Politically incorrect way to make violin?

[christian bayon]

2 minutes ago, baroquecello said:

I would agree with the player: it sounds very good!

As you say the thickened part under the f hole is the least efficient part, maybe you could elaborate on the other parts a little? I'm a little surprised by your bass bar, in particular when thinking of your triangular shaped bass bar which was somewhat heavier than a regular one, this one is likely a lot lighter, is it not? How do you explain that it works well nonetheless? Because the missing weight is in the braces?

And the most important question: have you tried this out on a cello yet?

The carbon bass bar is in the same objective than the previous triangle one, just a step further. It’s not light , bit more rigid. 
i have tried on cello, not the carbon fibre bass bar but the brace. Interesting because I made this new top after a normal top on the same cello. Same improvement.

[HoGo]

Is the carbon hollow tube or solid rod?

I guess the rod/tube is mostly under tension, wouldn't similar rod/bar made out of one piece of some strong wood have similar effect (avoiding the gluing of CF and wood blocks - is it epoxy or CA glue?)

[David Burgess]

16 minutes ago, HoGo said:

I guess the rod/tube is mostly under tension,

I was visualizing it more in terms of compression. He has a straight rod going almost all the way between the upper and lower blocks, which might take most of the longitudinal compressive load off the top.

[christian bayon]

30 minutes ago, David Burgess said:

I was visualizing it more in terms of compression. He has a straight rod going almost all the way between the upper and lower blocks, which might take most of the longitudinal compressive load off the top.

Right.

its a tube. I put some tension too, like I do with normal bass bar.

you get the point David, I did before between the blocks but it was a sound disaster, because it didn’t allow the whole body torsion.

[HoGo]

1 hour ago, David Burgess said:

I was visualizing it more in terms of compression. He has a straight rod going almost all the way between the upper and lower blocks, which might take most of the longitudinal compressive load off the top.

But his upper bout is pretty much immobilized in this respect with the "A-frame" and lower bout has the extra thick area near the other end so I mostly thought of bridge pressure and "bending" part of the forces where the bar would be under tension. I wonder if one can somehow measure what is really happening there.

I guess the stiffness of the bar (assuming the ends are perfectly fixed within the blocks) doesn't allow much relative displacement of the three anchored spots. I wonder what vibrational modes would this affect the most...

[David Burgess]

52 minutes ago, HoGo said:

But his upper bout is pretty much immobilized in this respect with the "A-frame" and lower bout has the extra thick area near the other end so I mostly thought of bridge pressure and "bending" part of the forces where the bar would be under tension.

 

I expect there is some of both going on, with the loads being different on opposite sides of the tube.

[Don Noon]

2 hours ago, David Burgess said:

I expect there is some of both going on, with the loads being different on opposite sides of the tube.

With just visual analysis, there should be bending stress on the tube, as it would be so much stiffer than the top for the bridge vertical loads.  While I suspect that the same would be true for the longitudinal string compressive loads, it is not as obvious to me.  I could envision that the tube might be in tension, as the arched spruce top gets deflected downward.  Spruce is pretty stiff longitudinally.

But I don't think any of that is meaningful in terms of tone, as static loads are mostly unimportant.  What I do see as tonally important is that all of the low-frequency modes appear to be outrageously stiff with outrageously high mass that needs to be moved, and the higher frequency zones light and free.  The recordings are nowhere good enough to tell what it sounds like, and we really need an A/B comparison with a good quality concert violin to do a better evaluation.

[scordatura]

Mario Beauregard makes "facetts" guitars that have ridges on both the top and back. He claims tonal benefits. He places the ridges on nodal lines. It has been patented at least for guitar.

http://beauregardguitars.com/guitars/facettes/

[David Burgess]

57 minutes ago, scordatura said:

Mario Beauregard makes "facetts" guitars that have ridges on both the top and back. He claims tonal benefits. He places the ridges on nodal lines. It has been patented at least for guitar.

http://beauregardguitars.com/guitars/facettes/

The issue I see with that, is that there is no such thing as fixed nodal lines, on an instrument which plays variable frequencies.

