Jump to content
Maestronet Forums

How Good is Our Theory?


Michael_Molnar
 Share

Recommended Posts

All sounds good, but I think you have to decide whether you're making violins for players or for audiences. This is a fantastically difficult question as players play better on violins they like, and audiences in trials tend to try to listen to the tone rather than the playing!

Then you have to decide whether you're looking for violins for concerto soloists (for whom volume is the number one priority) or string quartet players etc. We are all subjected to the agenda of the concert soloist, yet this is like saying that all singing should use operatic technique.

So I think even proper double-blind trials are useless unless there is a very specific aim (for example "do professional soloists express a preference for Stradivaris when they don't know what they're playing).

I think blindfolds and some blocking of sense of smell might be necessary.

And then there are all the factors which make a violin feel responsive but don't objectively affect sound .......

Nightmare! No wonder we let antique dealers decide what's a good violin.

Link to comment
Share on other sites

  • Replies 148
  • Created
  • Last Reply

Top Posters In This Topic

Speaking as a luddite ......

Science focuses on the physical properties of a given violin (or of a statistical sample).

...a pursuit of physical properties seems to be counter-productive when aiming for playing qualities.

It's a jungle because no two "great" violins have anything like the same physical properties, and the potential for massive variation even in the same model is huge.

That's like looking at a MN thread on varnish, and criticizing makers for focusing on only a small part of the violin.

Science can be applied to any aspect of the violin; playability is not being ignored, but it is not the most visible of the investigations. There was some discussion this year at Oberlin about an experiment where damping in the audio feedback to the player of an electric violin had a perceptable effect on playability. And, as a player myself (loosely speaking), I continuously think of how the physical properties of the violin affect what the player experiences.

In any scientific investigation, it is normal to narrow the focus to just a few variables to see if anything can be determined. Anything learned is better than nothing learned, as long as one does not lose sight of the complete picture or the actual goals. If you learn something useful about varnishing, that's a good thing... but it doesn't mean you can ignore arching.

As has been said before, science and theory are only tools to be used for understanding. If used appropriately, can more understanding be a bad thing?

Link to comment
Share on other sites

These all are great posts. B)

I got the idea to ask this question after making my 25th violin and realized that I no longer use plate tuning à la Hutchinson. In fact, I wish I never heard of plate tuning. I think it set my making back a few years.

The other impetus for my question are the various papers and posts by high powered researchers who love to flaunt the differential equations. After reading their treatises :blink: I come away and still am where I started - OK, how do I make a better violin?

Now, I am not shooting down all research which can help understand the behavior of violins. I am not a Luddite convert. For example, I think some of the things we are getting from Noon and Buen are very compelling and can eventually help. But IMHO I think many well-meaning, smart researchers have lost focus. Noon and Buen maintain that focus. ;)

Stay Tuned.

Mike

Link to comment
Share on other sites

The very foundation of the scientific method is that you control all the variables you can identify, then change only one thing at a time and carefully measure its effect. You can rigorously apply the scientific method to one violin, but not across different violins. Oded and Jack Fry (RIP) advocate applying the scientific method to one violin and can measure the effect of various adjustments of thicknesses. On the other hand, trying to apply the scientific method to the thicknesses of violins in general will not succeed because you cannot control all the variables, density, longitudinal stiffness, crossgrain stiffnes, the whole tensor of twisting moduli, wood age, etc. A multiplicity of different values for the variables may give the same result. You may deduce some useful trends, but you are largely in the realm of declaring psychology a rigorous science.

Link to comment
Share on other sites

But IMHO I think many well-meaning, smart researchers have lost focus.

They might just have a different focus: research and publish. They are often not that interested in actually making violins, it seems. I too find it difficult to see any practical use for many of the papers heavy with analysis and equations, but occasionally see something in them that might be of some importance. More knowledge can't hurt... at least, until your head explodes.

