Jump to content
Maestronet Forums

How does a violin reproduce overtones? - Theorizing a model


Andreas Preuss

Recommended Posts

  • Replies 646
  • Created
  • Last Reply

Top Posters In This Topic

3 hours ago, Andreas Preuss said:

I would rather say two L bars formed from linings plus the width of the channel and they are linked via the ribs. The overhang of the ribs doesnt really do anything.

Hmm...

We disagree on that.    The rib is the web of the I beam, the glued on edge works are the flanges.  The lining joins these.

This configuration reduces the buckling potential of the ribs. This also makes the structure stiffer and stronger in some directions, while remaining fairly elastically twistable and bendable in other directions.

Link to comment
Share on other sites

On 6/20/2021 at 4:50 AM, David Beard said:

Hmm...

We disagree on that.    The rib is the web of the I beam, the glued on edge works are the flanges.  The lining joins these.

This configuration reduces the buckling potential of the ribs. This also makes the structure stiffer and stronger in some directions, while remaining fairly elastically twistable and bendable in other directions.

I see it like this. On the left classical style edge work with a wider channel. On the right modern style edge work. (Both a bit exaggerated) 

i see the structural strength in the red areas. The overhang of the plates (marked in black) doesn’t do much. If you would align it with the ribs like a guitar the sound won’t change. However height and thickness of the linings including the glue surface contributes to the stability in that area. The portion of the ribs between the linings is a sort of necessary connector to form the cavity. 48A224B3-5F99-4F10-89D5-A78515D230DA.thumb.jpeg.a2978b986b8cbf8ab623fc5dff326cc9.jpeg

Link to comment
Share on other sites

4 minutes ago, Andreas Preuss said:

I see it like this. On the left classical style edge work with a wider channel. On the right modern style edge work. (Both a bit exaggerated) 

i see the structural strength in the red areas. The overhang of the plates (marked in black) doesn’t do much. If you would align it with the ribs like a guitar the sound won’t change. However height and thickness of the linings including the glue surface contributes to the stability in that area. The portion of the ribs between the linings is a sort of necessary connector to form the cavity. 48A224B3-5F99-4F10-89D5-A78515D230DA.thumb.jpeg.a2978b986b8cbf8ab623fc5dff326cc9.jpeg

Okay.  What you say is valid as it stands.  I don't disagree.

But remember, vibrational structures vibrate in all ways available to them. 

 

What I'm pointing to is not a contradiction to what your focusing on, it's an 'also'.

What I'm saying is if we take a blue pencil to your drawing and mark your red areas, but also mark the over hang of the edges, for both top and back edges, and we blue in the rib itself, then our blue area is the cross section of a an I bar shape.

This I bar shape, combined by the curved shape of the ribs, and helped by the blocks; it all gives the edge/sides structure a kind of strength.

These properties of the edge resist bending of the plane of the plate, but just along the edge following the line of the sides.

But, because of the channel shapes, the center of the plate is still very free to move out of plane.

Because of the way the archings descend to the channels, combined with the way the channels go to tje edges, and the complex shape of the ribs/edge structure, the motions of the ribs/ edge get to be a little different than the motions of the plate archings.

Basically the design resists the edge/sides responding to out of plane motion in the arching with similar out out of plane motion at the edges/sides.  Instead, that motion will tend to convert into a combo of the edges/sides moving in and out with motion directions parallel to the plain of the plates, and with twisting of the level of the edges as they sit across the rib.

Sorry.  This hard to say simply in words. I should make a picture like you did. But not for now.

 

 

Link to comment
Share on other sites

On 6/19/2021 at 8:59 AM, Andreas Preuss said:

I would rather say two L bars formed from linings plus the width of the channel and they are linked via the ribs. The overhang of the ribs doesnt really do anything.

You can describe any I bar that way.  You could always say one side isn't doing much, it's just Ls.  If you've got one side, you don't really need both.

You can say the overhangs aren't doing much.  But, they're the other sode of the I structure. They're full in height if not width. They contribute mass. They contribute stiffness.

Link to comment
Share on other sites

9 hours ago, David Beard said:

vibrational structures vibrate in all ways available to them. 

And how they want to vibrate will vary with frequency.

If you want to look at the entire rib zone as an I-beam, then the overhang will add to the stiffness (if you didn't have the overhang, it would be a C-section beam).  I-beams are very stiff vertically.

Andreas' illustration indicates he is more interested in side bending at the joint.  I-beams are not good for bending or twisting sideways, and the overhang wouldn't help.  Wider linings and thicker edgework will help here... but then the bending might just shift down to lower in the rib.

