Recurve and sound [Archive] - Mandolin Cafe Forum

sunburst

Oct-13-2012, 8:45am

I wish I had your confidence that I understood how top graduations relate to sound. I've built these things for over 20 years and a lot of what I thought I knew about how they work has turned out to be wrong or at least different in actuality. I've seen and heard powerful, good sounding mandolins with anything from and exaggerated re-curve to no re-curve. Stromberg archtop guitars were known for their power, yet many of them had no re-curve.
Arch height, arch shape, and graduations all play a part in how a top works and how the resulting instrument sounds, but they are all so interrelated, and the rest of the instrument is so interrelated, that I hesitate to attribute any specific part of the sound to the shape of the top re-curve alone.

Dave Cowles

Oct-13-2012, 9:41am

Actually, confidence doesn't come close to "fairly aware." :)) I should maybe say somewhat aware.That means I have some vague sense that there must be a connection...:redface: Your knowledge far exceeds my amateurish observations, John. I agree that the extent/degree of interrelated build issues in an archtop mandolin must be exponentially greater than say building a flattop guitar, and was only curious as to how much of a role the recurve plays. I'm prolly making some assumptions based on the total lack of a recurve in the cheap, pressed-top mandos that sound as cheap as their price tag. Thanks for your reply as I greatly respect your knowledge and experience.

Dale Ludewig

Oct-13-2012, 9:48am

Don't overlook the fact that the normal recurve you see on the outside, the dip down near the edge, can be accomplished by a thinning of the top/back in that section on the inside of the instrument. It will have, as far as I know, the same effect but just appears different from the outside.

Dave Cowles

Oct-13-2012, 10:08am

See how knowledgeably confident I am? I never even considered this possibility, Dale, I simply believed it was a top-only feature. Thanks!

Stephen Perry

Oct-13-2012, 10:12am

1. Arch shape is effectively that of the surface halfway between inside and outside, as a somewhat useful simplification.
2. The relationship of the differing arching of front and back plates is important in violins, and I suspect in mandolins

See, e.g., http://www.violin.uk.com/research/Controlling%20arch%20shape.pdf which I highly recommend.

3. Recurve can function to change the resistance of the box to deformation from force vectors along the plate.

Pretty complex stuff.

I bring this up because the one real "Loar" mandolin I had a chance to examine carefully did not have arching like most (maybe all) of the mandolins I've examined from other makers, showing some very interesting violin influences that may relate to modification of the "EAR" (see the link above).

I'm gradually whittling away at my ignorance, but I'm still using a hatchet on a log rather than a pocket knife on a twig.

Pete Jenner

Oct-13-2012, 10:37am

I was about to start a very similar thread.

I've been compiling a list of features commonly found on mandolins which, in my opinion, owe more to Gibson created convention than to any sound acoustic principle. Included in the list so far are - scrolls adjustable truss rods, adjustable bridges, 12th - 15th fret body markers, the virzi tone reducer, ( I stole that description), string gathering tail-pieces, inlays, points and fancy bindings. The re-curve is also currently on the list.

After two years researching mandolin building, I am rapidly concluding that re-curves are just more eye candy.

I read recently in 'The Acoustical Foundations of Music' (J.Backus, 1969), in a section called 'Miscellaneous Factors in Violin Tone', in the violin chapter, that "...it was found that a cheap violin could have it's tone improved and power increased considerably by cutting a narrow groove around the outside edge of the top plate and filled with thin strips of wood that are glued in..." Backus goes on to say that as the violin is played, the glue cracks and "...the purfling becomes sufficiently loosened (but not so loose as to fall out) and the tone of the violin is better". As there is no cut groove in a re-curve, this doesn't apply to mandolins.

I can't find any reference to purfling or re-curves in 'The Physics of Musical Instruments' (Fletcher and Rossing, 1998).
Trawling the interwebz similarly turns up no supporting documentation (apart from anecdotes) for the argument that the re-curve influences tone or volume. Perhaps someone here can provide a link to some empirical data on the subject.

Now don't get me wrong ...;) I think the re-curve is far and away the most sexy part of a mandolin and one of the most satisfying parts to carve. It can turn a dull looking musical box into a very desirable and curvaceous creature indeed. My #2 build will certainly have a re-curve - if it ends up sounding dreadful, it will look nice on my bedroom wall.

sunburst

Oct-13-2012, 10:55am

"...it was found that a cheap violin could have it's tone improved and power increased considerably by cutting a narrow groove around the outside edge of the top plate and filled with thin strips of wood that are glued in..." Backus goes on to say that as the violin is played, the glue cracks and "...the purfling becomes sufficiently loosened (but not so loose as to fall out) and the tone of the violin is better"...

