Jump to content
Maestronet Forums

Newly Published Book "The Sound of Stradivari"


pt3
 Share

Recommended Posts

Performers have given up looking at modern instruments because they are tired and recognize the futility of looking for diamonds in the proverbial garbage heap....

It seems that some performers aren't aware of this. ;)

...and makers are "tired of plate tuning cause that's been done and hasn't proved that fruitful.

It seems that quite a few successful makers aren't aware of this.

Link to comment
Share on other sites

8 pegs I thought I was having a flashback! What are you using to generate this image?

Yes, it is a Chinese made regraduated "Hardanger fiddle". It ended up having nine pegs, so there are 5 understrings. The picture is made using George Stoppanis modal anaysis programs.

The dots seen in the picture is the grid used for the measurements. An accelerometer is fixed in one position while an impact hammer is used hitting the grid points in a predetermined sequence. Then the set of admittance curves for each grid position are used for a calculation of the mode shapes through a fitting procedure.

Link to comment
Share on other sites

Joseph Curtin has shared some measured data from some fine old italian instruments in the Strad. I have these in a statistics file here along with a few data entries from some top contemporary makers. I enclose histograms for B1- and B1+ frequency data. The blue are the old italians + some 6 top contemporary instruments and a Vuillaume, 44 in all. You can forget about the rest, they are violin and Hardanger fiddle data I have collected. The numbers given are how many instruments have B1- or B1+ resonance frequencies within the given histogram bar, each with a 10Hz width around e.g. 430Hz, 440Hz, 450Hz etc.

It seems as the old italian B1- data mainly are within the 435-455Hz range, 27 out of 44 instruments are here, some 61% of them. a4 is at 440Hz, a#4 = 466Hz and g4 = 415Hz, so the B1- are close to open a in these instruments.

Now, for the B1+ data, there seem to be an overrepresentation in the 520Hz and 540Hz 10Hz histogram bands with a "dip" in the 530Hz band between them. There are 9 instruments in the 520Hz band, 2 in the 530Hz and 15 in the 540Hz band. There are some in the 550Hz and 560Hz bands as well, 9 in all. One may wonder why is there a dip in the region between 525-535Hz?

c5 = 523Hz, c#5 = 554Hz and d5 = 587Hz so the highest representation is in the 10Hz band below the c#5, but the next highest one is around c5. Usually the A0 lie just below or at c4 = 277Hz. The B1+ can give a wolf sometimes, so has there been a tendency for fixing wolf notes on these instruments? Or why are there two tops and a dip between them? How usual is it to play using quarter notes in classical music?

post-25136-0-39281600-1303393763_thumb.jpg

post-25136-0-75358200-1303393776_thumb.jpg

Link to comment
Share on other sites

bill, by tapping on it and youre right not every 100 yr old violin is tuned but most of them are, including even some of the cheapest crappiest ones, surprisingly a lot of regraduated violins are tuned by the regraduater, at least when it was done in the 1800s, and a few modern makers are tuned, even though they claim ignorance to the tuning, maybe its one of those secrets they don't want to let out, any one who doubts it and has access to 150 200 yr old violins, original thicknessing, i can show you how to hear the tuning if you call me at 909-793-8506, tuning new violins is not easy, takes a lot of patience and takes a lot more time than the physical regraduating but basically on the top its tuned G d a e1 and on the back often A e b f#1 as well as where the bridge feet contact the top and the soundpost contacts the back

Assuming what you say is fact, this would indicate that the makers prior to 1900 or so had it all figured out, and we have regressed since.

How come the number of great violins compared to mediocre or student grade violins is so low from this time era?

Link to comment
Share on other sites

I guess the book is to some extent about stability and sensitivity in instrument making. How to increase the consistency in making good instruments? In the data I have seen from good contemporary makers it seems as if they are very consistent and probably more so than e.g. Strads or del Gesus that has been taken care of in different shops through the years. One may wonder why the consistency seems to be so high. Some makers use the same log every time, others use(d) a "thickness regime", some may use a "tap tone and weight regime", yet another a narrow model range or combinations, there are probably lots of other examples.

I enclose a historgram of the difference between B1+ and B1+ mode frequencies for the set of now 46 old Italians in blue. The vast majority are within the 80Hz-110Hz range. Now I have a hypothesis or an intuition, that the distance between these modes to some extent can be determined by how the top and back plates are matched regarding their tap tones. (I already know that the back plate thickness play an important role here, but to tease the "anti tappers", and to follow the tread, i'll stick to the tap tones now)

From the regression formulas or lines based on Don's and Wilkins data from yesterday we may extract some useful information regarding where to end up with the B1 modes and their frequency distance based on their free plate tap tones. This is not exact, there are some fuzziness involved, but anyway.

If we want the B1- to end up around open a (440Hz), the mode 2 tap tone of the free top should end up around 155Hz. From Curtins analysis of free tops of some old italians we know that the mode 5 of the free top will end up at about 2.3 times the mode 2 frequency. (And the average mode 2 frequency was found to be 154Hz). So the top should not have tap tones below about 155Hz*2.3 = 356Hz.