The problem is that we don't have a good handle on where the stresses in the body structure are, and how they vary with frequency.  So theorizing about what's stiffer amounts to a lot of arm-waving in the fog.  Even experiments trying to test various things won't clear up the theory if you don't have a handle on the actual theory to start with.  At best you can find something you like or don't like (which is helpful!), but as for why and how, it's still going to be just guesswork.

One more point... even if you have the theory correct about structural stress vs. frequency and know where to diddle with the structure to make it stiff in a specific mode, you have the problem of determining the effect on all of the other modes, how the stiffness relates to the sound output, and whether or not the aggregate sound is good or not. It's really complicated.

Link to comment
Share on other sites

8 hours ago, David Beard said:

They contribute mass. They contribute stiffness

Mass for this portion is for the back less than 4g for the top less than 3. Negligibly small in my view. 
 

stiffness comes rather from the joint of rib garland to the top or the back than the additional wood in that area. Also negligibly small in my view. 
 

There are in the classical construction elements which were made for practical considerations rather than structural considerations. So just in my view the overhang at the border is one of those things. 
 

Back to the original topic: I am sure this doesn’t contribute anything to the overtones

 

Link to comment
Share on other sites

24 minutes ago, Don Noon said:

One more point... even if you have the theory correct about structural stress vs. frequency and know where to diddle with the structure to make it stiff in a specific mode, you have the problem of determining the effect on all of the other modes, how the stiffness relates to the sound output, and whether or not the aggregate sound is good or not. It's really complicated.

In an ironic sense you could say that we know too much and this DOES make things ‘look like complicated’.

(So, wouldn’t be the conclusion to look on the whole problem more naively?)

The thing is that we don’t deal with gravity and try to figure out why planets circle around the sun.

We also deal with a ‘complicated’ material 

—————————————

i think it is useless to think about triggering the behavior of single modes knowing that other modes might change in an unwanted way. 
 

—————————————

If I look on all the research which has been done, it is analyzing the ‘macro motions’ of a stringed instrument. With ‘macro motions’ I mean that they can be visualized with mode patterns. This helps to understand the ‘mechanics’ of the instrument. But this seems not to solve the problem because there are different ways the same mechanic can work good and bad. 
 

What I am trying to think aloud in this thread is to build any hypothesis which can serve as a practical ‘thought model’ 

Right now I would formulate it like this: 

the resistance in vertical direction against the strings is what triggers the overtones. If an instrument is too weak in this direction it just can’t produce overtones making the sound pretty dull. But any vibration on the body in this axis is automatically linked to vibration on the horizontal axis.  (As seen in holographic pictures). Now if the resistance or stiffness in horizontal direction is too weak there will be no springing force for a quick rebounce causing a dull sound 
 

In lengthwise direction one can stabilize the structure with a stiff table which produces overtones. But this is a sound with weak fundamentals, so it is thin and has no body. 
 

So to achieve the necessary lengthwise strength it must be done from the back. Therefore the recipe thick back and very thin top can work extremely well.

This model explains to me as well why it is possible that an instrument may not sound at all  at the beginning but start to sound eventually very well after a while because the stretching will give the structure the necessary vertical resistance. 

What seems to be important n this very simplified model is that vertical strength in didferent parts must be calibrated against horizontal strength to function best. Overtones come from the ability of the entire structure to rebounce rapidly.. It seems that the back is the best place for this something like storing energy which is used to pull everything back to zero. 
 

This seems to me as well the reason why very overtone rich instruments have almost always a good response. 
 

(to be continued, gotta get some sleep now)

Link to comment
Share on other sites

1 hour ago, Andreas Preuss said:

very overtone rich instruments have almost always a good response.

Overtones, i.e. higher frequencies, are inherently faster to respond to a forcing function.  So, yeah, that makes sense.

As for your mental model, it is nothing at all like my mine.  I tried reading it over 3 times, and it still goes nowhere for me.

I think mostly in terms of energy efficiency, which can be divided into 1) efficient radiation of sound from a vibrating structure (good stiffness/weight wood, arching for increasing antinode area) and 2) minimizing energy loss (low damping materials, preventing inefficient vibrations).  

Then there's balance... generally a thickness-related thing to get the desired bass/treble relation.  The more inherent overtones you can get, the thinner you can go to strengthen the low end for balance, and end up with more overall power.

That's the main part of my thinking, but there are also other bits picked up here and there that enter the mix of thought... ring modes, impedance, poorly radiating resonances to avoid (or shape tone), etc.

But it's all mostly trial-and-error, and the traditional shape, materials, and arching seem to work best... with the exception of torrefied wood as a substitute for 300 year old plates.

Link to comment
Share on other sites

5 hours ago, Andreas Preuss said:

overtone rich instruments have almost always a good response

That is not true and in general is the other way around.

