How does a violin reproduce overtones? - Theorizing a model

[Anders Buen]

21 minutes ago, Peter K-G said:

Don,

Have you come up with any ideas on the just under 1000 Hz peak, that is anoyingly strong on some violins?

I get wolfs at B5 on two of mine, not to disturbing, but if bowed quickly it real wobbles.

On both it falls right between B1- and B1+ an octave above.

This is speculation, but maybe it has something to do with Hutchins humming tone that appears between B1- and B1+ (at their relative strength)

Add some mass to your inner f-hole wigs and see if it helps.

[Anders Buen]

6 hours ago, Peter K-G said:

Plucked => No bowhair noise 

There may be some from the finger before the string is free. And, a transient will by definition drive all frequencies, also between the string harmonics. The nature and spectrum there depends on the time history og the pluck. With nail or finger tip. 

[David Beard]

1 hour ago, Marty Kasprzyk said:

I'm quite sure the noise between a note's harmonics are produced by the bow hair. Attached is a plot up to 8000Hz of a bowed violin open G string as a blue line.  Also plotted is an orange line showing the bowing of the violin bridge upper bass side edge which produces a "hissy" kind of white noise.  

The violin body/bridge filters the bow noise just like it filters the string harmonics of a note and the orange line of the bowed bridge edge follows the same envelope shape as the noise of the bowed note.

Also attached is the same plot going up to only 600Hz.  The A0 resonance peak (~280Hz) shows up as noise just like Peter had found.  All violins probably do the same thing.

 

As you say, the strings, the instrument, the whole system is fundamentally an energy transducer.

The primary (basically only) energy input is the essentially random stream of pull/slips from the bow action.

This source energy can be thought of as noise, an essentially random stream of energy kicks -- white noise.

Some of this white noise source energy will pass through the whole transducer and radiate as still simple white noise.

The big question for all stages of the instrument is 'how much of the white source energy falls into musical signal and how?'

A good part of the energy falls into signal initially at the strings. The quality of the strings and the player's action matter greatly at this stage.  And then, a portion of this signal energy will pass into the instrument through the bridge which will act as a filter.

But also, a portion of the white noise energy will pass into instrument still as white noise.  The question we don't ask enough is how well does the instrument keep capturing that noise energy into the signal?

 

[Peter K-G]

18 minutes ago, Anders Buen said:

Add some mass to your inner f-hole wigs and see if it helps.

I'll try that

[Peter K-G]

15 minutes ago, Anders Buen said:

There may be some from the finger before the string is free. And, a transient will by definition drive all frequencies, also between the string harmonics. The nature and spectrum there depends on the time history og the pluck. With nail or finger tip. 

True!

Even setting up for playing and "grabbing" the neck they show up on realtime FFT, very sensitive to any kind of excitement.

[David Burgess]

11 minutes ago, David Beard said:

 

The primary (basically only) energy input is the essentially random stream of pull/slips from the bow action.

This source energy can be thought of as noise, an essentially random stream of energy kicks -- white noise.

I will disagree. Initially, the source energy can be thought of as noise, but after even one round trip of the traveling string kink between the bridge and upper nut (or finger), a pattern starts to be established, which kicks the bow hair loose from the string in a regular pattern  corresponding with the frequency of the played note

[Peter K-G]

And now it's past bed time again, on the better side of the Atlantic, where the sun sets in the ocean.

Well, The gulf of bothnia to be presice, a couple of minor countries like Sweden and Norway (which are going to loose in hockey) in the way

Good night

[Anders Buen]

35 minutes ago, David Beard said:

The primary (basically only) energy input is the essentially random stream of pull/slips from the bow action.

This source energy can be thought of as noise, an essentially random stream of energy kicks -- white noise.

There may be a randomish behavior at the startup, as David B writes, then the behavior becomes periodic and phase locked. There is no randomness to that basic bowed string sound from a trained player.

i think part of the noise comes from the slip phase, but I an not sure. The noise may be randomish in some way.

[David Beard]

1 hour ago, David Burgess said:

I will disagree. Initially, the source energy can be thought of as noise, but after even one round trip of the traveling string kink between the bridge and upper nut (or finger), a pattern starts to be established, which kicks the bow hair loose from the string in a regular pattern  corresponding with the frequency of the played note

Yes. Where is the disagreement?  Unless you are claiming that the dynamic you describe is 100%.  On that we would disagree.  