Link to comment
Share on other sites

I ask again, how does one quantify human intuition, and how may one duplicate genius? The variables are legion, nonetheless the human mind is capable of sorting the lot (although some minds perhaps more so than others, I admit). As for our analysis, though, even the best of it is woefully rudimentary by comparison. In other words, it is little more than a distraction. Ask yourself, what is it you hope to duplicate? We all know the answer, of course (I hope...not sure all do, though), and so why apply a modern standard to a 300 year old paradigm? It just doesn't make any sense.

(And for those who may not know, what we aim to duplicate should not be an instrument in particular (the Titian Strad, for instance), but rather the essence of, which is a very different thing and the reason Stradivari, Guarneri and the like produced an original, every single time, even when made alongside another...and so we must as well, learn to think as they did, we must).

P.S. - One area, however, in which scientific analysis may yield some very positive and repratable results is varnish (ground coat and all). As for the remainder, though, it is an enigma.

Link to comment
Share on other sites

I ask again, how does one quantify human intuition, and how may one duplicate genius?

I think we all know the answer to that: you can't.

However, I also recall a phrase from Jim Ham: "educating the intuition". Someone may have an intuition about how a violin works (don't we all?), and after viewing the vibration animations on the Strad3D video, you would expect that the intuition might change, or be educated, about how the thing really works. The genius is to assemble all the bits of information and synthesize something from it. Even Einstein couldn't have done anything without a firm basis of facts about the way things really work. For example: the precise experiments showing that the speed of light is constant, and independent of the speed of the source or observer. Without that bit of hard evidence, he would never have been able to derive the theory of relativity. Even genius needs something to work with.

Link to comment
Share on other sites

These all are great posts. B)

I got the idea to ask this question after making my 25th violin and realized that I no longer use plate tuning à la Hutchinson. In fact, I wish I never heard of plate tuning. I think it set my making back a few years.

The other impetus for my question are the various papers and posts by high powered researchers who love to flaunt the differential equations. After reading their treatises :blink: I come away and still am where I started - OK, how do I make a better violin?

Now, I am not shooting down all research which can help understand the behavior of violins. I am not a Luddite convert. For example, I think some of the things we are getting from Noon and Buen are very compelling and can eventually help. But IMHO I think many well-meaning, smart researchers have lost focus. Noon and Buen maintain that focus. ;)

Stay Tuned.

Mike

You went further than I did, I've built about the same number as you have; when I first started I decided it was pointless to try and work with unpredictable variables. I proceeded to build based on a holistic, intuitive ( blind intuition) method, and I have no regrets. I did a lot of experimenting before deciding on the direction I wanted to go.

It seems to have been a steady progress, and every subsequent instrument seems to be a slight improvement over the previous. I'm thankful that I haven't experienced much regression which seems to indicate my feel for things is improving and my skills are becoming more finely honed.

Link to comment
Share on other sites

These all are great posts. B)

Agreed! Let me now ruin that record run!!! :o:lol:

I got the idea to ask this question after making my 25th violin and realized that I no longer use plate tuning à la Hutchinson. In fact, I wish I never heard of plate tuning. I think it set my making back a few years.

To me this is the single greatest contribution that Science has made to violin making is; that Science has, more than anybody else, eliminated the most 'smoking guns'.

Remember the furry caused by Bi-tri-octave tuning?

Then Science found other tuning 'patterns'.

You must have a particulate ground, or really light wood.

Well Scientific examination has shown other 'arrangements'.

Special 'Maundered' wood, with secret sauce, etc. etc., all and more have been refuted with the help of Science, either directly or indirectly.

So in short, Science has best answered the question of "How NOT to make a violin, better than anyone else."

The other impetus for my question are the various papers and posts by high powered researchers who love to flaunt the differential equations. After reading their treatises :blink: I come away and still am where I started - OK, how do I make a better violin?

Would we expect Science to answer "How do you get to Carnegie Hall?".

Surely we see this as a human endeavour, and do not seek an answer from Science.

So why do we seek an answer on Making?