It is pretty easy to produce overtone rich instruments through some thickness / arching/strings manipulation. But they won't have "good response".  Lots of violins can be coaxed into producing an usable tone. Finding players willing to put up with them is not easy.

Link to comment
Share on other sites

6 hours ago, Don Noon said:

I think mostly in terms of energy efficiency, which can be divided into 1) efficient radiation of sound from a vibrating structure (good stiffness/weight wood, arching for increasing antinode area) and 2) minimizing energy loss (low damping materials, preventing inefficient vibrations).  

This looks to me like the basis and on the bottom line you are saying there 'good wood is the key to everything'. (good stiffness/weight wood & low damping materials) No objection

Arching seems to be a more complex theme. 

6 hours ago, Don Noon said:

Then there's balance... generally a thickness-related thing to get the desired bass/treble relation.  The more inherent overtones you can get, the thinner you can go to strengthen the low end for balance, and end up with more overall power.

This is probably almost the same what I am thinking about. Thickness in my thought model influences the overall stiffness lengthwise and crosswise. And it is always more difficult to get sufficient cross stiffness in the violin body.

From experiments reducing thickness of the back from outside in playing condition it is clear that almost nothing happens when for example the back is too stiff and heavy. It is only on a very thin structure that you can observe clear changes in the sound or playing characteristics. (Not so surprising)

This model helps me very roughly to know where I have to find sensitive zones for the overall stiffness in both direction. On the back you have the c bouts and the area around the top block. 

The rest of the back looks to me like something which is only sensitive to overall thickness (In practical terms it is better to thin down the overall surface rather than looking for areas. Maybe a slight unbalance is desirable and therefore calculated thickness patterns don't make sense) 

6 hours ago, Don Noon said:

But it's all mostly trial-and-error,

Mustn't be. I am still convinced that a sound calibrating process can be elaborated. Roughly speaking it looks like this: You know what parameters you start with in terms of dimensions in mm and weight and you leave in those dimensions room for 'in-process-calibrations'. 

In those terms I don't think of the rib height as a pre-fixed measure any more, same for string angle (from the neck and lower nut) and bridge height, just to name the most important elements. The reason is simply that playing around with only these things you can alter the overall sound characteristics A LOT. I learned that the sound can change a lot under the condition that you have already pretty thin and light parts. It is more difficult to figure out how thin and light is a good starting point to have room for final calibrations. 

The only bigger problem I am still chewing on is how to find a good idea how to adjust the top arching to a given piece of wood. And that's the primary key for great sound instruments. (just IMO)

Link to comment
Share on other sites

1 hour ago, Andreas Preuss said:

>

The only bigger problem I am still chewing on is how to find a good idea how to adjust the top arching to a given piece of wood. And that's the primary key for great sound instruments. (just IMO)

What's the problem?  Wood properties are all over the map and people's sound preferences are all over the map too.

Violin making and selling is more like great dating matching than getting the right the crunching sound of Pringle potato chips.

Link to comment
Share on other sites

1 hour ago, Marty Kasprzyk said:

What's the problem?  Wood properties are all over the map and people's sound preferences are all over the map too.

Violin making and selling is more like great dating matching than getting the right the crunching sound of Pringle potato chips.

That's for sure. Will remember that!:D

However, you can mess up the top by making the arch too high or too low and I am not an advocate for making a uniform standard height arching on all my violins. Otherwise I would probably hire Mr. CNC as my next assistant. 

 

Link to comment
Share on other sites

18 hours ago, Don Noon said:

efficient radiation of sound from a vibrating structure

That’s pretty abstract. 
 

I am asking myself if the efficiency can’t be translated to some sort of simple model. 

At least it seems clear to me that if you do on a fantastic sounding instrument a wrong setup with the wrong string angle, wrong neck tilt, and the wrong bridge height you can kill the ‘efficient radiation’.  So this is to me an important part in the equation. 

Link to comment
Share on other sites

9 hours ago, Andreas Preuss said:

...I am not an advocate for making a uniform standard height arching on all my violins. Otherwise I would probably hire Mr. CNC as my next assistant. 

Mr. (or Ms.) CNC can make different arching heights too, you know.  He (or she) can carve any shape you want... but explaining exactly what you want him (or her) to do is the really difficult part.

30 minutes ago, Andreas Preuss said:

At least it seems clear to me that if you do on a fantastic sounding instrument a wrong setup with the wrong string angle, wrong neck tilt, and the wrong bridge height you can kill the ‘efficient radiation’.  So this is to me an important part in the equation. 

Those items I would say are involved with getting vibration from the string to the structure, and can certainly have significant effects on transmission and playability.  Once the vibrations get to the body, that's the "efficient radiation" I meant.

