Carving site recommendations for plate tuning

[Bits4Waves]

Greetings,

In a recent talk with a Luthier friend, he shared with me that plate tuning could get very tedious, because of the difficulties at each step to know exactly where to carve so as to obtain the desired impact on the final frequency.

Reading some papers, I came across a very interesting one, by Alan Carruth, a former student of Carleen Hutchins (attached carruth1991plate.pdf). In this article, the author gives a diagram with the impacts that carving might have on several different regions on the plate.

I then compiled this information in a spreadsheet, where you can input the desired final frequencies for the plates, and then feed it with the current ones. Based on this information, the spreadsheet will show diagrams with the regions that could be carved in order to achieve the final desired frequencies.

My friend liked it so much, that he encouraged me to share it to the Luthier community, so I'm sharing here with you.

Please feel free to post any questions, criticism, and any feedback is appreciated.

Thanks for your attention!

carruth1991plate.pdf

[uncle duke]

The best plate tuning demonstration I have ever seen is from the old Carlisle shop circa 1900"s.

[Don Noon]

5 hours ago, Bits4Waves said:

... plate tuning could get very tedious...

Yes, it can... and since there is no convincing evidence that particular free-plate frequencies are superior to others, I don't bother with all of that.  If someone proposes "desired frequencies", look for the research and evidence that was used to come up with them.  Good luck.  FYI, my personal method (at the moment) is a combination of caliper measurements, weight, absolute bending stiffness, and (yes) taptones... but mostly M5.  I have no formula to decide anything, just some vague idea of the zone.  And for violas, I have no good idea of the zone, and just guess.

There are about as many methods of graduation as there are violinmakers, with good and bad results that are seemingly unrelated to the method and more related to the number of years of experience and the innate talent of the maker.

 

[Nick Allen]

4 hours ago, Don Noon said:

Yes, it can... and since there is no convincing evidence that particular free-plate frequencies are superior to others, I don't bother with all of that.  If someone proposes "desired frequencies", look for the research and evidence that was used to come up with them.  Good luck.  FYI, my personal method (at the moment) is a combination of caliper measurements, weight, absolute bending stiffness, and (yes) taptones... but mostly M5.  I have no formula to decide anything, just some vague idea of the zone.  And for violas, I have no good idea of the zone, and just guess.

There are about as many methods of graduation as there are violinmakers, with good and bad results that are seemingly unrelated to the method and more related to the number of years of experience and the innate talent of the maker.

 

Such an inexact craft at times, violin making.  I've taken to just flexing the plate about in my hands. I'll weigh it too, and note the density, if I remember. But I haven't made enough violins to make any kind of authoritative statement. That's just what vibes with me and has had some nice results. I'm sure that my working method will produce a turd once in a blue moon, but hopefully will one day produce consistently good violins. 

[HoGo]

I thought current trend is predominantly about getting good arching shape and then thickness towards optimal weight/stiffness measured whatever way the maker chose (bending in hands or taptones). Experience is the key here. After making dozen(s) of violins consistent way some relationships appear and new instruments are more consistent around desired outcome. If they don't appear then either the maker has no talent for this or he needs to change the way he works or evaluates his results.

Tap tones/tuning of plates alone sound like magic wand but they've been around for many dacades now and we still don't see everyone building great instruments tuned to the perfect pitches. There is more than just the taptones behind work of a master.

[Violadamore]

Welcome to MN!!  Thanks for posting, but I already have effective methods of my own. 

[Goes back to, by the light of flickering torches, using a 13-luk keris to draw strange characters across what appears to be a singularly colorful and complicated Tibetan sand painting surrounding an obviously decrepit violin, while chanting Spell 151 from the Book of Coming Forth by Day, backwards.   Somewhere in the moonlit forest encircling the ancient mound, long predating the Spanish conquest, wolves begin to howl...........].   

[Peter K-G]

16 hours ago, Bits4Waves said:

Please feel free to post any questions, criticism, and any feedback is appreciated.

carruth1991plate.pdf 464.03 kB · 25 downloads

Consider this:

1. It's too late to aim for M5 and M2 if you haven't started when arching the outside

2. You most likely have  15 - 30 Hz higher modes after a couple of years if you take the violin apart and measure again.

3. Modes vary with humidity and thereby moisture content in the wood +/- 15 -to 25 Hz

[Marty Kasprzyk]

It can be frustrating to make something that sounds good and then not being able to duplicate it again.  Attached is a summary of some of the different plate thinning strategies for achieving better consistency.