But perhaps the scientifikimal-sounding strategy allowed him to sell more instruments than he would have otherwise? 

[ctanzio]

On 5/7/2021 at 2:30 AM, christian bayon said:

If all the stress ...  are absorbed by certain part of your violin, the part you want to vibrate will be more free.

Although the above is technically not correct using a strict application of physics, placing membranes under stress will alter the frequencies of the nodes. A simple example is to consider violin strings as one adds tension to them.

It is possible you are shifting the frequencies of many nodes above the typical 3kHz to 4kHz roll off of a violin into the lower frequency band. This increases something sometimes called the "spectral density" which can result in a stronger response. Whether or not it results in a "better sound" for the violin is anyone's guess, but the sound clips you posted do sound nice.

 

[Don Noon]

47 minutes ago, ctanzio said:

A simple example is to consider violin strings as one adds tension to them.

A better example would be a curved beam in tension.  In that case the vibration frequencies are not significantly changed.

Only in flat/straight things will tension always act to restore a vibration to the rest position, thus increasing its frequency.  With curvature, the tension tends to move the part away from the rest position on one half of the vibration cycle, cancelling out the restoring force action on the other half, thus leaving the frequency about the same as without tension.

This assumes the tension and vibration do not appreciably change the basic geometry, which would introduce nonlinearities in the physics.

[scordatura]

2 hours ago, David Burgess said:

The issue I see with that, is that there is no such thing as fixed nodal lines, on an instrument which plays variable frequencies.

But perhaps the scientifikimal-sounding strategy allowed him to sell more instruments than he would have otherwise? 

My thoughts also. I heard him talk about this in a "Luthier on Luthier" podcast. https://luthieronluthier.libsyn.com/53-mario-beauregard "Luthier on Luthier" is an interesting podcast series. For every adventurous violin maker, there are a bunch more on the guitar side. They seem less bound by tradition than we are. Some of the guitar crowd has embraced torrefied wood use for instance to get that "vintage Martin sound". Another thing I find interesting is the guitar makers use of different kinds of bracing. I enjoy listening to the podcasts.

[Marty Kasprzyk]

3 hours ago, ctanzio said:

Although the above is technically not correct using a strict application of physics, placing membranes under stress will alter the frequencies of the nodes. A simple example is to consider violin strings as one adds tension to them.

It is possible you are shifting the frequencies of many nodes above the typical 3kHz to 4kHz roll off of a violin into the lower frequency band. This increases something sometimes called the "spectral density" which can result in a stronger response. Whether or not it results in a "better sound" for the violin is anyone's guess, but the sound clips you posted do sound nice.

 

When you put on the strings and bring them up to correct pitch does the violin's frequency response curve change any way?  Do the A0, B1-, B1+ frequencies all stay the same as the string tension increases?

[christian bayon]

10 hours ago, Marty Kasprzyk said:

When you put on the strings and bring them up to correct pitch does the violin's frequency response curve change any way?  Do the A0, B1-, B1+ frequencies all stay the same as the string tension increases?

I guess yes, because, the wolf note (B1+?) is on the same note if you tune at 430 or 442hz.

[violins88]

I still think you should do the knock test and get the fft. Maybe there is change in the B1+,  but maybe just not enough change to provide enough change in the wolf.

[Peter K-G]

On 5/7/2021 at 5:05 PM, Don Noon said:

But I don't think any of that is meaningful in terms of tone, as static loads are mostly unimportant. 

IMO static load is one of the most important things when it comes to the players sense/feel for producing the sound.

When a (good) violin reaches its equilibrium of deformation, it tells the player a lot of things. (try listen to the Tucson again and look for the firm "bam.. " when the strings are attacked...

It takes a couple of days for this to happen when you string up a violin that has been resting a couple of years.

 

[christian bayon]

3 hours ago, violins88 said:

I still think you should do the knock test and get the fft. Maybe there is change in the B1+,  but maybe just not enough change to provide enough change in the wolf.

Sorry, I don’t do knock test.

[Don Noon]

5 hours ago, Peter K-G said:

IMO static load is one of the most important things when it comes to the players sense/feel for producing the sound.