Lets say the distance between B1+ and B1- should be around 93Hz (average in the set of 46). Then we need a formula to caluclate the effect the back plate has on that difference when the top is fixed around the given data above. Regresseoin formulas from yesteredays figure give:

B1+-B1-

= 0.594*B5 + 334-1.13*T2-264

= 70 + 0.594*B5 - 1.13*T2

Cremonese: B1+-B1- = 93Hz

=> 93-70=0.594*B5-1.13*155

=> B5=198/0.594 = 333 Hz

So the back plate should then have a mode 5 tap tone around 333 Hz to get the B1 modes right.

If we use 444Hz as the aim for B1- then the top should be around T2 = 160Hz and T5 = 368Hz and the back end up around B5 = 343 Hz

This is not exactly mode matching of the plates!

Now lets assume that the plates has been mode matched a-la Hutchins so the free top and back are at the same frequencies for mode 5, but that the cremonese 2.3 relation between mode 2 and mode 5 frequencies in the tops are used so the graduations become classic (and not thicker in the centre). What happens?

Ok, we start with the top mode 2, T2, at 160Hz and T5 at 368Hz so we get B1- around 444Hz. Then the B1+-B1- distance will be predicted to be:

= 70 + 0.594*368 - 1.13*160 Hz = 70 + 219 - 181 Hz = 108Hz that is in the upper scale of the cremonese range. B1+ should be around 444Hz+108Hz = 552Hz, close to c#5.

If we do it with octave matching a la Hutchins, but still stay at T2 = 160Hz (keeping B1- at 444Hz), we get: T5 = 320Hz and B1+ at 523Hz (B1+-B1- = 79Hz)

Then finally using Hutchins advisory T2 = 180 Hz and T5= 360Hz. Then B1- will end at some 468Hz and B1+ at some 548Hz (B1+-B1+ = ca 80Hz)

[Edit] Hutchins SUS295 has B1- at 466Hz and B1+ at 546Hz extracted from data presented in Wangs PhD taken my Kenneth Marshall from his modal analysis of that violin. Pretty close to my predicted values above.

The conclucion is that matching the free top and back plates a la Hutchins will tend to give a too high B1- freqeuncy and a too stiff top plate. B1 mode relations closer to the cremonese instruments will probably require a top that has a somewhat higher mode 5 tap tone than the back plate.

post-25136-0-89877700-1303405839_thumb.jpg

Link to comment
Share on other sites

The real note of quality is when musicians have the wisdom and integrity to buy something based on their OWN standards and recognition.

So modern day players have too-low standards, in addition to not enough wisdom & integrity? I think the only thing you left out is MONEY!

Link to comment
Share on other sites

My former post was a mess to read, so I have made two figures illustrating my point graphically. Using Hutchins recommended free plate tap tones will give a high frequency B1- mode and to some extent high B1+ modes (red lines). "Cremonese tap tones" around say 160Hz for mode 2 of the free top and some 345Hz ish for the free back plate mode 5 will in a set of many instruments give average B1 modes in the "Cremonese region".

I will return to what happens to the low frequency output of a violin with a top mode 2 as high as 180Hz as compared to the cremonese-ish 150-160Hz. The fiddle will become shallower sounding.

post-25136-0-35097500-1303410151_thumb.jpg

post-25136-0-30853600-1303410164_thumb.jpg

Link to comment
Share on other sites

Don't get me wrong... I think plate tuning is a useful tool, put not the panacea it is often represented as being.

One does get tired of newly discovered secrets of Stradivari, which people have been discovering since when, 1820-ish? They may be secrets, but not THE secret.

As for wholistic tap tuning, why is soundpost wood not important, but I should tap tune my endpin? rolleyes.gif

Strad varnish is important, but if you believe everything you read, it is only important between the f's, and his groundlayer is a glue-oil emulsion, because that's what the researchers wanted it to be. rolleyes.gif

Since an earlier post showed that this new book contradicts important data (#37) I think I'll pass.

Link to comment
Share on other sites

I really don't know how you got all that from what I said! Perhaps that's what you think, yourself, then?

I don't see it either Michael. But one thing I've learned in this business is that people almost always start with a preconceived position that is very, very hard to shift.

Link to comment
Share on other sites

I really don't know how you got all that from what I said! Perhaps that's what you think, yourself, then?

I was simply making the point that in the cases of both Luis Biava and Mr. Kreit the secret to success and fame lies in finding a niche in the marketplace for which there is a crying demand. In the case of Luis, that niche is inexpensive "rain" instruments that can be endorsed by lumenaries in the profession. In the case of Mr. Kreit, it seems to be a modern instrument that sounds like a Stradivarius.

Sorry if I mis-understood. Perhaps if your response were less abstruse?

Link to comment
Share on other sites

Sorry if I mis-understood. Perhaps if your response were less abstruse?

I can't help but notice you joined this forum just slightly over a month ago. Unless you've been lurking here long, you probably haven't figured who's who in these parts.