...I think the re-curve is far and away the most sexy part of a mandolin and one of the most satisfying parts to carve. It can turn a dull looking musical box into a very desirable and curvaceous creature indeed...

...and speaking of anecdotes...
Is there any evidence given to support the glue cracking, the purfling loosening and the tone getting "better" (whatever that is)? I think that is one of those things that gets written and repeated without much to back it up, and I'm not surprised that you weren't able to turn up any more writing to support it. I have heard that the purfling grove in a violin makes the edge of the plate more of a "hinged" edge, and the effects of a "hinged" edge as opposed to a tightly held edge have been demonstrated, but I don't know much about all that so I think it best to stay out of that discussion.

I think you've discovered the true purpose of the re-curve in mandolins when you say it "is far and away the most sexy part of a mandolin".

Pete Jenner

Oct-13-2012, 11:07am

...and speaking of anecdotes...
Is there any evidence given to support the glue cracking, the purfling loosening and the tone getting "better" (whatever that is)? I think that is one of those things that gets written and repeated without much to back it up, ...

I think you've discovered the true purpose of the re-curve in mandolins when you say it "is far and away the most sexy part of a mandolin".

I should have prefaced that by putting it in better context. Backus was using it as an example of a possible explanation of the commonly held opinion that playing a violin can improve it's tone. His opening sentence is: "It has been asserted (but not proven) that playing a new violin can improve it's tone".

Even the word 're-curve' is sexy. :)

sunburst

Oct-13-2012, 11:18am

...Backus was using it as an example of a possible explanation of the commonly held opinion that playing a violin can improve it's tone. His opening sentence is: "It has been asserted (but not proven) that playing a new violin can improve it's tone".

Ah yes, speculation. So often written down to be taken as statement of fact even it not intended so. It's one of the reasons it can be so hard to get to the bottom of what is and is not known about how instruments work.

Pete Jenner

Oct-13-2012, 11:23am

Yeah - that's why I got the books.

Oh ...and Dave Cohen told me to. :)

Marty Jacobson

Oct-13-2012, 1:22pm

Good discussion.... one thing that I haven't seen mentioned yet is that the origin of the Gibson-style recurve has a lot to do with the build sequence.
By planning ahead and leaving the plate thick there, there are a number of practical advantages. Dings near the edge of the plate, impressions from spool clamps, tearout from routing binding channels can all be sanded out with a little more breathing room.
Also, if you read Benedetto's process or Michael Lewis's notes here about how he does it, it is just easier and more useful for them to do final recurve graduation from the outside after plates and binding are applied.

I personally do not do final graduation after glue-up. Mine is done beforehand in fixtures which secure the edge of the plate (so I do not do "free plate" graduation). Of course it changes a hair after glue-up, with final sanding. Pretty negligible, though. And I do like to be able to take the final graduated plate out of the fixture and flex it by hand and tap it while holding it, just to confirm that it feels right. Not scientific, I know... just don't tell Trevor Gore and I'll be OK.

EDIT: Campanella "purfles" his mandolins, right? From his web site: http://www.campanellastrings.com/mandolin#
92712

Pete Jenner

Oct-13-2012, 1:49pm

EDIT: Campanella "purfles" his mandolins, right?

Ah yes I came across either him or another bloke doing the same thing last year. Even if Backus' speculation on violin purfling is correct, Campanella's purfling would be too close to the edge to take advantage of it (assuming he is using conventional linings).

The edges look a bit like mine did when I bound the top and back without routing a channel. Never again - too hard to french polish.

sunburst

Oct-13-2012, 2:52pm

Violin plates have their area of minimal thickness right at the rim and the raised portion to the outside of the rim (like the mandolin pictured above) forms the perceived re-curve. The purfling is over the lining/rim of a violin, not to the inside where it might "hinge" (and weaken) the plate more. Many of Orville Gibson's adaptations of violin design were either naively or deliberately different from violin construction, and many of the Gibson corporation's versions of them are too, and those things are a source of confusion for many trained violin builders who decide to try mandolins.

Pete Jenner

Oct-13-2012, 3:11pm

Someone should have said to him "It'll never fly Orville..." Oops ...sorry wrong Orville.