Is it because we see making as somewhat less of an art form than playing?

Why?

Is it a 'Leave no stone not turned' approach?

I think a lot of people have expectations of Science that are way too high.

In short, just let Science do the things that Science can do best, and forget the rest.

Now, I am not shooting down all research which can help understand the behavior of violins. I am not a Luddite convert. For example, I think some of the things we are getting from Noon and Buen are very compelling and can eventually help. But IMHO I think many well-meaning, smart researchers have lost focus. Noon and Buen maintain that focus. ;)

Stay Tuned.

Mike

I think that people like Noon and Buen to mention a few, are laying some important foundations, for someone later to come along, and build on.

So until then, we are all left with the choice between ....

davinci-hax-shirt.jpg

CNC_violin3.jpg

Making for players ....

tumblr_lhuscaksTW1qzpxv2o1_400.jpg

Who play for Audiences of ....

weird-japanese-pictures-03.jpg

OR ....

'Man is the measure'....

pg000b-Mandala.gif

.... And so anything that aids Man must also be a subject to Man, and not the Master.

So in the end we end-up with ....

stradivarius1.jpg

Which leads to ....

paganini_nicolo_03.jpg

And then this ....

Royal+Albert+Hall.jpg

So I leave you with the words of one Brian Lisus:

"My Career: After leaving Newark I chose to specialize entirely on the making of new instruments and have been fortunate enough to have sold my instruments all over the world. (USA, Canada, England, Germany, Taiwan, South Korea, Chile, New Zealand, Singapore, Namibia and South Africa.) The most rewarding part of my career has been all the wonderful friends I have made. Nearly all my commissioned instruments have involved the musician, sharing in the whole process, choosing a name, personal set-up preferences, etc. I sometimes feel as if they are right there in the workshop (even if our only correspondence has been at the other end of the world via e-mail), as every instrument somehow turns out to have the exact quality of sound they desire. This is still a great mystery to me, as working more by feel than science one can not analyze this logically." - Link to Quote

As for me I can leave good enough alone, and let the 'Heart' do what the 'Heart' does best! ;)

Link to comment
Share on other sites

Salvador Dali - Sacrament of the last supper - 1955.

Dali's last supper abounds with the golden ratio (phi). Golden triangles make a pentagon and 12 pentagons make a dodecahedron. The scene is inside a dodecahedron. The violin's design seems to be closely related to the golden ratio or at least the simple ratios the approximate phi such as 2/3 , 3/5, 5/8 etc.. Why did he use dodecahedron?

post-24376-0-50316800-1325095131_thumb.jpg

Link to comment
Share on other sites

I think we all know the answer to that: you can't.

However, I also recall a phrase from Jim Ham: "educating the intuition". Someone may have an intuition about how a violin works (don't we all?), and after viewing the vibration animations on the Strad3D video, you would expect that the intuition might change, or be educated, about how the thing really works. The genius is to assemble all the bits of information and synthesize something from it. Even Einstein couldn't have done anything without a firm basis of facts about the way things really work. For example: the precise experiments showing that the speed of light is constant, and independent of the speed of the source or observer. Without that bit of hard evidence, he would never have been able to derive the theory of relativity. Even genius needs something to work with.

Good point (and I do realize one cannot quantify human intuition, nor copy genius...that was the point of the questions, to bring to light the absurd).

We each stand upon the shoulders of another (to quote Newton, loosely) and one serious benefit Stradivari, Guarneri, Amati (Nicolo and son), Bergonzi, etc. had is the time and place in which they lived. This fact cannot be defeated, nor ignored. The training they received was certainly second to none, and the environment an ideal one in which to practice their art (and I use this word, "art", deliberately).

Link to comment
Share on other sites

I got the idea to ask this question after making my 25th violin and realized that I no longer use plate tuning à la Hutchinson. In fact, I wish I never heard of plate tuning. I think it set my making back a few years.

Stay Tuned.

Mike

Great story, Mike.