But there's also some disagreement of what "good" is just like many things in our lives--it's a matter of personal taste.  So it's not possible to please everybody.  One building strategy could be to try to make something in the middle of the road and actually accept poor consistency.  The scatter in the distribution curve might better appeal to the distribution curve of people's taste--more people could be satisfied by having more choices.

"Any customer can have a car painted in any color that he wants so long as it is black."

Different plate thinning strategies 7_25_2019 .pdf

[Bits4Waves]

As we are talking using the company's account here, we will proceed in this somewhat more professional approach from now on. This choice is made to clarify our motivations an positioning towards our goal here to share a tool with the Luthier community.

First, we would like to thank you all for the quality feedback! Our mission is to serve the community, and your feedback is our greatest guide in providing the most practical and efficient tools.

First of all, it's important make it upfront clear that we are not Luthiers, nor Scientists, but simply software developers.  Our main purpose is to provide tools that facilitate the practical application of research advancements in the construction of fine string instruments.

This tool in particular is based research results on plate tuning from the past decades.  A good summary of some of the main results was provided in A Retrospective on the History of Stringed Instrument Acoustics and on Plate Tuning Techniques, from the research papers by the Catgut Acoustical Society.

As stated before, our aim is to serve the community with scientific-based software tools. Thus, we feel to be important to address some of the issues that were raised.

In his article Tap Tones and Weights of Old Italian Violin Tops, Joseph Curtin asserts that

Quote

The tap tones of Old Italian violin tops and backs have long interested makers and researchers, both as clues to how the classical makers graduated their instruments, and as guides for graduating instruments today.

He then provides the measured frequencies for modes #2 and #5 of nine Old Italian violins, as can be seen on the image below, extracted from the article.

He cites the paper Stradivarius Plate Tests, by Carleen Hutchins, that she motivates with

Quote

This study provides measurable documentation of the mechanisms involved in the tapping and flexing long used by master violin makers as they listen and feel for desirable qualities in each pair of free top and back violin plates before assembly.

She also points out there that these results are

Quote

based on a detailed analysis of the lowest six normal modes in over 200 plate pairs as they have been carved, tested and adjusted during the making of violins, violas, cellos and the new instruments of the violin family, with good tone and playing qualities.

She adds that tests on "the best violins made thus far" suggest that the tap-tone resonance mode #5 should

Quote

have high amplitude, low damping and be at the same frequency in both top and back free plates, with the actual frequency between 360 and 370Hz. Also plate thicknesses should be adjusted so mode # 2, which usually lies somewhere near an octave below mode #5, should be an octave below and at exactly the same frequency in the two plates.

Sie Anton in Comment on the "Double Octaves" Tuned Violins, asserts that "ten finely crafted violins have been made consistently using 'Double Octaves' matching plates tuning", sharing considerations that were

Quote

summarized from a considerable number of professional violinists, teachers and advance students, who more or less have carefully played and tried the instruments.

He presented there a correlation between the players' preferences and the frequencies of modes #2 and #5, in the form of a table, reproduced below¹.

A later article by Carleen Hutchins, Some Notes on Free Plate Tuning for Violins, Violas and Cellos, summarizes this results in the table reproduced below.

We would like to finish by citing excerpts from Carleen Hutchins' article "The Acoustics of Violin Plates" (attached hutchins1981plate.pdf), where she establishes parallels between the manual bending and tapping of the plates that has been done for centuries and recent research findings:

Quote

These plate resonances, or normal modes, are created by the physical properties of stiffness and mass, which cause standing-wave patterns to be formed in response to vibration at discrete frequencies unique to each plate.

Quote

The modes that have so far been found to be most important in tuning violin plates are 1, 2 and 5. Mode 1 entails a twisting of the plate, with one corner up and the other down in opposite phase. Thus when a violinmaker holds a plate at each end, twisting it between his hands to feel its resistance, he is actually sensing the main stiffness characteristics of mode 1.

Quote

When a maker holds one end of a plate in both hands with thumbs on top and fingers spread out underneath across the wood, squeezing it and bending it slightly to assess the cross-grain stiffness of first one end and then the other, he is comparing the relative stiffness of mode 2 in the two ends.

Quote

When a luthier holds a plate around the two ends in his fingertips and pushes down in the middle with his thumbs, he is actually checking the principal stiffness of mode 5. The same test can be made by holding the plate around the edges and gently pressing the top of the arch against a flat surface to feel the bending.

