Jump to content
Maestronet Forums

William Fry Internal Scraping Method


dpappas

Recommended Posts

3 hours ago, Violadamore said:

I'll note that surface trimming the outside has nothing to do with Fry's interior prestidigitation with sandpaper.

If you can remove material from the outside, but then you have to re-varnish, then removing material from the inside instead would seem like a good idea.

So apparently the reason why what Fry did was not effective as claimed was, in addition to his theories mistaken, thus leading him to attempt to remove material perhaps from the wrong places, because he did not remove enough material to make a difference, and could not do so, given the kind of tool he had made.

Of course, it might be difficult to construct the kind of tool that would let you remove material from the inside with the amount of care and control required. A tool that could do that would presumably be a rather impressive mechanism, and obviously not something that Stradivari had access to.

Link to comment
Share on other sites

  • Replies 194
  • Created
  • Last Reply

Top Posters In This Topic

29 minutes ago, edi malinaric said:

Hi Violadamore and Don - a few comments...

1.0 "major refinishing" - tuning the body during the making stage doesn't involve any great changes.If anything, it's probably quicker to tune to the "tap" than carve to predetermined thicknesses. Knowing what and to tune to and where to tune it may well be "The Secret".

1.1 I suspect that varnishing -then becomes the final tuning step.

2.0 "surface trimming" vs "Fry's I P" - actually they are exactly analogous.

Stiffness is proportional to the thickness raised to the third power - so a little difference goes a long way. For example assume a thickness of 3.4mm, then thin it by 0.1mm to 3.3mm - the stiffness will change by 8.6% which is not a small amount.

The stiffness will care not a hoot whether the change comes from the inside or the outside.

3.0 "Bells and violins" - since both vibrate to create sound, they do share some things in common.

4.0 On "casting cannon from violins" - one could seriously consider casting violas from a cannon.

cheers edi

 

 

1.0  Remember that I don't make, I repair and restore, or as I used to put it, "resurrect".  Anything I scrape off, I have to cover up again.  I was noting why I hadn't done more with this concept.

2.0  I beg to differ.  While your engineering math is fine, you're forgetting that the material is inhomogeneous, anisotropic, has pronounced directional properties, and contains continuous fibers.  Where you cut the strands in a composite can definitely matter, right?  The other point to consider is related, in that removing material under compression can have different effects from removing material under tension.  Until proven otherwise under repeatable laboratory conditions, I wouldn't risk equating one method with the other.

4.0  Maybe from bronze, that could look rather attractive.  :lol:

If, OTOH, you're suggesting involving me in a circus stunt, :P.

Link to comment
Share on other sites

4 minutes ago, Violadamore said:

1.0  Remember that I don't make, I repair and restore, or as I used to put it, "resurrect".  Anything I scrape off, I have to cover up again.  I was noting why I hadn't done more with this concept.

2.0  I beg to differ.  While your engineering math is fine, you're forgetting that the material is inhomogeneous, anisotropic, has pronounced directional properties, and contains continuous fibers.  Where you cut the strands in a composite can definitely matter, right?  The other point to consider is related, in that removing material under compression can have different effects from removing material under tension.  Until proven otherwise under repeatable laboratory conditions, I wouldn't risk equating one method with the other.

4.0  Maybe from bronze, that could look rather attractive.  :lol:

If, OTOH, you're suggesting involving me in a circus stunt, :P.

Hi Violadamore - I wasn't casting nasturtiums :-)

As to circus stunts, I tried to work in "Big Top" into a rejoinder and failed - mainly because it relied on knowing that the Croatian for a cannon is "Top". Can't win them all :-(

I beg to differ from your beg to differ (this could become interesting...) Since the plate surfaces are, for all practical purposes, parallel to each other - inside/outside would be interchangeable as far as their effect on vibrations.

One of the advantages of tap tuning is that it deals with the inhomogeneity of the wood with no hassle at all. One arrives at a desired result without needing any knowledge of the exact properties of the material or whether one is working on the inside or outside of the plate. The trick is in knowing what the endpoint is - aaah - the secret!