[David Beard]

16 minutes ago, Anders Buen said:

There may be a randomish behavior at the startup, as David B writes, then the behavior becomes periodic and phase locked. There is no randomness to that basic bowed string sound from a trained player.

i think part of the noise comes from the slip phase, but I an not sure. The noise may be randomish in some way.

The noise comes from the imperfection of the phase lock you describe.  And yes, it greatly depends on the player.  

But what you describe is a state that emerges. It is not 100% of the action.

And, unless there is plucking or banging, 100% of the energy, including all noise, is originating with the bow action.

Bows also partial skate.  And don't forget, scrunch and other noise are deliberate parts of player bow articulations.

[David Burgess]

40 minutes ago, David Beard said:

Bows also partial skate.  And don't forget, scrunch and other noise are deliberate parts of player bow articulations.

Except when they are not.

Some players highly value a fiddle which can easily do note and bow direction changes rather seamlessly, and others place higher value on consonance and articulation

[jezzupe]

1 hour ago, David Burgess said:

I will disagree. Initially, the source energy can be thought of as noise, but after even one round trip of the traveling string kink between the bridge and upper nut (or finger), a pattern starts to be established, which kicks the bow hair loose from the string in a regular pattern  corresponding with the frequency of the played note

How is this possible do you think?, I'm not disagreeing, just wondering how something with such little mass {the string} could "send" energy transfer back to the bow, {way more mass} and start to drive the bow hair? or dictate a patterned skipping or skating, is it that the bow hairs act as individuals in such a dynamic state? 

[David Beard]

9 minutes ago, jezzupe said:

How is this possible do you think?, I'm not disagreeing, just wondering how something with such little mass {the string} could "send" energy transfer back to the bow, {way more mass} and start to drive the bow hair? or dictate a patterned skipping or skating, is it that the bow hairs act as individuals in such a dynamic state? 

No. It is the cycle of stick slip that tends to fall into pattern.

But, noise are still in the sound.  And virtually the energy comes via the bow.

And any noise in the system, of whatever origin, is energy that has the potential to be captured into the signal.

 

[David Burgess]

19 minutes ago, jezzupe said:

How is this possible do you think?, I'm not disagreeing, just wondering how something with such little mass {the string} could "send" energy transfer back to the bow, {way more mass} and start to drive the bow hair? or dictate a patterned skipping or skating,

What is your impression from this slow-motion video?

 

[jezzupe]

9 minutes ago, David Beard said:

No. It is the cycle of stick slip that tends to fall into pattern.

But, noise are still in the sound.  And virtually the energy comes via the bow.

And any noise in the system, of whatever origin, is energy that has the potential to be captured into the signal.

 

D. Burgess wrote "a regular pattern  corresponding with the frequency of the played note"

So that means it's matching the frequency cycle of what ever note is being played? and therefore will change note to note? 

[David Beard]

1 minute ago, jezzupe said:

D. Burgess wrote "a regular pattern  corresponding with the frequency of the played note"

So that means it's matching the frequency cycle of what ever note is being played? and therefore will change note to note? 

That is the tendency. But reality includes transitions and imperfections and complications.

Even that the most skilled players include 'consonances' in their playing. It isn't all vowels.

Just plain not as simplistic as described.

[jezzupe]

4 minutes ago, David Burgess said:

What is your impression from this slow-motion video?

 

Hmm yes, like the string is the road and there is a patterned earthquake you trying to walk across.

[Don Noon]

3 hours ago, Peter K-G said:

Don,

Have you come up with any ideas on the just under 1000 Hz peak, that is anoyingly strong on some violins?

2 hours ago, Anders Buen said:

Add some mass to your inner f-hole wigs and see if it helps.

The mode shape I have known for a long time, both from my own testing and the Strad3D animations.  Yes, the inner trebel wing moves a lot, as does the trebel side of the upper bout.

Less clear what what to do to attenuate it.  I tried many things, mostly unsuccessful, to knock it down on instruments where it was annoyingly strong.  It seemed to move around my "fixes", but not go away significantly.  For the last several years, I have not had a problem with it, which I can only attribute to more careful study of Cremonese archings and perhaps some graduation changes (I used to go thin in the bout cheeks, now I don't... i.e. reverse graduations in the upper bout).

 

[Marty Kasprzyk]

1 hour ago, Anders Buen said:

There may be a randomish behavior at the startup, as David B writes, then the behavior becomes periodic and phase locked. There is no randomness to that basic bowed string sound from a trained player.

i think part of the noise comes from the slip phase, but I an not sure. The noise may be randomish in some way.

