Carving site recommendations for plate tuning

[Don Noon]

2 hours ago, Carl Stross said:

Even theoretically, only in some specific cases. Violin does not seem to me to be one of those. Not even close.

I would like to see your theory and evidence of this.

From what I've seen, the violin is pretty darn close to a linear system that will reproduce only the input frequencies.

[David Beard]

3 hours ago, Marty Kasprzyk said:

A violin is indeed like a steel drum--if you hit it, it makes a sound.  

The frequency response curve of the sound coming from a hammer impact to the bridge is the same as when the strings are bowed in a glissando.  Both methods excite the various body resonances the same way.

The attached paper describes this more thoroughly.

AilinZhang_JimWoodhouse_JASAJournal.pdf 998.42 kB · 1 download

Yes. If you hit a table it makes a sound.  And when you hit a soundboard it makes a sound.  But when you play a piano, you don't normally hit the sound board.  You hit the strings.  The soundboard in turn is driven by the strings, a very eifferent behavior than hitting the sound board.

Likewise, yes you can hit a violin.  But, that is very very different than the primary behavior of bowing the strings.

When you hit the violin, then you excit the natural frquencies of it's resonance.  When you hit s violin, the frequencies of the resonances matter greatly.  Although, in an assembled violin with tuned strings, those frequencies will have moved around from what they were on the free plates.  Nevertheless, WHEN YOU HIT a violin the natural frquencies are very relevant.

But when you bow, the behavior is very different.  The signal from the strings drives the violin response.  The natural frquencies of the resonances are only starting points.  The resonance will be forced to follow frequencies from the driving signal if they are to take in enegy and vibrate.    The Q of each resonance becomes very important.  This determines how readily the frquency can bend to respond to thw drive.   WHEN YOU BOW a violin, the specific natural resonance frequencies become much less significant.

A hit violin is like a steek drum.  A bowed violin is NOT.

[David Beard]

2 hours ago, Carl Stross said:

Even theoretically, only in some specific cases. Violin does not seem to me to be one of those. Not even close.

Sorry Carl, but a bowed violin is a driven system.  Ot isn't a simple case, but it is clear cut and absolutely true.

The only resonances that get to sustain in a violin are the modes whose Qs are low enough to allow their natural frequency to bend to an energetic frequency in the signal.  There is no other way.

[Violadamore]

1 hour ago, Don Noon said:

I would like to see your theory and evidence of this.

From what I've seen, the violin is pretty darn close to a linear system that will reproduce only the input frequencies.

True, but, IMHO, analysis of the output spectrum is complicated by the lack of a clean look at the input spectrum.  The assumed input is tidily theoretical, but that's not what the player is actually usually producing, as one spends a lifetime learning ways to modulate with the bow.  This acts to confuse everyone who's looking at the problem simplistically. 

It's my considered opinion that sorting this matter out will require much improved observation tools to empirically establish what violins actually do when played (to microscale resolution), followed by an exascale computing project (similar to current state-of-the-art in geophysics), to mathematically model what is observed. 

The impact hammer approach, BTW, IMHO, is too different from playing to recreate what playing produces, adding to the confusion factor.

 

[Carl Stross]

59 minutes ago, David Beard said:

Sorry Carl, but a bowed violin is a driven system.  Ot isn't a simple case, but it is clear cut and absolutely true.

The only resonances that get to sustain in a violin are the modes whose Qs are low enough to allow their natural frequency to bend to an energetic frequency in the signal.  There is no other way.

I see. 

Well, I'll have to think about it some more. You could be right and I am often wrong. 

[Carl Stross]

1 hour ago, Don Noon said:

I would like to see your theory and evidence of this.

From what I've seen, the violin is pretty darn close to a linear system that will reproduce only the input frequencies.

I will give it some more thought. I'm not comfortable with this idea of a violin being a "linear system" but it might be just a preconceived idea. I found myself to be wrong quite a few times.

[Marty Kasprzyk]

1 hour ago, David Beard said:

Yes. If you hit a table it makes a sound.  And when you hit a soundboard it makes a sound.  But when you play a piano, you don't normally hit the sound board.  You hit the strings.  The soundboard in turn is driven by the strings, a very eifferent behavior than hitting the sound board.

Likewise, yes you can hit a violin.  But, that is very very different than the primary behavior of bowing the strings.

When you hit the violin, then you excit the natural frquencies of it's resonance.  When you hit s violin, the frequencies of the resonances matter greatly.  Although, in an assembled violin with tuned strings, those frequencies will have moved around from what they were on the free plates.  Nevertheless, WHEN YOU HIT a violin the natural frquencies are very relevant.

