Violinists can't tell a Strad from a new violin -- in a 30-second guessing game!

April 7, 2014, 9:59 PM · The New York Times headline says, "A Strad? Violinists Can’t Tell." USA Today's headline said, "Violinists can't tell new violins from old, study shows." The Daily Mail: "Is it all just a fiddle? World's leading violinists CAN'T tell the difference between a Stradivarius worth millions of pounds and a modern instrument."

They are talking about a study by Claudia Fritz, Joseph Curtin, Jacques Poitevineau, Hugues Borsarello, Indiana Wollman, Fan-Chia Tao, and Thierry Ghasarossian, whose results were published online April 7 in the Proceedings of the National Academy of Sciences. The study was conducted in September 2012 at the Auditorium Jean-Pierre Miquel (Coeur de Ville) in Vincennes.

Even the study itself, entitled "Soloist evaluations of six Old Italian and six new violins," claims, "The current study, the second of its kind, again shows that first-rate soloists tend to prefer new instruments and are unable to distinguish old from new at better than chance levels."

Is that really what this study proved? I thought it might be a good idea to read the actual study and look at the details of the process and supporting information.

Here are some thoughts, after doing so.

First, the luthiers and social scientists who did this study picked a nice group of people to participate: the 10 soloists who tested the violins were: Olivier Charlier, Pierre Fouchenneret, Yi-Jia Susanne Hou, Ilya Kaler, Elmar Oliveira, Tatsuki Narita, Solenne Païdassi, Annick Roussin, Giora Schmidt, and Stéphane Tran Ngoc. (That's seven males, three females). Based on the information given, two of those people normally play new violins, one plays both a modern and a new; and seven play old violins.

Ilya Kaler
This framegrab image from video, provided by Stefan Avalos, shows soloist Ilya Kaler wearing welder glasses so he can’t see the violin during a test of old and new instruments outside Paris in September 2012.

In two 75-minute sessions, players were asked to evaluate 12 violins -- six old, and six new -- and choose which one they would want, if they were replacing their own violin. In the final seven minutes of the last session, they were given violins in rapid succession and ask to identify which were old and which were new. The whole experiment was "double-blind," players wore goggles and could not see the instruments. The study listed no information about the preparation of the violins; whether the makers of the new instruments were hovering nearby, ready to adjust the soundpost and bridge, whether luthiers even looked at the Strads before they were used (this was a problem in the first test). Also, we don't know how old the strings were on any of the violins, and if they were all the same age. That can make a huge difference in the perception of a violin, whether the strings are brand-new; new but nicely broken in; or just plain old. I'd be curious, and maybe there's an answer for that.

When it came to preference, six violinists preferred new violins and four preferred old ones. The players did not do better than the roll of dice when it came to the seven-minute guessing game at the end.

What does all this prove? To me it proves that new violins do better than 300-year-old ones in brief, blind tests. I'm not in the least bit surprised by this. Time and time again, violinists tell me that the 300-year-old Italian violins, such as the Strads of the Golden Period, take time to learn to play -- years, even. The rewards come over a long period of active partnership, which also involves experimenting with set-up, strings, etc. If the period of time is a couple of hours, then I would imagine that yes, new violins would be easier and more preferable. Even so, just 6 of the 10 soloists in the study preferred them. It was nice, though, that the players had a bit more time to evaluate these violins than in the first study.

The soloists certainly were not given much time, when it came to identifying which violins were old and which were new. In fact, that part of the experiment took place for a total of 7 minutes, at the very end of the second of two 75-minute sessions. Instructions for the players were: "We will now present you with a series of violins one at a time in random order. Play each for 30 seconds then guess what kind of instrument it is." In other words, at the point where they would have the highest amount of fatigue from playing all these violins, they were given a half-dozen or more violins in rapid succession, in random order, and asked to identify them as being old or new in 30 seconds apiece.