Even so - and Hutchins lack of great success as a violin maker, look at the momentum still in existence.

Link to comment
Share on other sites

To me this is the single greatest contribution that Science has made to violin making is; that Science has, more than anybody else, eliminated the most 'smoking guns'.

I think a lot of people have expectations of Science that are way too high.

In short, just let Science do the things that Science can do best, and forget the rest.

I think that people like Noon and Buen to mention a few, are laying some important foundations, for someone later to come along, and build on.

As for me I can leave good enough alone, and let the 'Heart' do what the 'Heart' does best! ;)

Bravo!

[i just wish I understood half of what Buen and Noon write... ]

Re: Luddites, many science bashers today have (in previous topics) posted things like "I feel the response in the X-Khz range is important." Ain't that science? blink.gif

Double standard anyone?

I personally don't think CNC production is good science. Nor is relying wholly on plate tuning, but it will probably help identify a dud.

Re: Dali, the viewer is the 12th Apostle. wink.gif

Link to comment
Share on other sites

Great thread!

I had a long reply, but it looked way too tiresome and I deleted it in favor of a brief opinion statement:

Plate tuning: nope*

Chladni patterns: nope

Micro-tuning: nope

modal analysis: not yet**

wood properties: some help

... At the moment, the "trial and error" theory still looks very efficient.

... I got the idea to ask this question after making my 25th violin and realized that I no longer use plate tuning à la Hutchinson. In fact, I wish I never heard of plate tuning.

... But IMHO I think many well-meaning, smart researchers have lost focus. Noon and Buen maintain that focus. ;) ...

... To me this is the single greatest contribution that Science has made to violin making is; that Science has, more than anybody else, eliminated the most 'smoking guns'.

... Surely we see this as a human endeavour, and do not seek an answer from Science. So why do we seek an answer on Making?

... I think that people like Noon and Buen to mention a few, are laying some important foundations, for someone later to come along, and build on. ...

Science has given us a wealth of info, but few answers. Perhaps, as I think many have suggested, this is partly because violin making is at root a human art and not a science.

trying to get modern science to make a better stradivari is like trying to get jackson pollack or salvador dali to help you paint a better mona lisa or last supper ...

... Trying to achieve the Strad or Guarneri sound through immersion in the spirit of their time is, I believe, filled with at least as many dead-ends as any other route (or more so). ...

This I'm not convinced of. Dead-ends are part of any discovery method. In a sense, they don't really matter. What counts is that 'historic immersion' and 'hands on' yield a lot of good fruit.

As a new maker, learning now, and building my first instruments now, I find writings from the 'immersion' crowd to be helpful beyond all other sources: Sacconi, Darnton, Hargrave, et al.

The other source of supreme usefulness is direct reports and observations of the old instruments. In this, science plays a hand. The internet gives a wealth of access to sounds, pictures, words, and data. Science and computers enable our wonderful forum, and many of the tools that help us learn from the experience of others.

Only as a distance third have scientific studies seemed helpful in themselves. Yet, some of them are very helpful indeed.

Perhaps some of the weakness of scientific results is that science thrives on separating components of a problem to arrive at fundamental understanding, while the violin maker must be concerned with fitting the components together into an effective whole.

Science can play a part, but making is not equal to physics + chemistry + engineering. It is also part history, culture, and art.

For me, some of the most helpful scientific studies have placed information in a complex context. For example, some of the CT scan works present several coordinated maps giving thickness, density, and contour/elevation of the same instrument. To me, these are more helpful than any study that probes one of these separately. Similarly, there are some motion animations that show the coordination of movement through the whole instrument and the air motion for varied frequencies.