But when you bow, the behavior is very different.  The signal from the strings drives the violin response.  The natural frquencies of the resonances are only starting points.  The resonance will be forced to follow frequencies from the driving signal if they are to take in enegy and vibrate.    The Q of each resonance becomes very important.  This determines how readily the frquency can bend to respond to thw drive.   WHEN YOU BOW a violin, the specific natural resonance frequencies become much less significant.

A hit violin is like a steek drum.  A bowed violin is NOT.

I suggest you give your violin's bridge a fingernail tap and do an Audacity analysis of the recorded sound.  Then play glissandos on the E and G strings up to high positions and do an Audacity analysis of those sounds too.  The frequency response curves  will look very similar--all the resonance peaks will be at the same frequencies for both tests.

You can believe what I say or believe your own eyes.

[David Beard]

23 minutes ago, Marty Kasprzyk said:

I suggest you give your violin's bridge a fingernail tap and do an Audacity analysis of the recorded sound.  Then play glissandos on the E and G strings up to high positions and do an Audacity analysis of those sounds too.  The frequency response curves  will look very similar--all the resonance peaks will be at the same frequencies for both tests.

You can believe what I say or believe your own eyes.

When you tap the bridge, you are fairly directly tapping the violin body.  Under this circumstance the violin is not predominantly receiving a driving signal from the strings.  In this circumstance, your are playing the violin like a steel drum, and it is responding that way.   The natural frequencies of the resoanances of the assembled instrument will be very significant in this case, and related to the free plate resonances in the way shown by Gough.

Bur this is not the case when you bow the strings and stimulate the violin witj a driving signal.

The BOWED VIOLIN is not a steel drum.

[Anders Buen]

The bowed string is nonlinear. A linear translation becomes something periodic, eventually. That is nonlinear.

[Dennis J]

After reading all of these comments I get the impression that makers have trouble getting consistent results sound wise despite making instruments to the same specifications.

My guess is that without careful calibration of arching no two instruments will be the same.

 

[Don Noon]

27 minutes ago, Dennis J said:

After reading all of these comments I get the impression that makers have trouble getting consistent results sound wise despite making instruments to the same specifications.

My guess is that without careful calibration of arching no two instruments will be the same.

 

Even WITH careful calibration of arching, there will be differences.

How would you get wood that is identical in all properties, everywhere, in every direction?  I have quite a stockpile of same-tree spruce, and although the range is narrower than with from different trees, it still varies in the overall gross properties that I can measure easily.  There are a lot more properties that I CAN'T measure.

[Dennis J]

Well, density and mass can be quite accurately measured. And as far as I know they are the major factors at play.

I'm simply looking at arching as a structure that supports the bridge in a stable way, so control of that arching, both back and front, seems pretty important to me. And the synchronicity of the back and front inflection surfaces I've shown looks promising. 

 

[Marty Kasprzyk]

My main point is that if you duplicate the stiffness, mass and damping of every violin part you have a good likelihood of duplicating the sound too.

Just duplicating a few parts won't give you consistent results.

Matching plate mode frequencies isn't a worthless thing to do, its just not sufficient.

[Dennis J]

12 hours ago, ctanzio said:

This is a fundamental truth of driven systems that is difficult for many people to comprehend.

To try and state it more clearly: when you play a note on the violin, the standing waves that develop in the body have the same frequencies of the vibrating string, not the natural frequencies of the violin body.

 

That makes a lot of sense to me. What more do you need to know? How the body reacts to the bowed string is the important thing. So that comes down to how it is constructed.

 

[Peter K-G]

8 hours ago, Marty Kasprzyk said:

My main point is that if you duplicate the stiffness, mass and damping of every violin part you have a good likelihood of duplicating the sound too.

Just duplicating a few parts won't give you consistent results.

Matching plate mode frequencies isn't a worthless thing to do, its just not sufficient.

Like again! 

 

 

[Peter K-G]

Backplate in progress

density: 0.585

low speed of sound:  18 mm high arch

good cross grain strength

UV box at 343 Hz, 16h UV  -> 357 Hz after -4 g 

 

 

[Marty Kasprzyk]

Your graph is a perfect example why "plate tuning" is insufficient.  If you thin your top plate to have a mode 5 frequency of 357Hz you hit that at two different weights:  about 112g and then again at 101g.  I don't believe that violins assembled with plates having these different weights would sound or play the same.  

Weight should be controlled too.

[Peter K-G]

1 hour ago, Marty Kasprzyk said:

Your graph is a perfect example why "plate tuning" is insufficient.  If you thin your top plate to have a mode 5 frequency of 357Hz you hit that at two different weights:  about 112g and then again at 101g.  I don't believe that violins assembled with plates having these different weights would sound or play the same.  