Suffice to say Michael has well over 10,000 posts on this forum. Do a search on "blind test" and "Michael Darnton." You might learn something.

Link to comment
Share on other sites

I can't help but notice you joined this forum just slightly over a month ago. Unless you've been lurking here long, you probably haven't figured who's who in these parts.

Suffice to say Michael has well over 10,000 posts on this forum. Do a search on "blind test" and "Michael Darnton." You might learn something.

I very well know of Mr. Darnton's fine reputation, and suffice to say we pro'ly have 50 common acquaintances. Perhaps it's simply a case of ships passing in the night, or "humour" that doesn't translate well in cyberspace. Maybe it's the case that players can't have opinions contrary to the received wisdom here, in which case I am truly in the wrong place. I meant no offense, and maybe from a player's perspective I just perceived the conversational course of the thread differently. I've been doing my job longer than most of the people here (hence my screen name...), but if you think you already know everything you find interesting just let me know.

Link to comment
Share on other sites

Maybe it's the case that players can't have opinions contrary to the received wisdom here, in which case I am truly in the wrong place.

It's not that players can't have "opinions contrary to the received wisdom here" - as the case may be. It is more that any opinions (offered by anyone) must be percieved as valid or, at the least, be based in common sense or logic of one sort or another - in order to be considered.

If you want your opinions accepted because of who you are, instead of what the opinions are - then you probably are in the wrong place. I don't really see that you've been victimized here, or singled out or excluded because of your opinions "as a player".

Link to comment
Share on other sites

I don't really see that you've been victimized here, or singled out or excluded because of your opinions "as a player".

In fact, I (and many other makers here) rely specifically on the opinions of accomplished players, in order to perfect my (our) craft.

On the other hand, I don't want the opinion of any and all players - a lot depends on their astute and unbiased observational skills, and their temperment...

Link to comment
Share on other sites

I have a confession to make regarding the speculations around B1+-B1- delta in my post #81. Part of the argumentation there are based on a combination of two regression formulas, one between T2 and B1- and one between B5 and B1+. I have been working a bit more on the subject where I simply tested how good that combined regression formula is at predicting the measured B1+-B1- delta in Dons and Wilkins data, and it basically sucks at it, see the attached figure. If it was good at predicting the B1+-B1- delta, it would have had the data points along a more or less straight line, but this is basically a "gunshot".

The regression formulas for B1- and B1+ alone are probably sound, but not when combined. - A new lesson learnt!

I think the argument that Hutchins type high mode 2 (T2) plate tuning tend to give high B1 modes still holds. I am not quite sure if the second graphical illustration in post #83 is ok or not in combination with the first. It is made combining two regression lines too, only graphical, so maybe it will be bogus too.

Wilkins data are from five instruments where he has matched to top and back plate modes a-la Hutchins I guess. They are the rightmost data points in the regression figures in #83. I guess Dons data are from a mix of processed wood and some normal wood instruments.

post-25136-0-53128800-1303556859_thumb.jpg

Link to comment
Share on other sites

Don't get me wrong... I think plate tuning is a useful tool, put not the panacea it is often represented as being.

One does get tired of newly discovered secrets of Stradivari, which people have been discovering since when, 1820-ish? They may be secrets, but not THE secret.

As for wholistic tap tuning, why is soundpost wood not important, but I should tap tune my endpin? rolleyes.gif

Strad varnish is important, but if you believe everything you read, it is only important between the f's, and his groundlayer is a glue-oil emulsion, because that's what the researchers wanted it to be. rolleyes.gif

Since an earlier post showed that this new book contradicts important data (#37) I think I'll pass.

Funny but I was thinking exactly the same thing. In other words, if a maker was able to put together all the parameters that were, at one point or another over the last century, reported to make the so called secrets of a Stradivarius, this violin would be dangerous to play as it would be SO GOOD that people would drop dead at hearing the first note... :D

Link to comment
Share on other sites

...if a maker was able to put together all the parameters that were, at one point or another over the last century, reported to make the so called secrets of a Stradivarius, this violin would ...

... probably sound like all the other instruments made by that maker.

There's something about discovering "secrets" that reminds me of diet pills. A simple, immediate, wowie solution... even if totally bogus... seems much more attractive than: "get more excercise, consume less calories". Similarly, magic varnish, magic math, golden ratios, and and similar hoohah seem to be so much more intriguing (and give instant results) than: "use really great wood, build something reasonably decent, age 300 years". My opinion of the "secret", anyway. Maybe I should write a book and sell it.

Link to comment
Share on other sites

... probably sound like all the other instruments made by that maker.

There's something about discovering "secrets" that reminds me of diet pills. A simple, immediate, wowie solution... even if totally bogus... seems much more attractive than: "get more excercise, consume less calories". Similarly, magic varnish, magic math, golden ratios, and and similar hoohah seem to be so much more intriguing (and give instant results) than: "use really great wood, build something reasonably decent, age 300 years". My opinion of the "secret", anyway. Maybe I should write a book and sell it.

Don,

Don't throw out the baby with the water. There is still much to learn about the classic making methods before we know all there is to know.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...