Out of curiosity I searched 'wood fibre data". Sugar Maple - fibre length ~ 0.85mm, length/diameter ratio ~ 210. So in removing 0.1mm one has severed ~ 25 out of ~840 fibres (plate thickness 3.4mm)  

cheers edi

Link to comment
Share on other sites

38 minutes ago, edi malinaric said:

The trick is in knowing what the endpoint is - aaah - the secret!

And this is why I have been inclined to think, despite the doubts of some who are knowledgeable here, that either trimming a violin from the outside in the white, or trimming a violin from the inside if feasible, has potential.

Just as it's more difficult to find the nodal lines of Chladni patterns of the plates on a fully-assembled violin - but apparently not impossible, according to Oded Kishony - it's easier to find videos on YouTube of people playing fully-assembled violins by Stradivary than any kind of sound recording of the tap tones of the isolated belly plate of a Stradivarius.

So one has a better chance of knowing what the desired end-point is with an assembled violin than with a plate; then one is still left with the problem of knowing from where it would be useful to remove material, so a secret is left, but one that could be learned from much experience.

Link to comment
Share on other sites

7 hours ago, edi malinaric said:

Stiffness is proportional to the thickness raised to the third power - so a little difference goes a long way. For example assume a thickness of 3.4mm, then thin it by 0.1mm to 3.3mm - the stiffness will change by 8.6% which is not a small amount.

Although as stated, this is correct, the frequency is only raised proportionally to the thickness, due to the added mass...for bending stiffness.

Violin plates are NOT totally ruled by bending stiffness, as there is also stretching stiffness due to the arch and the overall box-shape of the instrument.  It turns out that frequencies controlled by stretching stiffness are completely independent of thickness, and only ruled by the radius of curvature and the material properties.

So in theory, frequencies will be affected by thickness by a factor between 0 and 1.

In practice, thinning does move some mode frequencies downscale, and the lower frequencies become stronger.  However, I still stand by my experience that the overall shape of the middle and higher frequencies, and thus the major tonal character, remains recognizably the same, except for really massive thinning (say over 10 grams for a top).  

I think it is also telling that nobody (to my knowledge) has been able to objectively demonstrate a targeted tonal change by selective thinning.  I have tried several times, with no success.

7 hours ago, edi malinaric said:

"Bells and violins" - since both vibrate to create sound, they do share some things in common.

Yes, both have modes of vibration and move air to create sound.  However, it is also extremely important to know the differences, and not get attracted down a dead-end path by the useless similarities.  

A bell is a single-note instrument, excited by impact.  The goal is to get as many of its vibration modes as possible into harmony with each other.  

A violin needs to respond to a complete spectrum of vibrations.  Any "tuning" of modes to each other (the bell idea), if accomplished, would accentuate one note at the expense of anything else, and make for terribly uneven response.  So how do you  "tune" a violin to everything all at once, evenly?  Certainly  not by thinking about it as a bell.  Or a speaker cone.

Link to comment
Share on other sites

1 hour ago, Don Noon said:

A violin needs to respond to a complete spectrum of vibrations.  Any "tuning" of modes to each other (the bell idea), if accomplished, would accentuate one note at the expense of anything else, and make for terribly uneven response.  So how do you  "tune" a violin to everything all at once, evenly?  Certainly  not by thinking about it as a bell.  Or a speaker cone.

2

This is a job for Artificial Intelligence:ph34r:

Link to comment
Share on other sites

1 hour ago, Michael_Molnar said:

This is a job for Artificial Intelligence:ph34r:

I'd rather go with experience and expert player evaluation.  Logic runs into a brick wall at the end, when humans decide what's good and what's not.  And in the middle, there's infinite complexity. :)

 

1 hour ago, Davide Sora said:

@Don Noon

I love your answers B)

I love your videos. :wub:

Link to comment
Share on other sites

38 minutes ago, Don Noon said:

I'd rather go with experience and expert player evaluation.  Logic runs into a brick wall at the end, when humans decide what's good and what's not.  And in the middle, there's infinite complexity. :)