You can tell me how legitimate that sounds to you. I welcome your thoughts, and certainly I'd recommend that everyone read the actual study before letting the headlines do the thinking.

Let me add, I would agree with many people who have said that the present moment is also a golden period of violin making. There are some amazing luthier/artists out there whose extraordinary work deserves the highest praise and whose instruments deserve to be in the hands of the finest players. I also respect the old Italian instruments for what they are: shining examples of the art of lutherie that have withstood the test of time.

Replies

April 8, 2014 at 05:59 AM · Greetings,

I think your analysis is spot on,

cheers,

buri

April 8, 2014 at 06:07 AM · Thanks Laurie for going to bat on this issue. I have had the opportunity to play on a large number of modern instruments and older ones as well. They have so many differences from the shape, age, and other factors as you mentioned. It usually takes a breaking in period for the instrument to get warmed up way beyond the limitations in this study. Under those conditions a student instrument would outperform the Guarneri and Stradivari in certain hands. The headline's perpetrate the weakness of the masses to jump to conclusions which is hardly the artistic sensitive subjective treatment this study needs. There are too many variables in this study and it is not in the least bit scientific. Therefore taste is everything. 30 seconds is like going to 31-flavors ice cream and getting all chocolate samples with your eyes closed. The choice is so arbitrary!

April 8, 2014 at 06:20 AM · You pretty much summed it up, Laurie.

You deserve endless prunes.

Regards,

Ellis

April 8, 2014 at 09:32 AM · Owning or playing a Strad is symbol of social status for violin players. Just like so many people have much fantacy about Rolex. Is it because Rolex has better "time" than other watches?

April 8, 2014 at 12:54 PM · "The soloists certainly were not given much time, when it came to identifying which violins were old and which were new. In fact, that part of the experiment took place for a total of 7 minutes, at the very end of the second of two 75-minute sessions. Instructions for the players were: "We will now present you with a series of violins one at a time in random order. Play each for 30 seconds then guess what kind of instrument it is." "

_______________________________

Players will often claim that they know right away whether an instrument is old or new, so I reckon this is what was being tested in that portion of the experiment. This is just one in a series of studies, examining various claims, and using varying approaches.

To me, it's all interesting stuff, if one can set aside the way some of the media has sensationalized it.

April 8, 2014 at 01:26 PM · Sounds pretty legit to me. I would also like to add that it takes time to develop a relationship with a modern fiddle as well. Nothing in this study was a surprise to me. Congrats on your book, Laurie! :)

April 8, 2014 at 01:29 PM · I wish that the luthiers were also mentioned along with the violinists. They deserve the recognition.

April 8, 2014 at 02:03 PM · What exactly is meant by "new" violin? Is it one right off the bench? One that's been broken in for 5, 10, or 20 years?

April 8, 2014 at 03:24 PM · David, it is interesting, for sure. I don't personally know any violinists who claim to be able to identify a Strad vs. new violin in 30 seconds, but then again I only know a few thousand violinists. ;) I suppose it's the kind of thing that a luthier might hear in the shop, more than a player would hear in a rehearsal (or an interview, or on a violin discussion board). Hearing it even once would undoubtedly be very irksome.

I tend to think that it's the luthiers who possess miraculous powers of identifying particular instruments, but probably more by sight than by sound.

Recently I spoke to a highly accomplished luthier (the violin he showed me was just gorgeous in every way) and he talked about the real frustration in having violinists compare everything that a luthier does to a Strad. He asked, as a violinist, does everyone compare you to Heifetz? Do you "suck" if you aren't as good as Heifetz? (Actually, with the advent of Youtube, we do all get compared to crazy high standards!) But I did see his point.

Still, all violinists compare modern luthiers to Strads? I doubt that. My hypothesis: Most violinists actually don't compare everything a luthier does to a Strad. Most have never played a Strad or heard one in person. People just want a nicely-crafted instrument that sounds lovely and serves them, and continues to do that over time.