Weight should be controlled too.

Spot on!

(I assume you understood that if I would have stopped at 112 g 357 Hz, it would show 108 g and ~370 Hz after 16 h of UV box)

The "target" in this case is 103 g and 345 Hz at 6% MC in the wood.

Thicknesses are in average 0.03 mm thicker than another plate with the same weight and M5, that is of 0.63 density.

 

 

 

 

[Don Noon]

1 hour ago, Marty Kasprzyk said:

Your graph is a perfect example why "plate tuning" is insufficient...

Weight should be controlled too.

I would like to see a graph of weight and taptone vs. tone score in blind testing.  I predict it would look like this: A very wide band of random results with no particular correlation to any weight or taptone scheme, with below-average results for things that are waythehell out at the fringes.  At least, that's about what I see in what I have made so far, and what I see in what others have made as well.  And nobody has shown proof of anything else, other than arm-waving or anecdotes that it should be better if you control things to a narrow band.

[Marty Kasprzyk]

25 minutes ago, Don Noon said:

I would like to see a graph of weight and taptone vs. tone score in blind testing.  I predict it would look like this: A very wide band of random results with no particular correlation to any weight or taptone scheme, with below-average results for things that are waythehell out at the fringes.  At least, that's about what I see in what I have made so far, and what I see in what others have made as well.  And nobody has shown proof of anything else, other than arm-waving or anecdotes that it should be better if you control things to a narrow band.

This kind of experiment was recently done in the Bilbao Project at the BELE Basque Violin Making School in Spain.  Violins were made with the Strad Huberman design using various combination of top and back plates having pliant(floppy), medium, and resistant stiffnesses.

 

The plates and assembled violins were throughly tested and blind listening tests were done to determine preferences but the data and results have not yet been published.

 

See the following: 

http://www.bele.es/en/bilbao-project-introduction/  

 

http://www.bele.es/en/bilbao-project-progression/  

 

I expect to see a lot of scatter in the listening tests because people have different tastes.  Of more importance to me is the player's opinions.

[Peter K-G]

1 hour ago, Don Noon said:

I would like to see a graph of weight and taptone vs. tone score in blind testing.  I predict it would look like this: A very wide band of random results with no particular correlation to any weight or taptone scheme, with below-average results for things that are waythehell out at the fringes.  At least, that's about what I see in what I have made so far, and what I see in what others have made as well.  And nobody has shown proof of anything else, other than arm-waving or anecdotes that it should be better if you control things to a narrow band.

Why would you like to see that? A blind test of one single violin would probably also look as scattered

[David Beard]

Is there an ideal plate weight plate weight shared by the best historicsk concert violins?

Are there ideal tap tone frquencies shared by these violins?

How then do you pick your target weight, your target frequencies?

Are there not bad violins with your target values?

If your targeting is not generally shared by good vioiins, does not steer you to a good violin, and does not steer you away from making a bad one, what is the purpose of control something that arrives at decent normal range values automatically?

[Peter K-G]

10 minutes ago, David Beard said:

Is there an ideal plate weight plate weight shared by the best historicsk concert violins?

Are there ideal tap tone frquencies shared by these violins?

NO

How then do you pick your target weight, your target frequencies?

Copy, Copy, Copy...... 10 violins, with different wood.

Are there not bad violins with your target values?

Don't know, probably, who cares?

If your targeting is not generally shared by good vioiins, does not steer you to a good violin, and does not steer you away from making a bad one, what is the purpose of control something that arrives at decent normal range values automatically?

What is normal? Range is for ex. Il Cannone and the Plowden.

Any maker that has made a few can steer in the direction left or right

 

[David Beard]

I do understand that for a copyist the frequencies can be viewed as just one more detail to copy in a specific instrument.  But I don't see any validity in believing the effort to tune plate tap resonances helps create an instrument with good sound or response.

The specific frequencies of the plate resonances do not appear to be a cause of good sound or response.

 

Beyond falling in the broadly scattered range of classical examples, the targeting of your frequencies does seem to matter.    

No matter how the frequencies fall within this broad scatter, you can potentially make a good violin and potentially make a bad one.  The outcome does not depend on the frequencies.

It's nearly as senseless as saying the secret to good sound is copying the purfling corner miters and bee stings.

[JacksonMaberry]

21 hours ago, Dennis J said:

After reading all of these comments I get the impression that makers have trouble getting consistent results sound wise despite making instruments to the same specifications.

My guess is that without careful calibration of arching no two instruments will be the same.

 

No two violins will be the same, no matter what is done. Full stop