Stephen Perry

Oct-13-2012, 4:42pm

peter.coombe

Oct-13-2012, 5:45pm

I'm fairly aware of the relationship of top graduation to ultimate sound of a finished mandolin,

Just saw this thread and my first reaction was - you are?????

I have not quite reached 20 years, but am not far off, and my new mandolins never cease to surprise me. Sometimes a new idea on how to build them works far better than you ever dreamed it would, sometimes it just tanks. Incredibly complicated and challenging instruments to make and several lifetimes work to understand. Personally I think the outside shape of the recurve plays very little, if any, any role in sound. However, factor in the thickness of the recurve (i.e. what you do on the inside as well as the outside of the plate) and you might be on to something because thickness is related to stiffness, and that does influence sound.

Dale Ludewig

Oct-13-2012, 6:18pm

Peter, I agree completely. If I understand things correctly, the recurve on the sides of the top and also to the back (based on grain direction stuff) at a right angle to the strings is important but much less important than the recurve area where you "round the curve" from that right angle area either toward the neck or toward the tailpiece. Stiffness is our friend and our enemy. On the outside of the ff-holes, it is there, but normally tonewoods are more flexible there. All builders know this. To allow the top to flex to its max and not collapse, builders are in the interesting place of how much to take away the strength in that area- where the stiffness of the wood is highest. Then you hit the area where the strings meet the tailpiece area, where the stress on the top is at an extreme where the recurve should ease off a bit. At least in my experience. FWIW, that line down from the bridge to the tailpiece should probably be left a little stronger due to the stresses involved.

JeffD

Oct-14-2012, 1:10am

Take a straight piece of wood, like a piece of violin lining. Put end against a fixed object and push. It is stiff.

Now bend that piece of wood into an S. Do the push trick. It will act more like a spring.

Would there not be a difference between an induced arch or recurve, which is more like bending, and a carved arch. The induced arch would have some springiness, because the sound pushes against bend fibers. The carved arch would have no bent fibers, so its akin to a block of wood, just carved into a fancy shape.

Just thinking out loud.

Stephen Perry

Oct-14-2012, 5:19am

Yes, but not completely.

Read the Harris article I linked to.

Pete Jenner

Oct-14-2012, 5:55am

Just saw this thread and my first reaction was - you are?????

However, factor in the thickness of the recurve (i.e. what you do on the inside as well as the outside of the plate) and you might be on to something because thickness is related to stiffness, and that does influence sound.

Peter, are you saying you carve a little thinner on the inside beneath where the recurve would normally be?

peter.coombe

Oct-14-2012, 4:24pm

Peter, are you saying you carve a little thinner on the inside beneath where the recurve would normally be?

Yes.

Willie Poole

Oct-14-2012, 8:55pm

While I am not a builder nor an Accoustical engineer I have always looked at the mandolin top recurve and compared its shape to a bell recurve, I have seen bells with out a recurve that had a "clink" when struck and the ones with a recurve sounded off much louder and more beautiful, might not have anything to do with the recurve, might just be the metal that the bell is made out of...Which also might be what the sound of a mandolin top is all about, what wood is being used...I do know that I play a custom made mandolin and a lot of people that do some building or want to ask about the recurve and say how good it makes the mandolin look...I let them think what they want to as far as the sound goes, it does sound great by the way...

Interesting topic....

Willie

Pete Jenner

Oct-14-2012, 9:38pm

Yes.
Not that I want you to give away all your carving secrets. :)

kjbllc

Oct-15-2012, 6:52am

playing in a new violin definitely improves the sound, I do question the purfling loosening, first its pretty much a tight fit when you put it in, and the hide glue last quite a long time, and on top of that, you have ground, sealer and then varnish. Putting strips of wood on the violin will change the sound, a cheap violin, is coarsely graduated, so adding things could very well help the sound. From what I have learned the graduations are the key and unique to every piece of wood.

Since we can use cnc to copy every great sounding violin to the .001" of an inch , why don't we see thousands of fantastic sounding violins being shipped from china. Just copying exactly all the dimensions does not take into account the complexities of every piece of wood. There are great violins being made today and the past, so there doesn't seem to be any secret "thing" about strads, other than the knowledge of how to make a great violin. just my opinion for what its worth.

Of course there is more than just one factor involved, and not sure how this applies to mandolins. But I am planning on building one and have been curious how directly violin design is related to mandolin. I am just beginning to learn myself after about 8 VSO's and expect that many more will be made before I start to get a feel for the plate tuning aspect.