Right but in general AI solutions to the "recommendation problem" are much better than what humans could do.  The first obvious problem here is no data to work with and no way to get it :)

~~~

I thought it was interesting in one of those videos above when the player lady says that at first she didn't know if a certain violin was the best she'd ever tried or the worst.

Link to comment
Share on other sites

8 hours ago, edi malinaric said:

Out of curiosity I searched 'wood fibre data". Sugar Maple - fibre length ~ 0.85mm, length/diameter ratio ~ 210. So in removing 0.1mm one has severed ~ 25 out of ~840 fibres (plate thickness 3.4mm)  

cheers edi

Ummmm.....tabulated data is nice, as far as it goes, but making a chair seat or a basket out of splits will indicate something else entirely.  Start pulling a sliver of spruce loose with hemostat or forceps, and see what happens.  Whatever the book says, the structures parallel to the grain can continue for long distances, and have a lot of tensile strength. :)                                                                                                              

Link to comment
Share on other sites

11 minutes ago, Violadamore said:

Ummmm.....tabulated data is nice, as far as it goes, but making a chair seat or a basket out of splits will indicate something else entirely.  Start pulling a sliver of spruce loose with hemostat or forceps, and see what happens.  Whatever the book says, the structures parallel to the grain can continue for long distances, and have a lot of tensile strength. :)                                                                                                              

Hi Violadamore - much the same question that was rattling around in my mind. Fibre length/cell length? How do the cells chain themselves together longitudinally?

Recently I was looking into the length of flax and hemp fibres and marveling about Dip.Eng. Nature.

cheers edi

Link to comment
Share on other sites

3 hours ago, Don Noon said:

Although as stated, this is correct, the frequency is only raised proportionally to the thickness, due to the added mass...for bending stiffness.

Violin plates are NOT totally ruled by bending stiffness, as there is also stretching stiffness due to the arch and the overall box-shape of the instrument.  It turns out that frequencies controlled by stretching stiffness are completely independent of thickness, and only ruled by the radius of curvature and the material properties.

So in theory, frequencies will be affected by thickness by a factor between 0 and 1.

In practice, thinning does move some mode frequencies downscale, and the lower frequencies become stronger.  However, I still stand by my experience that the overall shape of the middle and higher frequencies, and thus the major tonal character, remains recognizably the same, except for really massive thinning (say over 10 grams for a top).  

I think it is also telling that nobody (to my knowledge) has been able to objectively demonstrate a targeted tonal change by selective thinning.  I have tried several times, with no success.

Yes, both have modes of vibration and move air to create sound.  However, it is also extremely important to know the differences, and not get attracted down a dead-end path by the useless similarities.  

A bell is a single-note instrument, excited by impact.  The goal is to get as many of its vibration modes as possible into harmony with each other.  

A violin needs to respond to a complete spectrum of vibrations.  Any "tuning" of modes to each other (the bell idea), if accomplished, would accentuate one note at the expense of anything else, and make for terribly uneven response.  So how do you  "tune" a violin to everything all at once, evenly?  Certainly  not by thinking about it as a bell.  Or a speaker cone.

Hi Don - thanks for your note - thinking time ;-)

cheers edi

Link to comment
Share on other sites

9 hours ago, Michael_Molnar said:

This is a job for Artificial Intelligence.

It is true that a violin is not a bell; a bell is designed to have, as he noted, all the resonances either at the same frequency, or harmonically-related, so it will ring with a pure tone. The Stelzner violin, designed with elliptical, parabolic, and maybe hyperbolic surfaces so as to have a lot of reflections the same length, was mistakenly designed to be like a bell.

And indeed a violin is not a speaker cone; designed not to have any resonances in the area of interest, so that the sound that comes out will accurately resemble the sound that comes in as closely as possible. That doesn't mean that thinning plates around the edges is necessarily all bad, though.

But we don't need to go to Google's Alpha Zero or IBM's Watson to find out what a violin is. A violin is designed to use resonances to more effectively couple the vibrations of the strings to the air. It's all right if this means the frequency response is uneven, since that just makes vibrato more interesting.

