Jack Devereux

Basic Acoustics Resource

Recommended Posts

1 hour ago, curious1 said:

Yes, I use Don's method of measuring the speed of sound in the wood which is the same as the method in the study above.

It is not perfectly clear how the "Hitman HM200" is determining acoustic velocity.  It would either have to use a frequency analyzer or a pulse time interval measurement, but the end result should be the same (assuming straight, uniform logs).  With frequency measurement and our size of wood, the real effect that is being measured is the first longitudinal mode of vibration of the sample, which needs to have a uniform cross-section (i.e. extruded shape) for accuracy.  I don't know the limitations and accuracy of the pulse interval method, if that is even used.

Then, with measured density and the formula shown in the article, absolute longitudinal stiffness (modulus) can be calculated.  Simple.  Then it gets complicated again if you want to deal with bending stiffness or bending frequencies.

Share this post


Link to post
Share on other sites
1 hour ago, Don Noon said:

It is not perfectly clear how the "Hitman HM200" is determining acoustic velocity.  It would either have to use a frequency analyzer or a pulse time interval measurement, but the end result should be the same (assuming straight, uniform logs).  With frequency measurement and our size of wood, the real effect that is being measured is the first longitudinal mode of vibration of the sample, which needs to have a uniform cross-section (i.e. extruded shape) for accuracy.  I don't know the limitations and accuracy of the pulse interval method, if that is even used.

Then, with measured density and the formula shown in the article, absolute longitudinal stiffness (modulus) can be calculated.  Simple.  Then it gets complicated again if you want to deal with bending stiffness or bending frequencies.

Here's the abstract:   

https://wfs.swst.org/index.php/wfs/article/download/1650/1650

Share this post


Link to post
Share on other sites
37 minutes ago, curious1 said:

Yes, the details are there... it uses the longitudinal resonance frequency, not the time-of-travel method.  Interestingly, they  say the device actually uses  the second harmonic, but perhaps the first harmonic in the lumber of that size is too low of a frequency to measure well.

Share this post


Link to post
Share on other sites
On 9/8/2017 at 1:41 PM, Jim Bress said:

Given the five variables, density, arch height, thickness, plate weight, and stiffness, how would you (anyone) rank their effect size on the outcome on achieving your acoustical target?  >>>>

Just out of curiosity:  Why didn't you include violin outline  shape and size as variables?  Strad made many variations of these.  Why can't we?

Share this post


Link to post
Share on other sites
On 9/9/2017 at 1:41 PM, curious1 said:

If I had a piece of spruce with a modulus in the 9,000 -10,000 range and I was graduating it less than 2.4mm I'd start to worry if it could hold its shape.

I think an experienced maker would know when they are headed into the red. The trick is giving it a name.

 

On 9/10/2017 at 6:13 AM, Davide Sora said:

May I ask why?

I mean, regardless of type arching with wich you are dealing with?

I would be pleased to find out the trick to give a name to the maker.....:);)

Here is an example David, of why it might matter, drawn from real life.

Antonio Stradivari Huberman/Gibson 1713

Top density .35g/cc calculated by Steven Sirr based on CT scans. (If speed of sound does roughly follow density, the density of spruce is between .3g/cc-.5g/cc and the speed of sound 4800-6000m/s. Based on density it should be ~5200-5400m/s ?. It's modulus based on my formula 5200m/s^2 x .35g/cc = 9,464,000-10,200,000).

This violin fits into my criteria for low stiffness, low density, and thin graduations. Not withstanding it's fame it's arching has clearly collapsed and is being held up in part by the bell patch in the central area (my guess would also be that their was some arch correction when the violin was restored in the mid 1980s.

I am not a maker with a background in science. I know enough to get myself in to trouble but not enough to get myself out. 

So bring on the scorn and ridicule. :^)

 

IMG_8773.PNG

IMG_8775.PNG

IMG_8774.PNG

IMG_8771.PNG

IMG_8776.PNG

Share this post


Link to post
Share on other sites
9 hours ago, Marty Kasprzyk said:

Just out of curiosity:  Why didn't you include violin outline  shape and size as variables?  Strad made many variations of these.  Why can't we?