DougC

Oct-15-2012, 9:53am

It is always one of the first things I look for when deciding whether I'm interested in an instrument as it seems to me that those made with a more pronounced recurve deliver the biggest/best sound to my ears. I'm guessing that the thinning of the topwood along the edges really enhances the ability of the topwood to vibrate. Thanks for your thoughts.

After reading all of these informed comments, I go back to Dave's statement here and say, "No, 'recurve' does not indicate better sound, there is more involved."

Dave Cowles

Oct-15-2012, 10:06am

That certainly appears to be the consensus, Doug. I'm grateful for all the informed responses here, and truly believe that my minuscule knowledge of mandolin construction has been expanded exponentially, although still remaining in the nanobit realms

Stephen Perry

Oct-15-2012, 4:50pm

recurve reduces stiffness compared to no recurve with the same graduation. Graduation and recurve don't necessarily correlate.

Mark Seale

Oct-20-2012, 9:28am

sunburst

Oct-20-2012, 9:33am

On an f-hole mandolin, would recurve actually matter on the top? Or at least in the area bound by the relief of the f-hole? The back is acting like a firing speaker, so it makes sense that the recurve is shaping the ability of the back to "fire." However, on the top, after introducing holes in the plate, it seems that you are effectively changing the area that can be put to vibration when a note is sounded. Thoughts?

Neither the top or the back "fires" like a speaker. The physics of mandolin plate movement has been studied enough, and the findings published for all of us to read and at least try to understand. The f-holes seem to make no difference in the plate modes and the recurve doesn't seem to make any difference either.

Stephen Perry

Oct-20-2012, 11:27am

Plate modes can't be everything.

Pete Jenner

Oct-20-2012, 12:43pm

Plate modes can't be everything.

As far as I understand it, the sound that reaches your ears from a mandolin, is produced as a result the vibrational modes of wood (plate resonances) and air (from the Helmholtz resonator).

For the purposes of this discussion, we can discount the (single) air resonance as it is determined by the volume of the box, the area of the sound holes and to a lesser extent (or not at all because I think mandolin bodies are regarded as neckless Helmholtz resonators), the thickness of the edges of the sound holes.

So in this case, plate modes do appear to be everything.

The question then is - 'Does a recurve affect plate modes?'.
So far the consensus answer to that question appears to be - 'not perceptibly'.

Dave Cohen

Oct-20-2012, 10:02pm

"Plate modes can't be everything."

The normal vibrational modes of an object are the vibrational modes characteristic of that object. The object undergoes those normal modes, and does not vibrate in any other modes of motion. In that sense, the normal modes are all there is. They are not the only way to look at the physical function of an instrument, however.

The normal modes of the component parts are pretty much intact in an instrument. That is, the string modes stay in the string, the top plate modes stay in the top plate, the back plate modes stay in the back plate,..., and so on. They do, however, interact with each other, and are said to "perturb" each other. That is, they exchange energy, which causes frequency shifts, etc. To really understand what an instrument is doing, you have to consider coupled oscillators, i.e., oscillators that are distinct, but interacting with each other, exchanging energy, as if they were connected by springs (in an abstract way, they are). The top and back plate modes are the modes of plates clamped at their edges,... well, almost. The edges, aka "ribs" also move some, and that perturbs the plate motions. The air in the body cavity and in the soundhole area(s) also exchanges energy with the plates, another perturbation. The (neck + ribs) assembly also undergoes its own (more or less) bending motions, and in principle, those bending motions can interact with the body motions if the symmetries are "right" and the frequencies of the bending modes just happen to be close to those of the body modes. So far, I have found just one bending mode that does actually interact with a body mode. Not published yet, so I can't release it here. There may be others at higher frequencies, but I'm not sure yet.

There is also the possibility of nonlinearities. Rossing showed me a Chinese gong that he had studied. Had a little dome in the middle. That thing was very nonlinear. You bonked it, and the frequency (pitch) went up as the amplitude died off. Gave that 'boing' sound that we associate with the gongs in Chinese opera. So, I set about looking for nonlinearities in arched mandolin plates. Sat on the main body resonance and varied the driving force (i.e., the current to the coil of the "stinger"). There was no nonlinearity. Pushed the coil as hard as I dared without frying it, and the peak frequency of the mode just sat there, constant. So, Stephen, that should end your conjecture about the recurve giving an extra push. The relatively broad curvature of the plate, combined with the damping properties and anisotropy of the wood, did away with any possibility of nonlinearities.

http://www.Cohenmando.com

Stephen Perry

Oct-21-2012, 10:27am

Standing wave development from edge reflection.