Hi Anders,

I bowed the G string very close to the bridge by tilting the bow so that only a narrow band of hair touched the string.  If I used very little bow force I was able to get the bow to slide across the string without producing the G note--all I got was the same kind of noise that bowing on the bridge edge produced.   ApparentlyI was using a bow force below the "minimum bow force" line in Schelleng's  famous bow force vs. bow-to-bridge distance diagram.

The noise again had some higher peaks at the A0 frequency like Peter had found but the peaks were randomly spaced again.

So I think you're right in thinking noice comes from the fly back slip phase in the bow/string interaction.  The video David Burgess gave us shows the bow hair stays in contact with the string.

I wonder if good and bad violins have different signal to noise ratios.

[Evan Smith]

On 5/26/2021 at 4:08 PM, Don Noon said:

I'd say my recent test replacing a normal maple back with one made of MDF convinced me that the top is definitely the most important, as the general tonal character remained, although the signature modes changed significantly.

When someone parts out a scrapped Strad, you can keep the back and I'll take the top.

I'm not so easily convinced.

Replacing a normal maple back with a piece of mdf and hearing the same tonal characteristics doesn't surprise me at all. I really don't consider that any kind of a test.

Replacing the back of a fantastic violin with mdf will show you what you loose.

Using a back that works to it's full potential still will not change the general tonal character of the top either, but it will bring out things and fill out the sound with a dimension unattainable with something of lesser quality.

Another Dimension Don,,, the twilight zone,, the realm of inanimate possibilities and supernatural fiddles.

My first foray into this world was about 30 years ago. I destroyed a lot of violins learning about them, I had a lot of questions that no one could answer,,,I like answers, ones that give simple solutions to relevant problems, I'm not about to make a fiddle from scratch until I get a handle on this thing.

There was this old German thing with a horrible nasty spot on the D string,,,I spent  a lot of work trying to figure it out,, three tops later,, nothing had changed. I had foolishly believed that the back couldn't cause that much trouble. I made a new back and the problem was completely gone. The back had some strange grain to it, it just didn't work right, it was plenty thick. I have glued struts on the outside of backs and made fiddles work fantastic with the help of some really good professional players. But struts on the outside cannot be duplicated on the inside. It just shows what is possible to accomplish with a back plate.

I have done lots of tests with back plates, and I can't see anything significant in the mdf test.

I'm just sayin that there is more to it,, that's all.

[Don Noon]

58 minutes ago, Evan Smith said:

I have done lots of tests with back plates, and I can't see anything significant in the mdf test.

I'm just sayin that there is more to it,, that's all.

The MDF back certainly made a difference... bad, IMO, but there was still a lot of the original character remaining.  I also have a MDF top plate (but needs some work to complete the F's and surrounding grads).  I can put the original back on again, and replace the top plate with MDF.  Yawanna bet on which version sounds the worst or changes the most?

But I really don't have time for any more of these fun tests that don't contribute to finishing the orders waiting.

[Andreas Preuss]

1 hour ago, Don Noon said:

The MDF back certainly made a difference... bad, IMO, but there was still a lot of the original character remaining. 

 

2 hours ago, Evan Smith said:

Using a back that works to it's full potential still will not change the general tonal character of the top either, but it will bring out things and fill out the sound with a dimension unattainable with something of lesser quality.

I see the back as a sort of equalizer of the sound. So there are many variables between too thick and too thin which can work well as long as the center is thick enough. But I would do more fine tuning of the sound on the top than on the back, because it is in my view the primary sound source. All the rest of the instrument including the back acts as a modifier.

[Peter K-G]

This is now kinda confirmed:

The body modes are only shown because of 

- Bow noise, slip, attack etc.

- Plucked, start, release etc.

 

Open G plucked, recording started when it's just the string vibrating (close distance and yet pretty clean)

 

[Anders Buen]

22 minutes ago, Peter K-G said:

This is now kinda confirmed:

The body modes are only shown because of 

- Bow noise, slip, attack etc.

- Plucked, start, release etc.

 

Open G plucked, recording started when it's just the string vibrating (close distance and yet pretty clean)

 

Probably not quite correct, becuse the bridge and body «see» or get a «kick» every time the triangle on the string hits the bridge, and get reflected there. 

Edited May 28, 2021 by Anders Buen
Corrected bridge to string

[Peter K-G]

Yup