Scott, that's an interesting question, and the study doesn't identify the age of the new violins or their makers. Here's what the study says: "We assembled a pool of 13 new violins (from professional makers around the world) and 9 distinguished Old Italians, including six by Stradivari (made between 1710 and 1730) and two by Guarneri del Gesu (circa 1740). " They then selected 6 from the 13 new violins, and 6 from the 9 old ones. Also, there's one old violin they don't identify by maker.

April 8, 2014 at 03:53 PM · There can never be a study that determines conclusively whether new or old violins are better until "better" can be defined. Whether the inability to tell which of two violins is older stands as proxy for "neither is better" is highly questionable. The relationship between player and instrument (which includes the bow!!) is too nuanced. As long as the ethical requirements for the use of human subjects is met, then collection of data poses no harm. It's the interpretation of data that requires great care.

Seven males and three females -- okay, but how was the group divided with respect to the use of the shoulder rest?

April 8, 2014 at 05:07 PM · "Still, all violinists compare modern luthiers to Strads?"

No, but a small percentage of people who are considering the purchase of a modern violin actually do compare them side-by-side with a Strad or two. ;-)

A more fundamental reason behind this series of tests though (and there are more than have been published) is that most luthiers and players have long considered Strads to be kind of a reference standard for high quality. So an early goal was to be able to make violins which sounded like Strads.

Along the sometimes tortuous and convoluted path in search of that goal, some potential problems emerged with the method: Strads could sound very dissimilar from each other, and the Strads used for reference weren't always preferred over the new instruments by either players or listeners. What to do about that? The first step was to try to get a better handle on it by doing more tests, more carefully, and using larger sample sizes. That's kind of where things stand right now, with this research group, along with trying to take "voiceprints" of the violins which are preferred.

Should they ultimately determine that "Strad sound" isn't the most useful goal, I don't know what will take its place. Hopefully, not a kazoo. ;-)

April 8, 2014 at 05:22 PM · I was in Paris for this study and shot extensive footage of it for a larger documentary I'm working on. As a filmmaker AND violinist, it was amazing to be able to be so intimately involved.

For those interested, a 'brief' version of the doc is available to be seen (if you haven't already watched it). It answers some of the pervasive questions and comments people have.

Something important to note that is ALWAYS left out of the media stories: ALL of the violins sounded amazing. Just because one didn't fair as well as another doesn't change that fact in the slightest.

a five minute version:

A longer version - likelier of more interest to violinists and enthusiasts than just the five minute version.

April 8, 2014 at 05:32 PM · Thank you for an interesting article. It says six violinists (out of 10) felt that they had enough time with the instruments while eight people were confident in their choices so overall I don't think the study has been hastily done.

Table S1 shows that although their individual tastes differ, they are remarkably consistent regarding which subset of violins (both new and old) to reject, e.g., N2, N9, and O12.

Overall, there doesn't seems to be strong statistical evidence to choose old over new violins in terms of utility. It is very encouraging to violin students and their parents that the best modern instruments can now compete with Strads and Guarneris.

April 8, 2014 at 05:39 PM · Once upon a time, when I was working as a reporter on a television program, I created a taste test for April Fool's day. My cameraman and I set up a table in a grocery store and filmed folks telling us which jar of peanut butter they preferred. The trick we played was that the two samples came from the same jar. But all the testers had strong opinions about the 'differences' that they tasted.

More recently, in my job as an audio engineer, a number of us tried a 'blind' test of microphones while trying to decide which model to buy. Several voice actors recorded the same material on each of five or six mics. Within minutes, we realized that there was no way we could make any kind of objective comparison. Too many sounds, too many nuances. We practically drove ourselves crazy.

Thanks for digging below the surface of the headlines. We should all do that more. If we did, we'd probably trust our newspapers less, though.