So you basically want lots of resonant peaks, starting at the frequency of the open G string (just below 200 Hz). They should be closely spaced and strong. The peaks should continue on up to also affect the overtones.

Ideally, the violin should have a "Bridge-Hill", with the peaks from 1700 Hz to 3200 Hz, approximately, being higher than those above or below.

As far as the fundamentals of the notes on the violin go, however, the violin doesn't start having a constant stream of closely-spaced strong peaks until about 800 Hz or so. Below that, there will be three big wide peaks at about 200, 400, and 500 Hz, with a lot of other fairly tiny peaks. That can't be helped, that is just the way a violin sounds.

Link to comment
Share on other sites

50 minutes ago, Quadibloc said:

It is true that a violin is not a bell; a bell is designed to have, as he noted, all the resonances either at the same frequency, or harmonically-related, so it will ring with a pure tone. The Stelzner violin, designed with elliptical, parabolic, and maybe hyperbolic surfaces so as to have a lot of reflections the same length, was mistakenly designed to be like a bell.

And indeed a violin is not a speaker cone; designed not to have any resonances in the area of interest, so that the sound that comes out will accurately resemble the sound that comes in as closely as possible. That doesn't mean that thinning plates around the edges is necessarily all bad, though.

But we don't need to go to Google's Alpha Zero or IBM's Watson to find out what a violin is. A violin is designed to use resonances to more effectively couple the vibrations of the strings to the air. It's all right if this means the frequency response is uneven, since that just makes vibrato more interesting.

So you basically want lots of resonant peaks, starting at the frequency of the open G string (just below 200 Hz). They should be closely spaced and strong. The peaks should continue on up to also affect the overtones.

Ideally, the violin should have a "Bridge-Hill", with the peaks from 1700 Hz to 3200 Hz, approximately, being higher than those above or below.

As far as the fundamentals of the notes on the violin go, however, the violin doesn't start having a constant stream of closely-spaced strong peaks until about 800 Hz or so. Below that, there will be three big wide peaks at about 200, 400, and 500 Hz, with a lot of other fairly tiny peaks. That can't be helped, that is just the way a violin sounds.

I recommend that the resonance peaks should never be octaves apart (261.6, 523.3, 1047 etc, C note for example).  This would cause some notes be overly loud (maybe wolfs) because all of its harmonics would have high amplitudes.  The valleys between the peaks should also not be octaves apart which would cause some notes to be very weak.  

One of the qualities of a good violin is that it has a fairly even note loudness so that the player doesn't have to constantly make bowing adjustments from one note to the next.

The frequencies of the resonance peaks and valleys hence their spacings appears to be hard to control and seems to be a matter of chance given the variability of wood.  Even great makers like Strad sometimes had good and not so good spacings hence good and not so good violins which nobody wants to talk about.  (oops I forgot, "great" and "not quite so great")

 

 

Link to comment
Share on other sites

1 hour ago, Marty Kasprzyk said:

I recommend that the resonance peaks should never be octaves apart ...

Although I agree that it would be nice not to have the octave spacing, I also have experience indicating it's not all that critical... at least for A0/B1+ alignment... as long as the rest of the overtones are strong enough so that the signature modes aren't the only things you hear.

My two highest placing violins in tone competition at VMAAI (one currently played by Annelle Gregory) both had A0 and B1+ an octave apart at C#.

It may be that the lower frequency resonances aren't as important, as the ear is less sensitive at those frequencies... and if there are peaks and dips at higher frequencies, those alignments may be more critical.  But those higher resonances and dips are also far more difficult to control.

Link to comment
Share on other sites

2 hours ago, Marty Kasprzyk said:

Even great makers like Strad sometimes had good and not so good spacings hence good and not so good violins which nobody wants to talk about.  (oops I forgot, "great" and "not quite so great")

Given that the owners of the not so good ones have a lot of money tied up, I suppose they have an interest in word not getting out. Although I think your reasoning is sound, I suspect Don Noon is right, and the two peaks in question are almost unavoidably in a harmonic or near-harmonic relationship on most violins. But at least they're broad peaks, with a low Q, and that may help.

Link to comment
Share on other sites

3 hours ago, Don Noon said:

It may be that the lower frequency resonances aren't as important, as the ear is less sensitive at those frequencies... and if there are peaks and dips at higher frequencies, those alignments may be more critical.  But those higher resonances and dips are also far more difficult to control.