Quote

Holding the plate at the midpoint of one end and tapping with the soft part of a finger around the upper and lower edges will activate the sound of mode 1 quite clearly, because the holding point is a node and the curves of the upper and lower edges are antinodes for that mode. Holding at one of the four points where the nodal lines of mode 2 intersect the edges and tapping on the antinodal area near the midline of either end of the plate activates primarily mode 2. Holding at a point along the nearly oval nodal line of mode 5 and tapping in the centre of the plate causes the sound of mode 5 to predominate.

To conclude, we would like to repeat the words of Carleen Hutchins

Quote

It cannot be emphasized too strongly that before one can apply such methods one must learn the violinmaker's craft, so that the basic instrument is built according to the principles of fine violinmaking. I learned violinmaking in the 1950's under the tutelage first of Karl A. Berger and then of Simone F. Sacconi with Rembert Wurlitzer's encouragement. It was eight years of slow, painstaking work.

Our mission as a software company is to do our best to help the community, by making it more practical and straightforward to apply the state-of-the-art research findings in the making of fine instruments. Thank you all for the attention, and please give us all the feedback and criticism!

Thanks to @Violadamore for the kind message!
 

---

Footnotes

1. The author calls the modes #2 and #5 "X" and "O", respectively. This is related to the fact that some authors (see attached carruth1991plate.pdf in the original post)

 

Quote

prefer to give them descriptive names like 'T', 'X', and 'O' [...] because the [Chladni patterns of the] modes assume different shapes in other instruments or occur in a different order

 

hutchins1981plate.pdf

Edited September 3, 2020 by Bits4Waves
clarify the approach

[Shunyata]

I started with certain faith that plate frequencies had to tell you something useful and waived aside every suggestion the the contrary. I have renounced that faith.

I just built two violins with wildly differently pitched plates.  Both sound very good with no difference in tone that I would attribute to plate tuning.  In fact the one that is most "wrong" from a tuning standpoint sounds best.

What DOES consistently make a difference is plate weight and careful, barrel shaped arching.  The better sounding violin has these two characteristics more right.

Plate tuning may have some meaning after you get everything else right, but I am a long way from that!

[Shunyata]

...and this is what Mr. Noon tried to tell me all along, but I wouldn't take his word for it.

[Violadamore]

On 9/3/2020 at 1:20 PM, Bits4Waves said:

As we are talking using the company's account.......... our goal here to share a tool with the Luthier community.

............First of all, it's important make it upfront clear that we are not Luthiers, nor Scientists, but simply software developers........................

Our mission as a software company is to do our best to help the community, by making it more practical and straightforward to apply the state-of-the-art research findings in the making of fine instruments. Thank you all for the attention, and please give us all the feedback and criticism!

Thanks to @Violadamore for the kind message!

 

IMHO............

Having considerable scientific, engineering, and software development experience (military and consumer firmware, business applications, scientific applications, and gaming), as well as some more than nodding acquaintance with luthiery, I feel it necessary to warn you that your expert system approach is endangered from at least two hazards you may not be aware of:

1).  Both the data and the design approach which you are basing your algorithms on is currently considered dated (some would say obsolete), as well as controversial, rather than authoritative, by the luthier community you project as a customer base.

2).  The possibility of repeatably designing or modeling violins via calculation has been exhaustively discussed on this forum.  To constrain a very complicated situation to a necessarily brief explanation, the current verdict seems to be that static approaches fail, while neither the theoretical tools to attempt the dynamic solutions required, nor the budget to develop better ones, are available.  The bottom line is that due to the number of extremely multivariate simultaneous equations to solve, any available real computer will founder before it gets there, and we don't even know what all the variables are, yet.

While we're at it, it doesn't help that few to none of the stakeholders agree on what a "great" violin should sound and respond like.

My tongue-in-cheek (as well as cheeky ) comment earlier, picturing a violin resurrection proceeding through occult means, was intended to evoke a sense of what I perceive as the current reality of the violin business.  Making excellent violins is still a black art performed by adepts, which defies scientific analysis, and bodes well to continue doing so. 

[Don Noon]

6 hours ago, Violadamore said:

Making excellent violins is still a black art performed by adepts, which defies scientific analysis, and bodes well to continue doing so. 

I mostly agree, with two minor comments...

-The "black art" consists mostly of trial-and-error with lots of trials, with good players providing the feedback as to what is good and what is an error.  After a decade or so of that, you can get pretty good.  The time can be shortened by leeching off of a skilled maker who has already been through all of that.

-Scientific analysis can perhaps give slight nudges in directions that might be beneficial... and also nudge into dead ends.  The evaluation part is critical, rather than assuming the theory is correct.  I have yet to see a solid evaluation of whether all this taptone stuff results in a better instrument, and my instincts tell me it is unlikely I will ever see one.