Hi Marty, For a particular model (so far I've only used the "P" form) the outline is a constant. For a different model/outline I would use a different arch height based on the model, then use the variables to tweak the arch height if I thought the wood characteristics warranted it.  Still a guessing game for me.  Taking notes and making more violins.  Eventually I hope to understand enough to make to make well reasoned decisions.  BTW, thanks for the article link.  Very interesting.

-Jim

Share this post


Link to post
Share on other sites
9 hours ago, curious1 said:

 

Here is an example David, of why it might matter, drawn from real life.

Antonio Stradivari Huberman/Gibson 1713

Top density .35g/cc calculated by Steven Sirr based on CT scans. (If speed of sound does roughly follow density, the density of spruce is between .3g/cc-.5g/cc and the speed of sound 4800-6000m/s. Based on density it should be ~5200-5400m/s ?. It's modulus based on my formula 5200m/s^2 x .35g/cc = 9,464,000-10,200,000).

This violin fits into my criteria for low stiffness, low density, and thin graduations. Not withstanding it's fame it's arching has clearly collapsed and is being held up in part by the bell patch in the central area (my guess would also be that their was some arch correction when the violin was restored in the mid 1980s.

I am not a maker with a background in science. I know enough to get myself in to trouble but not enough to get myself out. 

So bring on the scorn and ridicule. :^)

For violin tops I was thinking that the normal range of density was 0.33 g/cc - 0.40 g/cc, and for backs (maple) something like 0.50 g/cc - 0.65 g/cc.  That's my frame of reference for low, middle, high density wood.  Feel free to correct me if I'm wrong here.  Using the formula you provided, what would be considered withing the "normal" range of stiffness?

Thanks,

Jim

Share this post


Link to post
Share on other sites
13 minutes ago, Michael_Molnar said:

How does the Huberman/Gibson sound? If it's a dog, end of discussion.

Listen to Joshua Bell.  He plays it.  To my ear, it sounds nice and responsive, but lacks something in the sparkle department that I'd like to hear.  In the Strad article on this instrument, Sam Z. hints that the high end has been a bit weak.

Share this post


Link to post
Share on other sites
1 hour ago, Jim Bress said:

For violin tops I was thinking that the normal range of density was 0.33 g/cc - 0.40 g/cc, and for backs (maple) something like 0.50 g/cc - 0.65 g/cc.  That's my frame of reference for low, middle, high density wood.  Feel free to correct me if I'm wrong here.  Using the formula you provided, what would be considered withing the "normal" range of stiffness?

Thanks,

Jim

Spruce density .3-.5g/cc, speed of sound ~4800-6000m/s

maple density ~.55-.75g/cc, speed of sound ~3500-4500m/s

average would be for me right in the middle of those ranges.

Share this post


Link to post
Share on other sites
39 minutes ago, Michael_Molnar said:

Curious1,

How does the Huberman/Gibson sound? If it's a dog, end of discussion.

 

26 minutes ago, Don Noon said:

Listen to Joshua Bell.  He plays it.  To my ear, it sounds nice and responsive, but lacks something in the sparkle department that I'd like to hear.  In the Strad article on this instrument, Sam Z. hints that the high end has been a bit weak.

Warm and flexible is think how Sam Z. describes it.

My point in bringing it up as an example is there is an interplay between the stiffness and arching/graduations. That the Gibson has strayed outside the norms (whether by the maker's intent or by over zealous regraduation we cannot know) is indicated by the bell patch and arch shape. And perhaps by the tone.

The thread is titled Basic Acoustic Resources. I think measuring the modulus of the wood is a simple procedure that can benefit ones work when applied even in the most general of ways.

Share this post


Link to post
Share on other sites
4 minutes ago, curious1 said:

I think measuring the modulus of the wood is a simple procedure that can benefit ones work when applied even in the most general of ways.

I definitely go along with the idea of knowing the properties of the material you use, if only to look back over ones that turned out good (or bad) to see what you did, and perhaps to help guide what you use in the future.  Knowing stuff is usually  better than not knowing, unless you obsess over it and let the numbers take over without getting a feel for what it all means.

I don't pay much attention to modulus, although it's a simple calculations and it is on my spreadsheet.  I mostly look at density, speed of sound, and radiation ratio when deciding what wood I'm going to use.  It's kinda all related, though.  I'm still playing around with arching, and don't have any rules that I follow.