Much character comes from the 2500 to 5000 Hz domain.

Conformal coatings have a tremendous effect (see the general work on violin grounds)

Plate modes aren't everything. If they were, arching and recurve etc wouldn't be the subject of discussion, would they?

Dave Cohen

Oct-21-2012, 11:33am

"Standing wave development from edge reflection"

Stephen, what exactly do you think the normal modes are?

"Much character comes from the 2500 to 5000 Hz domain."

What exactly do you mean by "much"? In bowed string instruments, the Helmholtz motion from the bow plus higher frequency components, e.g., 'double slip', 'multiple flyback', results in greater amplitudes of the higher string harmonics, in turn pushing the higher frequency motions in the instrument's body and bridge. The so-called "bridge hill" is in that frequency region. But the bowed string instrument bridge is light and flexible. Not so for the bridges of plucked string instruments. They are massive and rigid by comparison. The single string pluck in plucked string instruments results in much lower amplitudes for the higher string harmonics generally. I have said this before, but I guess I will have to say it again. Compared to bowed string instruments, plucked string instruments are low frequency animals. By low frequency, I mean that most of what happens in the body and air modes of those things is contained in the frequency range from 0 - 2 kHz. You can find stuff at higher frequencies, but unlike bowed string instruments, you have to push plucked string instruments really hard to see any high frequency stuff at all. All of my work on mandolins has agreed with that. A good thing for me, because all of the other work on plucked string instruments in the last 50 yrs has supported that as well.

"Conformal coatings have a tremendous effect (see the general work on violin grounds)"

It is tempting to not even dignify that with a response. Most of what the components of film finishes do is damping. Schleske found that the ground and the first few coats of violin varnishes harden the surface a bit, possibly increasing the ratio of Young's modulus to density, generally a good thing. But after the first two or three coats, all film finishes just add mass and especially damping. which means that most of the contribution of a film finish to sound quality is in the form of subtraction, and the effect of the first few coats is reversed and then some. That is, the finish has the effect of damping some of the harsher higher frequency components that we don't want to be there. In any case, it is a tremendous overstatement to say that "conformal coating have a tremendous effect." I have built over 70 mandolin family instruments and guitars. Strung all of 'em up in the white so that I could tweak them if necessary before finishing. Applied finish, and of course played them and tweaked the final setup. Never once heard a "tremendous" difference in the sound of any plucked string instrument from before to after finishing. So, show me your data, complete with measurements.

http://www.Cohenmando.com

sunburst

Oct-21-2012, 11:46am

Never mind, Dave covered it much better than I did while I was typing.

Stephen Perry

Oct-21-2012, 1:22pm

What do standing waves in the boundary layer have to do with plate modes?

I'm not attempting to be an expert or anything, I just can't see plate modes as everything. What about the effect of B zero manipulation? That's pretty clear.

As to manipulation of the relative energy into that nice singer's formant region, that's very easily and quickly done on a mandolin without touching the bridge. For example, the technique demonstrated by Fry for bowed strings is surprisingly effective. Can hear the clarity and volume of that range increase nicely, which players in my shop have reported results in greater projection. Worth trying. I can give a quick overview, if anyone is interested. Relies on the setup being pretty solid.

I don't have any data on conformal coatings on mandolins. Maybe "tremendous" is an overstatement, but there's lots and lots and lots of threads on boxes producing sound that point to the bad effect of bad finishes. The most interesting effects to my ears in bowed strings (which I pluck to work on) are from fast grounds, mineral matrix. There's some very interesting work from Bud Purvine.

My point is that plate modes are not the only thing going on in an instrument. Bridge modes and clarity of those modes must be important - the instruments sound better and cleaner if the components of the bridge are working well independently (a complex example: marimba bars). The B zero effect is quite pronounced - Roscoe Morgan commented on this long ago. Unfortunately the mass required to put B zero where I like it is quite large on a mandolin.

I can't believe for a second based on the personal work I've done and on the extensive consideration of other factors in the literature that plate modes are the only thing that counts - the only concept presented above I'm trying to refute. Unless one lumps all the other aspects into "plate modes" - something that would kind of blur "plate modes."

I don't have time, equipment, or energy to do fundamental research at the moment. I do have time to try wacky ideas that come along. Some of them turn out not to be so wacky. So far with mandolins, I've had success in finding impacts from the general techniques presented by Spears, Fry, Smith (Idaho), Purvine (analogy to electro-acoustic transducers and enclosures).

