maker to put a violin up for blind testing against the Titian! Hats off to him if he did ....

Actually, no. There's nothing to lose except for the owner of the Strad. He's the brave one.

[Evan Smith]

The results of this test are interesting and maybe not surprising but I would warn against drawing too many conclusions from it or previous tests that gave similar outcomes.

My first objection to this test is that there is no 'control' to test the experiment itself.. If the idea was to see whether some of the finest makers of today can match Strad & co in a blind test ( however it was worded)there should have been one or two good cheap factory fiddles in there well set up with a cost of about £3000 half of that being on set up. My hypothesis is that such fiddles would have held their own in this test.... Imagine the reaction then...! A lot more people would be questioning the validity of this test for sure.

My second objection to the test is that regardless of price no player would buy an instrument given the limited time testing for each violin... and the paper documents players saying they need more time..It's a bit like only being allowed to test drive a car around the dealers lot. A players initial reaction to an instrument is not always the same when they have played it for two weeks. That is why it is quite normal for musicians to try a violin for at least that period of time before they decide to financially invest in it or not. The test is only really a test of initial reactions...violins are a bit like partners in love...often what initially seems very obviously attractive soon can feel shallow, tarty and unattractive and vis versa. Some players are real violin Jockeys and can assess a fiddle quite fast but I've never seen one who'd put their money down immediately if making a major investment in a principal instrument to play on via the criteria of this test

My third objection to this test is that as the authors say it was not easy to obtain the Strads & del Gesu for the experiment and they made do with what they could get. On the other hand unless advised otherwise the modern violins might have been hand picked & the best and most fettled the makers could offer? It is not a surprise to me that in the conditions of the experiment with a relatively short time to asses each of the many violins that moderns and olds came out with no big distinction between the two...I would also be interested in the psychological effect of the experiment on the players.... coming from visual disorientation etc It is not normal for players to be blinded and asked to judge many violins. I hate to say it but the blindfold is a well used tool of disorientation in interogation. The players themselves may have felt some pressure in case they cannot tell a Strad from a modern...It is not a natural environment at all..

My fourth objection to this test is that it involved moderns selected from some very fine makers against what appears to be a random sample of Strads & del Gesu. I don't have any problem with the idea that good modern violins can match some old Cremonese..In fact my work is based on doing that. Often some of the old Cremonese don't play well....maybe a bit tired! On the other hand. I have personally heard/ Played?assessed with virtuoso player a few Strads and Del Gesu that seemed to have supernatural qualities that I have never experienced elsewhere...It would be impossible to get a full set of instruments like this into the test but the result would be different. I strongly doubt that the best of moderns can yet match the best old Cremonese.

There is a disturbing cult like tendency amongst modern makers to give each other plaudits and awards, to think we knew how the old guys did it and we are as good as them. …..It's no time to be patting ourselves on the back and thinking we are as as good as Strad.

I'll say it again. A $1500 factory fiddle selected for looking right and with 1 $1500 makeover would have given a good account of itself in this experiment....

Exactly !

Very well said !

If your gonna make big claims you need 500% more control than this!

If your just gonna say ,,some players can't tell the diference,,It's good enough,,

just another one of """those""" tests.

You seem to be the only one that gives setup the attention that it deserves.

Bravo!

[lyndon]

well it seems that 6 test subject, roughly 25% of the testers not only picked one strad, but according to the articles all three old instruments from the new ones. i hate to say it but i wouldnt expect more than 25% to get it right, its not easy to do, given the test conditions, and i know some of you will try to say that those 6 were because of random chance, well if we knock out the 6 votes as random, we have to knock out 6 people who voted for modern as random, and so we only have 25% roughly prefering modern

the tests are all using the same choices, strads vs top top modern, how about top modern vs antiques in the same price range, how about classic eh roths vs strads, how about cheap chinese vs expensive modern etc, unless we expand the survey to include different categories of violin, we have no idea if the test situation is generating random result, or consistently predictable ones

from a scientific standpoint, the same study needs to be repeated 3 or 4 times with totally different subjects, if the results are basically the same every time, we have a lot more to go from, however as the study stands right now, we have no idea how much pure chance and wild guessing had to do with the study.

[Michael.N.]

This is nonsense. Your $3,000 Violin (or any other Violin) is not a 'control' in the strict sense. In this test it simply becomes another Violin, made in a slightly different method to the others but with a much lower price tag. That in itself does not make it a 'control', it just widens the study. Nothing more, nothing less. It wouldn't make the study any more or any less valid. It simply wasn't part of the remit. You may well say that it would be interesting to include other Violins, that's for another study and of course there has to be a limit on the numbers. Seriously, throughout this thread there have been dozens of bullets fired at our Fritz Violin test, pretty much all of them are hitting the double Bass stood in the corner.

[martin swan]

As Michael N says, people are increasingly missing the point of this test and getting caught up in the sideshow.

As I understand it, Fritz's research is an attempt to broaden the criteria by which violins are judged to include a parameter called PLAYABILITY.

This test was examining how players assess violins (irrespective of what they are) for playability. The test and the use of statistics relate as much to social sciences as to physics, and this experiment should be seen as part of a wider attempt to find ways of testing violins which relate more closely to players' experience and the needs of real musicians. Up until now we have been relying largely on physics and art history.