Share this post


Link to post
Share on other sites
9 minutes ago, curious1 said:

Spruce density .3-.5g/cc, speed of sound ~4800-6000m/s

maple density ~.55-.75g/cc, speed of sound ~3500-4500m/s

average would be for me right in the middle of those ranges.

 I tend to aim for the middle.  Given the broader range of "normal" it appears I've been aiming a bit low.  I would not have thought a 0.5 g/cc top or a 0.75 g/cc back would be usable.  I have a 0.75 g/cc single piece viola back billet (sugar maple) that I didn't think would be good for anything except furniture.  I guess the real point your're making is that it's the combination of density and speed of sound that needs to fall with in the normal range.  Thanks for the info.

-Jim

Share this post


Link to post
Share on other sites
24 minutes ago, Don Noon said:

I definitely go along with the idea of knowing the properties of the material you use, if only to look back over ones that turned out good (or bad) to see what you did, and perhaps to help guide what you use in the future.  Knowing stuff is usually  better than not knowing, unless you obsess over it and let the numbers take over without getting a feel for what it all means.

I don't pay much attention to modulus, although it's a simple calculations and it is on my spreadsheet.  I mostly look at density, speed of sound, and radiation ratio when deciding what wood I'm going to use.  It's kinda all related, though.  I'm still playing around with arching, and don't have any rules that I follow.

Do you ignore, Q, the damping factor?

Share this post


Link to post
Share on other sites
23 minutes ago, Jim Bress said:

 I tend to aim for the middle.  Given the broader range of "normal" it appears I've been aiming a bit low.  I would not have thought a 0.5 g/cc top or a 0.75 g/cc back would be usable.  I have a 0.75 g/cc single piece viola back billet (sugar maple) that I didn't think would be good for anything except furniture.  I guess the real point your're making is that it's the combination of density and speed of sound that needs to fall with in the normal range.  Thanks for the info.

-Jim

If you went to the middle you'd have

spruce .4g/cc, 5400m/s

maple .65g/cc, 4000m/s

i would then calculate the stiffness using this formula which incorporates the density and speed of sound

speed of sound ^2 x density

spruce 11,664

maple 10,400

stuff above those is going to be above average and stuff below that obviously below average

if I was an amateur maker i would consider these to be average numbers that would accord well with these 

Top arch 15mm, thickness 2.7mm, weight ~68g w bar, pitch F-F#

Back arch 14.5mm, 4.5mm in center, 2.3-2.5mm in lungs, weight ~100g, pitch E-F.

 

 

Share this post


Link to post
Share on other sites

C1, As an amateur maker, I find the information you provided extremely valuable for helping me grow.  Thank you very much for your generosity.

-Jim

Share this post


Link to post
Share on other sites
2 hours ago, curious1 said:

Warm and flexible is think how Sam Z. describes it.

Actual quotes:  "Warm and resilient response"...  "... contrasted with the crystalline sound of some other Stradivari violins..."

Diplomatic speak for "would be better if it had more high end", which you'd expect in a published description of your client's violin.   In the segment "Adjusting the Huberman", Sam makes several references to efforts to increase brilliance.

1 hour ago, Michael_Molnar said:

Do you ignore, Q, the damping factor?

No, I look at that too.  I just thought that would be beyond most folks reading here, and I don't really DO anything about it, just look at it and note it for future reference... like just about everything else.

1 hour ago, curious1 said:

...

Top arch 15mm, thickness 2.7mm, weight ~68g w bar, pitch F-F#

Back arch 14.5mm, 4.5mm in center, 2.3-2.5mm in lungs, weight ~100g, pitch E-F.

I think these are pretty reasonable values, not far from what I'm doing.  I have been using higher arching, but gradually getting lower, and torrefied wood can get the same modulus with lower density, thus lighter plates (tbd if that's "better" or not).

Share this post


Link to post
Share on other sites
13 hours ago, curious1 said:

Here is an example David, of why it might matter, drawn from real life.

Antonio Stradivari Huberman/Gibson 1713

Top density .35g/cc calculated by Steven Sirr based on CT scans. (If speed of sound does roughly follow density, the density of spruce is between .3g/cc-.5g/cc and the speed of sound 4800-6000m/s. Based on density it should be ~5200-5400m/s ?. It's modulus based on my formula 5200m/s^2 x .35g/cc = 9,464,000-10,200,000).