It depends on the loudness / distance from the violin. At usual sound pressure levels achieved by the violin the b1 frequency is in one of the most sensitive regions. The sensitivity starts to decrease significantly below the A0 frequencies. Sorry for the German language, this was the easiest way to post an isoprene graph.

IMG_0222.JPG.39897124a3de5b2d782b14e86279746b.JPG

Link to comment
Share on other sites

I made a Hardanger fiddle using my Native King Billy Pine , I thought the top could be a bit thicker so I glue a big spruce patch between sound holes and replaced bass bar with a spruce one , when I strung it up again the sound had changed a little but I was surprised the character was still the same, not only did I change thickness but wood species as well

Link to comment
Share on other sites

8 hours ago, Don Noon said:

Although I agree that it would be nice not to have the octave spacing, I also have experience indicating it's not all that critical... at least for A0/B1+ alignment... as long as the rest of the overtones are strong enough so that the signature modes aren't the only things you hear.

My two highest placing violins in tone competition at VMAAI (one currently played by Annelle Gregory) both had A0 and B1+ an octave apart at C#.

It may be that the lower frequency resonances aren't as important, as the ear is less sensitive at those frequencies... and if there are peaks and dips at higher frequencies, those alignments may be more critical.  But those higher resonances and dips are also far more difficult to control.

Does she especially like the C# note?

Link to comment
Share on other sites

35 minutes ago, Marty Kasprzyk said:

Does she especially like the C# note?

She didn't mention it, and I didn't notice that note jumping out either in her playing or the several months that I had that fiddle myself.  Here's my favorite solo piece, and you can listen for yourself to see if you can find the A0/B1+ alignment.  Maybe it is there, but so what.

What she did mention, and I think is noticeable, is the power overall on the G and D, which I don't think is a function of the A0 and B1+, but the cumulative effect of all the other overtones, and there are lots of them.

So maybe it would be a tiny bit better without the octave spacing of these major resonances... but I wouldn't want to compromise anything else to get it.

 

Link to comment
Share on other sites

On 8/14/2018 at 3:55 AM, Quadibloc said:

And this is why I have been inclined to think, despite the doubts of some who are knowledgeable here, that either trimming a violin from the outside in the white, or trimming a violin from the inside if feasible, has potential.

Just as it's more difficult to find the nodal lines of Chladni patterns of the plates on a fully-assembled violin - but apparently not impossible, according to Oded Kishony - it's easier to find videos on YouTube of people playing fully-assembled violins by Stradivary than any kind of sound recording of the tap tones of the isolated belly plate of a Stradivarius.

So one has a better chance of knowing what the desired end-point is with an assembled violin than with a plate; then one is still left with the problem of knowing from where it would be useful to remove material, so a secret is left, but one that could be learned from much experience.

Of course you are right, in my opinion. I never had a strong enough stomach, or bladder, to finish Fry's infamous video but it doesn't take long to see what he thought he was doing. He believed in his method enough for everyone. Well...maybe his tool and method were caca, but who hasn't regraduated instruments to get a better result? Who has?

I am surprised that no one admits to leaving key areas a bit thick and shaving the areas down as needed later--as a de facto working method (not as an afterthought years later). But not like Fry, cheap and dirty and all voodoo magick fakery, but actually popping the top and re-setting up etc. Either no one has the patience or no one will admit to a working method that has some amount of uncertainty and subjectivity and slippery slopery built in.  Obviously it's not possible to organically add wood back into a too-thin area so would it not make sense to take some precautions especially if the spruce is all weird acting? 

I'm not one to ask either, because it totally makes sense to me...but not the way Fry did it. 

 

Link to comment
Share on other sites

18 hours ago, Michael Szyper said:

It depends on the loudness / distance from the violin. At usual sound pressure levels achieved by the violin the b1 frequency is in one of the most sensitive regions. The sensitivity starts to decrease significantly below the A0 frequencies. Sorry for the German language, this was the easiest way to post an isoprene graph.

IMG_0222.JPG.39897124a3de5b2d782b14e86279746b.JPG

What does the red line show?  The only German I can remember is ein grosse bier

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.



×
×
  • Create New...