I can't help but be interested in this - since I joined Maestronet I have been making what I feel is an important point. Physics can help us to identify some tonal characteristics in individual violins, but so far it hasn't contributed anything to our understanding of playability. Professional or experienced players will always rate this quality of playability over "static tone" , and for all the talk to the contrary, they will always choose a violin which sings for them at close quarters and which makes music (assuming the label's right).

The parameters explored by this test in 6 good violins were

"Playability" : a subjective assessment which is of paramount importance to a player

"Projection" : we can debate whether an experienced player can tell this from a nearfield test, but in the context of this test I think it means something more like "volume"

"Tone Colours" : note the deliberately wide classification, which accepts that tone is not a single or static parameter

"Response" : also a player's term, which signifies a very personal amalgam of otherwise measurable parameters

Let's forget for a while about what the violins were and concentrate on what's revolutionary about this test. It's a peer-reviewed experiment, well conducted and published in a major scientific journal, which is not about impact hammers, ffts and spectrum analysis, but about what players actually look for in violins. Maybe people should take a time out and allow that idea to actually sink in .....

I'm fascinated by the findings of acousticians and other scientists who post here, but for me Fritz's investigation into playability is an essential counterbalance to the cult of measurement.

Martin Swan Violins

[Anders Buen]

I am not without critical opinions on this study either.

E.g. the article states in the abstract: Player’s judgments about a Stradivari’s sound may be biased by the violin’s extraordinary monetary value and historical importance, but no studies designed to preclude such biasing factors have yet been

published.

I do not think that can be correct. There are several studies published on the matter, but basically using just listening in blind or double blind tests. A famous old example is Sauders works with Heifetz in the 30-40ties. To blindfold the players playing on real instruments might be new. Weinrich used measured radiativity spectra to reproduce the sound from a 'dead played violin' through headphones. I guess using convolution of the played signal with impulse responses of different quality instruments. But that would not give a realistic feel, and the sound quality would probaly also be limited byt eh computing capabilities at that time (80-90ties). Fritz et al have also used such techniques in their work, e.g. in doing tests of how large the changes to spectra of violins need to be before players can detect it (from the sound).

Other posters here have mentioned that this is one out of a series of experiments in this matter done using different approaches. The BBC Stern experiment is an example. However also a listening test. Some have participated in unpublished experiments and 'play around'.

[Anders Buen]

As I understand it, Fritz's research is an attempt to broaden the criteria by which violins are judged to include a parameter called PLAYABILITY.

This test was examining how players assess violins (irrespective of what they are) for playability. The test and the use of statistics relate as much to social sciences as to physics, and this experiment should be seen as part of a wider attempt to find ways of testing violins which relate more closely to players' experience and the needs of real musicians. Up until now we have been relying largely on physics and art history.

No, not really. It is my impression that the playability aspect is a new one included in this test as compared to what I can read out of the former articles and tests I have seen data from. Basically my own observation. I can be somewhat impreceise and I am quite frankly not reading all material through in detail. Often studies contain more aspects than those reported in the publications. E.g. they can be a part of a larger study where more overall conclucions might be made across the studies. I am speculating a bit here.

I think the main purpose in Friz et als study are reflected in the sentences:

This study explores player preferences under two sets of conditions. One set, designed to maximize ecological validity, emulated the way players choose instruments at a violin shop, where they typically try a selection of

instruments before selecting one to take home for further testing.

..

Our second set of test conditions, designed with the statements of Weinreich and Langhoff in mind, asked subjects to assess instruments rather quickly.

Weinreich and Langhoff are quoted: Weinreich (1) argues that any experienced player can classify a violin as a “student,” “decent professional,” or “fine solo” instrument; furthermore, “the judgment would not take more than about 30 s, and the opinions of different violinists would coincide absolutely.” According to Langhoff (13), “any musician will tell you immediately whether an instrument he is playing on is an antique instrument or a modern one.”

These senteces are from works in the 80ties and 90ties, some 20ish years ago in the high time for plate tuning etc. I think the modern violin makers have reached further through these years. That issue is not discussed in the article.

Basically the test is about how fine old versus fine new violins get rated by their sound and playing qualities alone.

[David Burgess]

Exactly !

Very well said !

If your gonna make big claims you need 500% more control than this!

Silly researchers. One thing they failed to take any data at all on was fingernail length, to see if that might have been related to the instrument preferences.

[JohnCockburn]

Since so many people in a field of interest have real or perceived conflicts of interest, the approach in medical research is for everyone involved to file their conflicts of interest in writing, so they are available to people who review the data, publishers, and people who read the study results. If someone has a vested interest (stock holder, or could otherwise materially benefit from the results), they should be excluded from any critical vote or result. In clinical trials they usually leave the room during critical discussions and voting, as their very presence may bias other participants who might not want to offend the vested person. So if a maker placed a violin in the trial and then participated in the testing, their conclusion must be excluded from the data analysis. IMHO they should not have been present during the test. In the test under discussion if probably didn't make a difference, since most gave the wrong answer or couldn't tell.

In this context, "conflict of interest" refers to the authors of the paper, not participants in the trial.

By the way, it doesn't matter to me one jot whether Curtin should have declared conflict of interest or not. Someone else raised it, and I just supplied the relevant section from journal policy.