RonM;576112 Wrote: 
> Actually, anyone who has passing familiarity with concert violins knows
> that Dan is quite right.  There have been innumerable blind tests
> involving Stradivari, and fine instruments from Stradivarius and other
> makers up to the 20th C have been shown to be indistinguishable to the
> ears of masters.  Nicely summarized in the Wikipedia article at
> http://en.wikipedia.org/wiki/Stradivarius.

This is a fairly long post and I did not have time to get it up until
now, but I had spent several hours researching this issue before RonM
posted and feel I came across some pretty interesting information.  So,
for anyone on here with an iota of intellectual curiosity, I thought I
would bother to post it.  Obviously this is not aimed at RonM.  Just
quit reading now if you are reading Ron, and go back to believing you
already know everything.  Honestly, though, starting your posting by
saying anyone with "passing knowledge" knows something?  What's that
old saying?  Oh, yeah: a little knowledge is a dangerous thing.  Pretty
apt here in my opinion.

First off, the notion that all "good quality" violins are so close in
sound quality (timbre) that it is reasonable believe that they would be
indistinguishable is easy to disprove with objective measurements. 
While I will certainly grant that audible differences due to cables
care slight, this is not the case with acoustic instruments.  Sure you
can find pairs of instruments that are very close in tone color, but in
general if you randomly pick two "good quality" violins they are going
to have distinct sonic signatures.  The single best site that I found
on this subject is:
http://www.schleske.de/en/our-research/introduction-violin-acoustics.html

The site is somewhat difficult to navigate, but not only does it
explain the basis for sound differences in great detail, it provides a
significant number of examples of actual measurements showing sound
differences between high quality violins.  For example:
http://www.schleske.de/en/our-research/introduction-violin-acoustics/sound-analysis.html

I also recommend reading these two papers:
http://www.schleske.de/fileadmin/user_upload/doc/CAS_Empirical_I.pdf
http://www.schleske.de/fileadmin/user_upload/doc/CAS_Empirical_II.pdf

RonM provides a link to a single Wikipedia article to "back up" his
beliefs.  He claims it is a good summary, but when I had read it I
judged it one of the weaker Wiki articles that I had ever consulted. 
RonM claims there have been "innumerable" blind tests but doesn't cite
a single one.  What about Wiki?  It claims that there have been
"many:"

> However, the many blind tests from 1817 to the present (as of 2000) have
> never found any difference in sound between Stradivari's violins and
> high-quality violins in comparable style of other makers and periods,
> nor has acoustic analysis.

So, "many" blind tests.  Just how many of these many/innumerable tests
are listed?  Two!  And here I thought that innumerable meant so many
that it is not possible to list them ALL--not that you don't actually
know any and so cannot cite anything.  :)  Wiki cites some book on
violins for the above line, but it obviously isn't possible to make any
judgments about the veracity of the book's claims since it is not
available online, and it is certainly not possible to make any
judgments about the reliability of any of these supposed many tests. 
(I note the book's author has no science credentials, being a composer,
so is not in a position to be evaluating listening test validity
anyway.)

So what about the TWO tests that are listed?  One was done for a BBC
Radio program in 1974 (some sources say 1977!).  Wiki describes it as
particularly famous," so one must assume that it had a big impact on
this "conventional wisdom" that particular violins cannot be
distinguished from other good violins.  Also this is the SOLE test that
I could find any reference to that involved actual concert violinists,
despite this being a key element of Dan's claims about these tests
(plural).  Most online descriptions of the BBC test are very sketchy. 
The best description I could find is here:
http://www.thestrad.com/Article.asp?ArticleID=1647

> Another example of a listening test (again, single-blind) was arranged
> and broadcast in the mid-1970s by the BBC (see Fiddle Faddle). In this
> case the quality of the judging panel was high - Charles Beare, Isaac
> Stern and Pinchas Zukerman. The instruments played were a Stradivari,
> a Guarneri del Gèsu, a Vuillaume and a Ronald Praill that was barely a
> year old. Two excerpts were played by Manoug Parikian: the start of
> the Bruch G minor Concerto and a segment of the Bach Chaconne.
> 
> Before giving their answers, the judging panel spent some time
> pointing out many of the deficiencies in the testing procedure. The
> two excerpts played were too short and limited in tonal possibilities,
> there was no chance to revisit each instrument for extended
> comparisons, the studio represented only one of many possible
> listening environments, and so on; all very valid comments as the test
> was undoubtedly far too limited. However, the good-natured panel
> proceeded to give out their judgements, many of which proved to be
> incorrect - the Praill was mistaken for both the Strad and the
> Guarneri (Beare and Stern did the best with two out of four correct).

So this is the particularly famous test?  (In fact Science Daily, in
the article linked from Wiki, refers to it as "the most serious of all"
such tests!)  Three subjects, given one chance to correctly identify
four violins, under less than optimal and certainly not reproducible
listening conditions.  Wow, I guess that sure proves that no concert
violinist can ever ID a particular violin.  What an absolute joke!  Of
course it should be plainly obvious to anyone with half a brain that
this test is far too limited to establish any result with any useful
level of confidence.  Yet this is a "particularly famous and serious"
example of these kinds of tests?  Yeah, boy, real science at its best. 
NOT!  Of course it was never billed as a scientifically valid listening
test.  It is people like Dan that make that sort of totally bogus
claim.

Now Dan also claims that the results of these kind of tests is "always
at the level of pure chance."  Is that true here?  If I did my
caluclations correctly, choosing totally randomly (unable to
distinguish instruments at all), the probability of correctly
identifying <2/4 violins is nearly 3/4.  Yet here 2/3 of the judges
correctly identified 2/4.  So while the sample size is way too small to
draw reliable conclusions, if anything it contradicts Dan's contention
about chance.

Now on to the next of the many, er, yes really only two tests.  Wiki
refers to it as the Biotech violin blind test.  It was conducted just
last year.  Unfortunately, the descriptions of this testing are again
rather sketchy, but it definitely did not involve famous concert
violinists (it was held at a forestry conference).  The one result that
is clearly spelled out in all accounts is that in a blind test of five
different violins (including one Strad), 90/180 listeners all rated the
same violin as "the best" (it was not the Strad but a new instrument
treated with wood fungus!).  Now hopefully the Wiki article was
written/edited by multiple people because this description of 90/180
people selecting the same 1/5 violin comes almost immediately after the
above sentence claiming that  "the many blind tests...have never found
any difference in sound between Stradivari's violins and high-quality
violins." A bunch of us had quite a good laugh at that contradiction. 
Never found any difference in sound, yet 90/180 selected 1/5 violins. 
Right!  Sorry again, Dan, not even close to "always at the level of
pure chance."

The other problem with the "fungus violin" test is that it was not a
test to try to see if listeners could DISTINGUISH different violins, it
was designed to determine which violin's sound was PREFERRED.  In fact,
that is what virtually all of tests that I was able to find described
on the web involved (I was able to find descriptions of only about
three other tests).  I would have no problem if Dan had claimed that
people didn't always PREFER Strads under blind conditions, but that is
not the claim he made.  He claimed they could not distinguish ("tell
the difference") them under blind conditions.

So that is a "quick" summary of what I found when actually bothering to
look into the subject of violin listening tests instead of just taking
Dan's word.  Until somebody can present some significant additional
evidence about other tests, I will consider Dan's claims about violin
testing extremely misleading and inaccurate.


-- 
ncarver
------------------------------------------------------------------------
ncarver's Profile: http://forums.slimdevices.com/member.php?userid=15905
View this thread: http://forums.slimdevices.com/showthread.php?t=81616

_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/audiophiles

Reply via email to