April 8, 2014 at 05:52 PM · I am not surprised by the results. A well made modern violin costs at least $30K to start and could go up to 100K, especially if the luthier is the winner of a major prize competition. The wood used on these "new" violins are top quality and usually aged for more than 50 years or so.

I would suggest a more meaningful test for the general violinist population; mainly for the intermediate level violin students and serious amateurs. Please compare the violins made by big name modern luthiers, that sell for more than $30K each, against properly tuned up handmade violins, in the $3k to $10k range. Please conduct the same test employing professional violinists.

A common question for the parents of promising young violin students is how much do I need to spend on a violin for my child so that he/she can continue her progress without the instrument itself being the hindrance/roadblock to success? I believe a high quality violin is still a requirement for the more technical violin music. The response and resonance of the violin needs to be "up there" in order for a difficult piece music to be properly executed.

The kind of answer I would like to see is if the test could show that beyond a certain price range, let's say $3k,the professional violinist would start to pick their violin of preference.

April 8, 2014 at 05:52 PM · You summed it up perfectly.

April 8, 2014 at 05:57 PM · This violin comparison event is currently being discussed on the computer geek forum Slashdot:

http://entertainment.slashdot.org/story/14/04/07/207236/elite-violinists-cant-distinguish-between-a-stradivarius-and-a-modern-violin

As is usual on Slashdot, the discussion frequently departs wildly from the subject in hand (entertainingly so at times) but for once on this occasion there is a small handful of musicians posting knowledgeable comments.

April 8, 2014 at 10:21 PM · Lots of wise comments here... yes, it would be very interesting to compare the big names of modern makers to those less known and even to cheaper chinnese instruments, and also do the same tests with good old and new bows. But meanwhile we should accept the results even if they crash with our previous belief. I guess the most important issue is that when people- great players in this case- don't try an instrument under the previous influence of name, fame, legends, origin and price, can much more appreciate it for its sounding qualities. As Elmar Oliveira says in the video, a great fiddle is a great fiddle, and that's it.

April 9, 2014 at 04:02 PM · Unfortunately even Stradivarius would not even be able to recognize the sound of his instruments today because all surviving Strads have been "modernized" with different necks, bass bars and strings, and played with greater string tension. Their intended, more intimate sound would be quite different and has been lost forever. Although they still have the Stradavarius name they don't sound at all like the maker wanted them to. So this comparison really isn't old vs new, it's modern vs modernized.

April 10, 2014 at 12:29 PM · I thought the test was well conducted and that perhaps some people's responses to it have been swayed by media coverage (that often missed the point of the experiments) and by the "belief" that Strad's are best.

The time spent with the violins was around 150 minutes for the first two parts of the experiment. Given that the participants said (before the test) that they'd need (on average) 50 minutes to evaluate a dozen violins, the time allotted seems reasonable. The test asked players which violin they preferred there and then and would like to take home for further evaluation. That is a reasonable facsimile of what would happen in reality, when looking for a new instrument. A player might take years to get to know a Strad, but not if they didn't initially like it's sound.

Of course that also brings up another question - do players get "more" out of a Strad because they believe it to be superior and so work harder at getting more out of it. It would be interesting to see what happens if the same effort were put into a newer violin. . .

The thirty second "guess the age" part of the study provides a snap shot of first impressions, which the authors agree may not correlate with long term satisfaction, but without a makers label for "guidance" the first impression might be the only playing a violin gets when being chosen for further investigation. I am also pretty certain that had the results demonstrated that the Strads were picked out more often than chance would expect, then there would be a lot of "see, see, they are stand out instruments." The fact that the instruments don't stand out immediately must irk those that assume they would and so they pooh, pooh this part of the trial.