[Violadamore]

1 hour ago, Don Noon said:

I mostly agree, with two minor comments...

-The "black art" consists mostly of trial-and-error with lots of trials, with good players providing the feedback as to what is good and what is an error.  After a decade or so of that, you can get pretty good.  The time can be shortened by leeching off of a skilled maker who has already been through all of that.

-Scientific analysis can perhaps give slight nudges in directions that might be beneficial... and also nudge into dead ends.  The evaluation part is critical, rather than assuming the theory is correct.  I have yet to see a solid evaluation of whether all this taptone stuff results in a better instrument, and my instincts tell me it is unlikely I will ever see one.

Yup.  My "black art" analogy is common to any arena where empirical, experience-based craftsmanship continues to outperform engineering based on number crunching.  There's something about achieving something through "feel" that can't be done better with calculation (or fully explained from theory), that seems to offend the crap out of certain people.  My usual response to that is tell somebody to go back to the lab or the field to fill in the obvious blanks in their paradigm, rather than waste time shooting the messenger. 

I feel that the CAS experiments were groundbreaking in their day, but what was produced at that time falls obviously short of being a "Unified Violin Theory".   Continuing to refold the same steel reaches a limit beyond which it is useless, and even detrimental. 

 

[Marty Kasprzyk]

1 hour ago, Don Noon said:

I mostly agree, with two minor comments...

-The "black art" consists mostly of trial-and-error with lots of trials, with good players providing the feedback as to what is good and what is an error.  After a decade or so of that, you can get pretty good.  The time can be shortened by leeching off of a skilled maker who has already been through all of that.

-Scientific analysis can perhaps give slight nudges in directions that might be beneficial... and also nudge into dead ends.  The evaluation part is critical, rather than assuming the theory is correct.  I have yet to see a solid evaluation of whether all this taptone stuff results in a better instrument, and my instincts tell me it is unlikely I will ever see one.

I very much agree with Don's statement that trial and error is crucial.  The book "How Innovation works" by Matt Ridley gives numerous examples of repeated failures that eventually led to successes and that science understanding often followed afterward to explain it--thermodynamics theory followed the invention of the steam engine is an example.

Ridley's crucial point is that failure is not so bad.  If you are not free to fail then you can't do something different and then no progress is made because no trial and error process happens.    We see that today in violin making.  Great makers can not afford to make a bad instrument for a commissioned client.  This would damage their reputation and future business.   Therefore they can't take any risks of trying new ideas that might fail.  They're stuck doing the same thing over and over again--tradition.

 

 

[David Beard]

While good old Hutchins has kept folks excited and busy for many decades, try some update with Colin Gough.

Those carefully worked resonances just don't stay put when the violin works as a whole.  But that's just the beginning of difficulties.

Most of all, there is no correlation between 'good instruments' having any particularly narrow range of resonance values, or relationships between the values.  What's seen in good instrument scatters somewhat broadly.  Further, there are no correlations showing that any certain values of resonances belong to good instruments but not to bad instruments also.  So the resonances look like passengers on both good and bad instruments, not determinants.

Lastly, a violin is not played by stimulating the resonances with random energy and letting the resonances ring at their natural frequencies.  Only the stopped and unstopped strings are stimulated that way.   

Instead, the strings drive a signal into the instrument.  The instrument resonances then respond to this drive.   In such driven behavior, the Q of each resonance is as significant as its natural frequency.  What is more important than the specific natural frequencies are how well the resonances bend to respond to the driving signal.

In short, the violin is not a xylophone.

[Bits4Waves]

Thank you all for your answers! It's great to see so much love for the making of fine instruments!

We hope the spreadsheet will help you tune your plates by telling you where to remove wood, so you can achieve the frequencies you want your plates to have.

In our previous post we cited some references on how to choose those frequencies and how they relate to the category of the instrument you want to make.

We'd love to hear about the effectiveness of the recommendations.

Please give us any feedback here, on send us a private message!

Edited September 7, 2020 by Bits4Waves

[HoGo]

On 9/5/2020 at 3:16 AM, Violadamore said:

Yup.  My "black art" analogy is common to any arena where empirical, experience-based craftsmanship continues to outperform engineering based on number crunching.  There's something about achieving something through "feel" that can't be done better with calculation (or fully explained from theory), that seems to offend the crap out of certain people.  My usual response to that is tell somebody to go back to the lab or the field to fill in the obvious blanks in their paradigm, rather than waste time shooting the messenger. 