This violin fits into my criteria for low stiffness, low density, and thin graduations. Not withstanding it's fame it's arching has clearly collapsed and is being held up in part by the bell patch in the central area (my guess would also be that their was some arch correction when the violin was restored in the mid 1980s.

I like to take the Huberman/Gibson as an exemple, because it almost perfectly match the P1705 form (there is the very useful ribs scan on the poster to testify this) that is one of the forms that I use for my violins, so I may have some direct "empirical" clue from my work.

The top arching height of this violin at 14,55 mm for me is too low seen the characteristics of the wood in question, what I would do to "improve" the sound (more brilliancy) and to assure a better resistance to deformation (collapse) would be to increase the arching height of about 1.5/2.0mm and reduce the radius of the central cross section bringing it from the actual 9.7 mm (roughly) to perhaps 8.7/8.5 mm radius (approaching these measures to that of the Betts as an example).

I don't know what changes in Modulus or Sound Velocity these modification may give, but surely there would be an increase in M5 tap tone frequency that I roughly may try to estimate in a full tone increase (this is a gamble, but empiricism involves these risks...:)) that I empirically relate to an increase in stiffness.

It would be interesting to know what frequencies the Huberman/Gibson free top plate has, but unfortunately it is not given to know, I may try to roughly estimate a frequency around E (330 Hz) or anyway in the range from E to F, that I associate with a warm and dark sounding violin (another gamble...).

From my modification I would expect a rise in frequency up to F# (370Hz) that I try to associate with more brilliancy.

This is not a tuning attempt but only an empirical way to associate frequency to stiffness, that of course may work only on a personal work known and metabolized.

With this I do not mean that the "scientific numbers" do not work, I also like to know them because having as much data as possible helps to understand,
but surely it is more likely that ancient makers were referring to things like tap tones frequency and not to abstract numbers, and this can still be valid today.
 

 

Share this post


Link to post
Share on other sites
2 hours ago, curious1 said:

if I was an amateur maker i would consider these to be average numbers that would accord well with these 

Top arch 15mm, thickness 2.7mm, weight ~68g w bar, pitch F-F#

Back arch 14.5mm, 4.5mm in center, 2.3-2.5mm in lungs, weight ~100g, pitch E-F.

 

 

Then by this definition I'm an Amateur :)

Share this post


Link to post
Share on other sites
1 hour ago, Davide Sora said:

I like to take the Huberman/Gibson as an exemple, because it almost perfectly match the P1705 form (there is the very useful ribs scan on the poster to testify this) that is one of the forms that I use for my violins, so I may have some direct "empirical" clue from my work.

The top arching height of this violin at 14,55 mm for me is too low seen the characteristics of the wood in question, what I would do to "improve" the sound (more brilliancy) and to assure a better resistance to deformation (collapse) would be to increase the arching height of about 1.5/2.0mm and reduce the radius of the central cross section bringing it from the actual 9.7 mm (roughly) to perhaps 8.7/8.5 mm radius (approaching these measures to that of the Betts as an example).

I don't know what changes in Modulus or Sound Velocity these modification may give, but surely there would be an increase in M5 tap tone frequency that I roughly may try to estimate in a full tone increase (this is a gamble, but empiricism involves these risks...:)) that I empirically relate to an increase in stiffness.

It would be interesting to know what frequencies the Huberman/Gibson free top plate has, but unfortunately it is not given to know, I may try to roughly estimate a frequency around E (330 Hz) or anyway in the range from E to F, that I associate with a warm and dark sounding violin (another gamble...).

From my modification I would expect a rise in frequency up to F# (370Hz) that I try to associate with more brilliancy.

This is not a tuning attempt but only an empirical way to associate frequency to stiffness, that of course may work only on a personal work known and metabolized.

With this I do not mean that the "scientific numbers" do not work, I also like to know them because having as much data as possible helps to understand,
but surely it is more likely that ancient makers were referring to things like tap tones frequency and not to abstract numbers, and this can still be valid today.
 