I just use pretty standard bridge height, arch height, overstand, etc., so other than cutting the bridge, I don't think much about those things.

Link to comment
Share on other sites

10 hours ago, Marty Kasprzyk said:

Wood properties are all over the map and people's sound preferences are all over the map too.

True to some degree... but the map of population density tends to show most people living in Power City, surrounded by the suburbs of Projection and Response.  Some might live way out in Dullsville, but I think it's uninhabited farther out.

Link to comment
Share on other sites

20 hours ago, Marty Kasprzyk said:

What's the problem?  Wood properties are all over the map and people's sound preferences are all over the map too.

Violin making and selling is more like great dating matching than getting the right the crunching sound of Pringle potato chips.

Marty, 

The Finn is telling you, if the purpose is going on a date, you better sure you have a sexy instrument to show up with, do you? From what I'v seen your's are bigger than AVG, are they?

Link to comment
Share on other sites

On 6/25/2021 at 6:53 AM, Andreas Preuss said:

I see it like this. On the left classical style edge work with a wider channel. On the right modern style edge work. (Both a bit exaggerated) 

i see the structural strength in the red areas. The overhang of the plates (marked in black) doesn’t do much. If you would align it with the ribs like a guitar the sound won’t change. However height and thickness of the linings including the glue surface contributes to the stability in that area. The portion of the ribs between the linings is a sort of necessary connector to form the cavity. 48A224B3-5F99-4F10-89D5-A78515D230DA.thumb.jpeg.a2978b986b8cbf8ab623fc5dff326cc9.jpeg

Well, think again!

Link to comment
Share on other sites

8 hours ago, Don Noon said:

Those items I would say are involved with getting vibration from the string to the structure, and can certainly have significant effects on transmission and playability.  Once the vibrations get to the body, that's the "efficient radiation" I meant.

we might ponder if the model of ‘efficient radiation’ is what performers really want. 

 All I am getting from high professionals is more about ‘sound control’. We talked before about energy loss when a violin gets too light. You explained that a too-light violin absorbs the energy too fast and only over-heavy bowing could kick it to a somehow acceptable sound.

i think this effect needs to be in a tiny degree on an instrument for best performance qualities. And for the same reason I think one of the most neglected parts in getting the ‘right’ sound out of an instrument are those setup parameters which can trigger this effect enormously.

Then, when I take just the word ‘efficient’ it is somehow like a smoothly running engine. What I get for example from genuine del gesu type of instruments is rather something which is ‘disfunctional in an interesting way’ making the kind of textured sound.

Link to comment
Share on other sites

13 hours ago, Peter K-G said:

Marty, 

The Finn is telling you, if the purpose is going on a date, you better sure you have a sexy instrument to show up with, do you? From what I'v seen your's are bigger than AVG, are they?

 

Modern big grand pianos are louder than the small ones made three hundred years ago.

But violins are still small little wimps-- therefore orchestras have to have many violin players to get enough sound.

If you made violins louder you wouldn't need as many players and many of them would become unemployed and many violin makers, shops, and schools would also go out of business.

That's why I'm getting death threats for making larger violins.

Link to comment
Share on other sites

  • 4 weeks later...

@Don Noon @Marty Kasprzyk @Anders Buen  @ctanzio @Carl Stross @sospiri @uncle duke @GeorgeH  @David Burgess @Peter K-G  @Bodacious Cowboy @jezzupe @David Beard @Evan Smith @JacksonMaberry @christian bayon @Violadamore@Rothwein @Roger Hill @martin swan

 

So I am back to this thread.

Question: Is it really coincidence that the area where Colin Goughs Stress diagram shows the highest stress is the zone where Don Noon measured most of the overtone output? 
 

With just common sense logic a thin zone under high stress should react more ‘nervously’ to vibrations than a a thick zone with no stress. 

Link to comment
Share on other sites

28 minutes ago, Andreas Preuss said:

With just common sense logic a thin zone under high stress should react more ‘nervously’ to vibrations than a a thick zone with no stress. 

Common sense doesn't always work that well for analyzing stress and vibrations.  One part I think is right, kindof... vibrations are generally stronger with less mass.  With "thin" equating to "less mass", then it sould be generally true.

Stress has nothing to do with it.  Loudspeakers are not stressed, yet respond very well to vibrations.  Only in specific cases will static stress matter... like a banjo head, where membrane tension makes all the difference in the vibration properties.  That is not the case for the curved shells of the violin.

What is this Colin Gough Stress Diagram you refer to?  And even without seeing it, I'd say yes it is a conicidence that static stress and active vibration areas appear related.

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...
  • Recently Browsing   0 members

    • No registered users viewing this page.



×
×
  • Create New...