I feel that the CAS experiments were groundbreaking in their day, but what was produced at that time falls obviously short of being a "Unified Violin Theory".   Continuing to refold the same steel reaches a limit beyond which it is useless, and even detrimental. 

 

Being a mathematician I see that there are many areas of math or computer sciences that could very possibly outperform the "black art" but that would require teaming top level builders with some top level mathematicians (I don't think you'll find many luthiers that can do this alone) and produce many instruments (both good and bad) to create good base of data to crunch on. This is the part that is almost impossible to reach as top makers don't have time to produce data and generally don't like to share too many details. It's the human part where such project fails, the science is out there waiting for it's chance.

Just imagine all the capabilities of various computer systems in areas that were considered "human only" just few years or decades ago... Police cameras can recognize faces in a mass of people or analyze hundreds of cars per minute (recognising license plate and type of car and coparing to database if they fit etc.)

[Violadamore]

1 hour ago, HoGo said:

Being a mathematician I see that there are many areas of math or computer sciences that could very possibly outperform the "black art" but that would require teaming top level builders with some top level mathematicians (I don't think you'll find many luthiers that can do this alone) and produce many instruments (both good and bad) to create good base of data to crunch on. This is the part that is almost impossible to reach as top makers don't have time to produce data and generally don't like to share too many details. It's the human part where such project fails, the science is out there waiting for it's chance.

Just imagine all the capabilities of various computer systems in areas that were considered "human only" just few years or decades ago... Police cameras can recognize faces in a mass of people or analyze hundreds of cars per minute (recognising license plate and type of car and coparing to database if they fit etc.)

Unfortunately, before you can computerize something, as you note, you have to have data to process.  With faces, or cars, or license plates, you have images to compare.  Even if the "top makers" had data on hand to contribute, exactly what would that data have to be?

How do you even define a "good violin", much less dictate a prescription for carving one, first time, every time?  Please explain this to me.  If you are running on the assumption that the job at hand is to recreate Cremonese classics from new materials, then please give me the list of numbers to code to functionally replicate a Stradivarius from a stack of tonewood. 

[GeorgeH]

7 minutes ago, Violadamore said:

please give me the list of numbers to code to functionally replicate a Stradivarius from a stack of tonewood. 

The number is 42.

[GeorgeH]

1 hour ago, HoGo said:

Being a mathematician I see that there are many areas of math or computer sciences that could very possibly outperform the "black art" but that would require teaming top level builders with some top level mathematicians (I don't think you'll find many luthiers that can do this alone) and produce many instruments (both good and bad) to create good base of data to crunch on.

I think that this could ultimately be done with carbon fiber violins because you would have a reproducible starting material. I don't think you could do it with violins made of wood.

But, as @Violadamore said, first you must answer the Zen Koan "What is a good violin?"

[Violadamore]

2 minutes ago, GeorgeH said:

The number is 42.

Shoot, if you're working at that level, forget the silly violins, and PM me the blueprints and manual set for a hyperdrive.   

[Violadamore]

20 hours ago, GeorgeH said:

I think that this could ultimately be done with carbon fiber violins because you would have a reproducible starting material. I don't think you could do it with violins made of wood.

But, as @Violadamore said, first you must answer the Zen Koan "What is a good violin?"

I feel that "What is a good violin?" belongs more to Mikkyo/Shingon, than to Zen. 

[Don Noon]

Even if one was able to define a "good violin" (or at least settle on one example), there's the problem of infinite variables necessary to completely define it, and I feel it is unlikely to be captured by just the ones that are normally tabulated.  How do you tablulate the inflection point of the arching around the perimeter?  Is crossgrain damping as a function of frequency important?  What about the L-T shear modulus of the spruce?  Etc.

Maybe making great violins is a black art as well as a bit of luck in having the uncontrollable variables work out well.  Even great makers have some level of inconsistent results.

[Violadamore]

25 minutes ago, Don Noon said:

Even if one was able to define a "good violin" (or at least settle on one example), there's the problem of infinite variables necessary to completely define it, and I feel it is unlikely to be captured by just the ones that are normally tabulated.  How do you tablulate the inflection point of the arching around the perimeter?  Is crossgrain damping as a function of frequency important?  What about the L-T shear modulus of the spruce?  Etc.

Maybe making great violins is a black art as well as a bit of luck in having the uncontrollable variables work out well.  Even great makers have some level of inconsistent results.

I just pitched what could just as well be an answer to this thread into the "sound under your ear" thread (it all ties together), and estimated a solution would consume $20 billion and around 9 years of development..  You think I underestimated?