 

Hi Davide, I give this fiddle as an example because it is so well known and archived. It shows how some very basic techniques can aid us in improving our work.

my guess is that the arch was original a bit higher if we account for the sag in the middle. Perhaps 1 mm maybe more but probably solidly in the 15+mm range. This would certainly help hold up the arch but I think the principal culprit here is the thin graduations. Strength comes from a combination of arching and graduations. With .35g/cc anything less than 2.5mm is asking for trouble.

changes in the arch shape do not change the speed of sound in the wood or the modulus of elasticity. Those qualities are inherent to the material and not shape.

i believe they call the strength of the shape the 'section modulus' (?). (Can I get some help?:))

The language of science helps give a name to empiricism.

Share this post


Link to post
Share on other sites

 

13 minutes ago, curious1 said:

Hi Davide, I give this fiddle as an example because it is so well known and archived. It shows how some very basic techniques can aid us in improving our work.

my guess is that the arch was original a bit higher if we account for the sag in the middle. Perhaps 1 mm maybe more but probably solidly in the 15+mm range. This would certainly help hold up the arch but I think the principal culprit here is the thin graduations. Strength comes from a combination of arching and graduations. With .35g/cc anything less than 2.5mm is asking for trouble.

changes in the arch shape do not change the speed of sound in the wood or the modulus of elasticity. Those qualities are inherent to the material and not shape.

i believe they call the strength of the shape the 'section modulus' (?). (Can I get some help?:))

The language of science helps give a name to empiricism.

I agree with the density / thickness ratio you suggest and the fact that in the Huberman this may be the main cause of deformation.

But with regard to speed of sound I am perplexed that is not affected by the arching, because it is a date inherent to the material as a log with straight and continuous fiber.

These fibers is no longer continuous in the arching, so I expect a lowering in sound velocity propagation that may spoil the calculation of real Modulus in the finished plate, also considering that the number is squared in the formula.

But my background in science is very scarce, so I'm almost shooting in the dark....:rolleyes:

 

Share this post


Link to post
Share on other sites
27 minutes ago, curious1 said:

Hi Davide, I give this fiddle as an example because it is so well known and archived. It shows how some very basic techniques can aid us in improving our work.

my guess is that the arch was original a bit higher if we account for the sag in the middle. Perhaps 1 mm maybe more but probably solidly in the 15+mm range. This would certainly help hold up the arch but I think the principal culprit here is the thin graduations. Strength comes from a combination of arching and graduations. With .35g/cc anything less than 2.5mm is asking for trouble.

changes in the arch shape do not change the speed of sound in the wood or the modulus of elasticity. Those qualities are inherent to the material and not shape.

i believe they call the strength of the shape the 'section modulus' (?). (Can I get some help?:))

The language of science helps give a name to empiricism.

The side view cross section view of the Huberman shows that the sag in the middle portion of the longitudinal arch makes that portion nearly horizontal.  This may or may not be detrimental to the violin's sound character.

The attached paper gives some evidence that a long horizontal arch produces a "bridge hill" increase in sound output at about 2.5kHz which is often found with good violins and absent in poorer ones.

Perhaps this creep deformation was actually good and maybe it's a mistake to prevent it by making the center portion of the top plates thick.

vibsys_2016-ch19.pdf

Share this post


Link to post
Share on other sites
1 hour ago, curious1 said:

believe they call the strength of the shape the 'section modulus' (?). (Can I get some help?:))

Young's modulus is what is related to the material.  Section modulus is something related to beam cross-sections, for calculating bending stiffness (by combining it with the Young's modulus of the material).

42 minutes ago, Davide Sora said:

But with regard to speed of sound I am perplexed that is not affected by the arching, because it is a date inherent to the material as a log with straight and continuous fiber.

These fibers is no longer continuous in the arching, so I expect a lowering in sound velocity propagation that may spoil the calculation of real Modulus in the finished plate, also considering that the number is squared in the formula.

There's the modulus of the material... highly direction dependent, but a property of the material alone.  When it is cut into funny shapes, the shape and anisotropic material properties will definitely cause wide swings in the "effective modulus" of the remaining material.   Looking at the cross arch, if you have a 45-degree slope in the arch, it would locally be something like 10% of the modulus of the flat part... assuming normal quartered spruce.

Share this post


Link to post
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.