Link to comment
Share on other sites

hutchins simply reported what many others were noticing and knew already, that the front and the back of great violins were tuned to musical notes, not out of tune between notes, and the top and back tended to be seperated by a semitone or tone, i think that data is still true, her big mistake was thinking she could commission relatively cheap violins, fiddle with the tap tones, varnish them and make them sound like strads, in this respect she failed miserably, but it still doesnt discount the scientific evidence that the plates are tuned at least on many examples, what it does show, though is you cant take any cheap violin, tune the plates similarly and make it sound like a strad

Link to comment
Share on other sites

hutchins simply reported what many others were noticing and knew already, that the front and the back of great violins were tuned to musical notes, not out of tune between notes, and the top and back tended to be seperated by a semitone or tone, i think that data is still true...

I am not aware that this actual data has ever been presented, and I was a subscribing member of CAJ back in the taptone heyday. It is my impression that this was wishful thinking more than any actual data. Any data I have actually seen does not support this tuning hypothesis. The whole idea of the importance of free-plate tap tuning to very specific numbers is so scientifically (and empirically) flawed that it was one of the reasons I gave up on CAJ membership.

Link to comment
Share on other sites

this data is meticulously recorded on almost all unaltered old master violins, all you need is the intelligence to tap on them and record what notes they are, however mrs hutchins was only on to one tap tone per plate, ive recorded 5-6 tap tones on each plate on many violins, ive talked about it before, you probably just ignored me like you usually do, don, oh well

Link to comment
Share on other sites

I am not aware that this actual data has ever been presented, and I was a subscribing member of CAJ back in the taptone heyday. It is my impression that this was wishful thinking more than any actual data. Any data I have actually seen does not support this tuning hypothesis. The whole idea of the importance of free-plate tap tuning to very specific numbers is so scientifically (and empirically) flawed that it was one of the reasons I gave up on CAJ membership.

....now you're in trouble!

Link to comment
Share on other sites

Years ago, one of the people doing good work in computer intelligence (I can't now remember his name) displayed and explained some new work to a group of people who promptly dismissed it as "clever, but not intelligence". That was apparently the last straw for him, and he responded "goddammit, if a horse had just done this you'd say the horse is intelligent!"

From that came a sort of sarcastic rule of thumb now standard in the field: if it can be explained, it's not intelligence. Intelligence is, by definition in most people's minds, something that's not explainable. The corollary being that no computer can display intelligence. If a computer can do something, it's explainable and therefore the computer is not displaying intelligence.

I wonder whether a similar rule might apply to fiddlemaking: if it's being done today, it's not as good because, by definition, only Stradivari and GdG did it right (and possibly the Amati Bros and Stainer, if whoever's talking has had a dram and is feeling generous). The corollary being that all theory is worthless because any theory that can be reduced to practice is not the one true theory, since that was, of course, vouchsafed only to Stradivari and GdG (and possibly etc).

(My formal training was in social, clinical, and cognitive psych. Which are sciences despite the fact that the subjects of our study are constantly self-modifying, not unlike fiddles but completely unlike rocks, quantum electrodynamics, and similar)

Link to comment
Share on other sites

The violin : how to choose one, Honeyman, Wm. C. (William Crawford), 1845-1919.

Pages 62–79 contain:

The Acoustics of Violin Making by Frank Devoney.

Wow! 1893, plate tuning. Worth reading.

Yes, interesting it is. Full of the flowery, self-assured tone of the writing style of the time.

I'm glad we now have the secrets of the highly sought-after Frank Devoney violins, and don't have to put up with those over-rated Strads, made by some guy who apparently failed to grasp "the system:"

Even Stradivarius thought and thought till his life was nearly gone,

and failed, because the beauty of his outline is got at the

expense of sacrificing some of the best notes which the plates

demand.

Ummm... yeah. Huh?

Link to comment
Share on other sites

Ummm... yeah. Huh?

I didn't say the book wasn't yet another ridiculous 19th c. work on violins... rolleyes.gif

Just that there's a step between "plates give different notes" of c. 1800 and tap tuning of 1960. How much of Savart is nonsense as well? They had some answers, just not THE answer.