@Laurie, the FAQs for the first study says that the violins were set up as their owners preferred them to be, so one would imagine that a Strad would be well set up with decent strings. As they go on to say, adjusting the instruments before the experiment would bring another variable (the adjuster's tastes) into the melting pot. So it is reasonable to assume that they instruments were set up well, for both experiments with whatever strings, sound post adjustment etc is deemed (by the owner) to suit the instrument. As the experiment is mimicking the process of choosing a violin for further investigation, then this is not unreasonable. Also if, for example, the same strings were used then there would be the criticism that such and such strings don't suit that Strad, it should have had string X. We have to trust that the owners (who according to the literature play the instruments) have them set up well, but of course to their preference.

@Scott - one of the violins was a few days old according to the FAQs for the first trial, so when they say "new" they're including freshly made.

@Paul - the trial was designed to try and determine whether Strads are (or aren't) "better" by how much they are preferred by soloists, as better violins are the ones that a soloist would choose and/or identify as being better. It is of course a subjective appraisal, but that does not invalidate the results. In many research fields subjective opinions can be collated to provide objective information. The fact that two trials have provided broadly the same result may be pointing towards some consensus.

I think the trial was well conducted, and provides useful evidence that Stradivari violins are not necessarily the be all and end all of violin making. Of course there is much pressure for the results to be wrong, or invalid - there's a lot of money and many reputations tied up in instruments, but that in itself does not make the results invalid, neither does the overstatement of the results by the media.

One final thought - "just 6 of the 10 soloists in the study preferred them" (new instruments). If the result had been the other way round might that sentence have read "over half of the soloists preferred the Strads"?

April 10, 2014 at 02:57 PM · Great post Liz.

April 11, 2014 at 02:22 AM · Ditto Liz.

April 12, 2014 at 01:42 PM · How many modern violins would you need to play before you found a great one? How many antique Italians? For a top-tier player, is it less of an effort to get one's hands on a great fiddle if one includes the moderns in one's search? Aren't great instruments rare, whatever your search parameters? The moderns are less costly, generally, but if you've got a patron would you still go for the Strad? Because, by definition, high market value greatly inflates the risk of any loss, are the gains worth the risks? And there's always that chance of getting TAZED.

April 12, 2014 at 03:32 PM · I strongly agree with Liz that the time parameters of the test seem reasonable. The length of time compares well to the amount one would spend considering alternative instruments to buy at a dealer or auction, and in my experience, one needs to switch fairly rapidly between instruments to have a memory of how to compare them. The overall length also seems comparable to the shopping experience. Admittedly, it is a first impression, but that is how one decides whether to e take an instrument home to try it in other halls, have others play it, etc.

I think it is also safe to assume the instruments were configured appropriately for their properties. It would have been extremely odd to go to such lengths to get important performers to compare many important instruments and then not bother to check if the strings were old or the bridges were misaligned! I also would expect they used their own bows.

What is not being pointed out in this Strad vs. New discussion is that if $30-100k new violins are facing up well to some Strads, then it is likely they often could blow away the "mid-price" ranged fiddles of the 19th century, which, as nice as they are, are often 3-15x more expensive than a good new violin. I would like to see a comparison of new fiddles to that category, rather than to the very greatest older violins.

April 12, 2014 at 06:44 PM · The recent trend on the part of foundations, festivals, and patrons to award great modern instruments to their winners makes good sense. It promotes modern luthiery and gets great instruments where they need to be. It also solves many other problems -- security, the high costs of insurance, customs issues, etc.. It's the way to go, for players and their supporters . . .

This entry has been archived and is no longer accepting comments.

Our Kokopelli
Please support Violinist.com
through your
one-time donation or
sponsorship campaign.

Violinist.com is made possible by...

Shar Music

Yamaha V3 Series Violin

The Potter Violin Company

Coregami Performal

Metzler Violin Shop

Connolly Music

Corilon Violins

Anderson Musical Instrument Insurance

Dimitri Musafia, Master Maker of Violin and Viola Cases

Fiddlerman.com

Fiddlershop

FlexTux

Heifetz International Music Institute

Long Island Violin Shop

Nazareth Gevorkian Violins

Pro-Am Strings

Wangbow Violin Bow Workshop