It takes a couple of days for this to happen when you string up a violin that has been resting a couple of years.

I think you are confusing static load and the transient effects static load has on damping.  A player could well sense a different static load on strings (i.e. different tension), but I see no way anyone would sense how the load is distributed in the instrument body.

5 hours ago, violins88 said:

I still think you should do the knock test and get the fft. Maybe there is change in the B1+,  but maybe just not enough change to provide enough change in the wolf.

2 hours ago, christian bayon said:

Sorry, I don’t do knock test.

I have done plenty of them, with strings, without strings, and with strings at half tension.  Without strings, there is some difference, likely due to the missing bridge, tailpiece, and string mass, but I don't find any change to the body mode frequencies with changes in string tension.

 

 

 

[violins88]

Don,

Thanks for that information. 

[David Burgess]

1 hour ago, Don Noon said:

I think you are confusing static load and the transient effects static load has on damping.  A player could well sense a different static load on strings (i.e. different tension), but I see no way anyone would sense how the load is distributed in the instrument body.

I have done plenty of them, with strings, without strings, and with strings at half tension.  Without strings, there is some difference, likely due to the missing bridge, tailpiece, and string mass, but I don't find any change to the body mode frequencies with changes in string tension.

I've messed around with this quite a bit too, and also didn't find any consistently measurable change in the mode frequencies. However, there does seem to be a major change in sound and playing characteristics, and I don't think this is attributable to increased or reduced string tension under the player's fingers, since one can increase or reduce the tension of the A string, and a player will perceive quite a difference when playing only the G string.

I am not yet able to explain the mechanism by which this happens, but have done enough back-and-forth experimenting to believe that it's real.

[Carl Stross]

2 hours ago, Don Noon said:

but I don't find any change to the body mode frequencies with changes in string tension.

That's strange, to say the least....  There should be some change.

[Marty Kasprzyk]

Doesn't putting in a loose sound post (low tension on the plates) differ in sound from one which is tight (high tension on the plates)?

[Peter K-G]

It's when the violin reaches its equilibrium state of deformation.

I think that is what the "playing in" really is about.

(sure humidity cause of sweating and breathing is a big factor but that has other characteristics, that actually lowers the modes and gives the violin a darker tone)

[ctanzio]

The string example was mean to demonstrate that changing the tension in a structure changes the frequency at which it vibrates. I was not making any inference on how string tension specifically alters the frequencies of the plates.

The stress distribution in a fully loaded violin is hardly uniform. Specifically, the stresses in the upper and lower bouts of the belly are significantly different than those in the center. Even going from left to right across a bout will see fundamental differences in stress conditions. So it would not be surprising to find some modes whose frequency is mostly unaffected by changes in string tension.

If the entire violin plate was being uniformly stretched or compressed, then one might notice changes in frequency across many modes.

I think Christian Bayon is exploring an interesting approach to "tuning" the plates, although it seems to be mostly in the "try it and see" phase. Modern analytical tools fall short in this endeavor because they need some quantitative goal to direct each iteration. People cannot agree on what a great violin sounds like, so they cannot derive quantitative measurements to direct the analysis.

[Don Noon]

1 hour ago, Carl Stross said:

That's strange, to say the least....  There should be some change.

Some, perhaps... significant, no.

I just performed a test: tuned, then D and A detuned to an octave low (~1/4 normal tension, I believe), and then back up to repeat the tuned state.

A0 was 275 Hz in all conditions, as you might expect this mode to be least affected.

B1- was 432, 428 detuned, and 415 tuned back up (the B1- in this instrument I think has two close peaks, and likely the lower one dominated when tuned back up). Repeating the re-tuned test:  433 Hz.

B1+ was 545, 540 detuned, and 543 retuned. Repeating the re-tuned test: 536 Hz.

So, OK, there are some frequency differences in the signature modes... but not much more than measurement scatter, and I would hardly call them significant.  Far more significant is what goes on with the amplitudes, especially at the higher frequencies, and that changes much more.  Throw into the mix the known transient (over days) effect that a change in static stress has on (primarily) damping, and you can a range of tonal effects... but I still think that the body mode frequencies stay essentially the same.