OK, now, raise your hand if you have never tapped or bowed a plate. That's what I thought. tongue.gif

Luddites! rolleyes.giflaugh.gif

Link to comment
Share on other sites

Years ago, one of the people doing good work in computer intelligence (I can't now remember his name) displayed and explained some new work to a group of people who promptly dismissed it as "clever, but not intelligence". That was apparently the last straw for him, and he responded "goddammit, if a horse had just done this you'd say the horse is intelligent!"

From that came a sort of sarcastic rule of thumb now standard in the field: if it can be explained, it's not intelligence. Intelligence is, by definition in most people's minds, something that's not explainable. The corollary being that no computer can display intelligence. If a computer can do something, it's explainable and therefore the computer is not displaying intelligence.

I wonder whether a similar rule might apply to fiddlemaking: if it's being done today, it's not as good because, by definition, only Stradivari and GdG did it right (and possibly the Amati Bros and Stainer, if whoever's talking has had a dram and is feeling generous). The corollary being that all theory is worthless because any theory that can be reduced to practice is not the one true theory, since that was, of course, vouchsafed only to Stradivari and GdG (and possibly etc).

(My formal training was in social, clinical, and cognitive psych. Which are sciences despite the fact that the subjects of our study are constantly self-modifying, not unlike fiddles but completely unlike rocks, quantum electrodynamics, and similar)

It's interesting that you mention this. One of my interests is unsupervised machine learning. One of my buddies is a psychiatrist who did some work in fMRI before that line of inquiry hit the proverbial wall.

I may be going half-cocked on this, but it seems to me there might be an experimental way to determine what might be the best instruments. You have a bunch of good players and a collection of instruments. The players are not told anything about the instruments beforehand (and they're blindfolded).

Whichever instruments light up the brain's fMRI the most would likely correspond to the superior instruments. You can also ask the players to give the instruments a score (without them seeing the fMRI). The scores could also be used to correlate the fMRI, to confirm that the most lit up brains corresponded to the highest scored instruments.

Then we'd have a basis for getting the scientific discussion started.

Link to comment
Share on other sites

I am a bona fide scientist and read the various theoretical proposals carefully and with great interest. I am also on TOBI-L.

My question is: How much has our various pursuits improved making a superior violin? For example, how has "plate tuning" helped to produce a violin competing with a Stradivari or Del Gesu? Show me one area that truly produces an improved instrument? Are we scientists/engineers just blowing smoke?

I realize that the Luddites will jump on this thread and try to turn this into a field day parade. So, I want my fellow engineers/scientists and experimenters to reply.

Stay tuned.

Mike

It's a great thread you started Michael.

Good science demands rigor,discipline and reflexivity...It also requires quite an anarchist spirit to think outside the box of given norms. Academic pursuits can sometimes take a life of their own where social criteria come into play in regard to what gets status/ published etc. My formal education ended with a BA in Social Anthropology where I studied the scientific methodology as well as things like the sociology of aesthetics or the fetishism of objects..( quite relevant to violins)..if anything it taught me to watch the ongoing debate as much as a social/political construct as a source of information and to grade all the information accordingly.

I try to learn every day...after 30years of recording tap tones there has been no magical bullet...some other things like arching, wood choice and varnish do seem to help as well as many set up things in the lives of my instruments or ones I maintain..

One of the first things we must look at is what kind of questions we are asking ourselves and trying to solve. The Hutchins articles I have show instruments that have no relation in arching to classic Cremonese...that was a blind spot.

Another problem is that makers try to make an instrument that 'sounds nice' in their amateur hands or the nearest ham player at hand. Or even worse they try to evaluate the instrument without the player being there. If we want to replicate the 'sound' of great instruments, a great player is needed to do the test because the great player is the only way to tell...If they are happy to play Paganini God Save the King, Bach, and some Mozart...Then it is a good fiddle and satisfies a lot of important criteria. If we have not worked with this kind of players on a very honest basis...no science will help and we are tilting at windmills.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    No registered users viewing this page.




×
×
  • Create New...