I'll stand by my opinion that the character of a mandolin doesn't result entirely from plate modes. Adding the caveat unless "plate modes" include ribs, bridge, neck, recurve, arching, finish, bar modes and shape, and other things I haven't thought of. That just plain doesn't make sense to me. If plate modes were it, then we wouldn't have all these discussions of bridges, tailpieces, finishes, etc and their effect on response, volume, and tone.

I'm not trying to pick a fight. This subject really doesn't appear worth major discussion. There's more to an Italian dinner than the pasta, there's more to instrument tone, response, and volume than the plate modes. Either that, or we're spending lots of time chatting about and working on things that make no difference at all!

sunburst

Oct-21-2012, 1:53pm

Stephen, can you show me anywhere in this thread where anyone said plate modes are the only things that count when it comes to instrument sound?

peter.coombe

Oct-21-2012, 5:33pm

This seems to be a misunderstanding of terminology. Normal modes of vibration (or normal vibrational modes) is not the same as plate modes yet you (Stephen) are treating it as if they are the same. Plate modes are a subset of normal modes of vibration. I don't think anyone would claim that plate modes are everything. I certainly don't and I can't see that anyone else in this thread does either.

Stephen Perry

Oct-21-2012, 5:52pm

Implied here:"The f-holes seem to make no difference in the plate modes and the recurve doesn't seem to make any difference either."

Perhaps I misunderstood.

It's a complex subject, especially for me. I just look for things that work really nicely!

sunburst

Oct-21-2012, 6:20pm

To attempt to clarify;
A plate has normal plate modes regardless of whether there is a recurve area or if there are f-holes cut in the plate. Those modes are basically the same, though there is natural variation of the shapes, for any plate, recurved, f-holed, of oval holed. Mode frequencies may change with graduations, thicknesses, arching, holes, any difference in the plate, mode shapes may change with those variations, but "modes is modes" and that is what vibrating plates do. Is that all there is to a mandolin making sound? Of coarse not, and I don't see how that statement could imply that it is.

Dave Cohen

Oct-21-2012, 6:35pm

One of Stephen's problems is that he is taking some stuff from violin research and acting as if it were a sure thing with mandolins or other plucked string instruments. The bridge stuff, f'rinstance. First, a violin bridge weighs a few grams at most, and it is very tall, and also very thin. Mandolin bridges are massive by comparison, ranging from >11 grams to about 17 grams at most for adjustables. One-piece mandolin bridges weigh anywhere from about 5 grams (maple) to maybe 8 grams (ebony). They are all extremely rigid by comparison with violin bridges. I fitted a whole bunch of different types of mandolin bridges to one mandolin and did accelerance spectra for all of 'em, one at a time. There just weren't any features all the way up to 5 kHz. All of the bridges - one-piece, conventional adjustable, sliding wedge - behaved essentially similarly, with no bending motions discernable. With a little thought, you can see why that would be true. Mandolin bridges are no taller than about 7/8", all much thicker and more rigid than violin bridges. Any modes of motion in such structures would not be expected to appear any lower than, what, 20 kHz. I certainly didn't see any bending modes in any of the bridges all the way up to 5 kHz.

The B0 mode Stephen is referring to is the nomenclature used by some for the first beam-like or bar-like bending motion of the whole instrument - neck plus body. I alluded to doing some work on those modes (B-modes) in mandolins in my post #33 above. Since I haven't published that stuff yet, I can't disclose it prior to publication. I will say that I compared bending motions for a single mandolin with two radically different necks, both in mass and in stiffness. Found a lot of bending motions. So far, only one of them interacts with body or "corpus" modes, and it is not the B0. If Stephen thinks he can manipulate that one by shaving a bit of dust off the neck or adding some epoxy or clay, or even lead (!), he is at best deluding himself, or at worst, blowing smoke.

I won't touch the stuff about "manipulation of the relative energy into that nice singer's formant region,...", as it is too much like magic for me to know where to start.

http://www.Cohenmando.com

Dave Cohen

Oct-21-2012, 9:17pm

"What do standing waves in the boundary layer have to do with plate modes?"

Now I see your terminology problem. (1) What you are calling a boundary layer is not a boundary layer at all - not in any graduate or undergraduate course that I took, anyway. It is just part of the plate. (2) The normal modes of motion of plates clamped at their edges are standing waves, and their origin is similar to the origin of standing waves in strings. The difference between strings and plates is that the plates are nominally 2-dimensional, with vibrational motion in the third dimension. Strings are nominally 1-dimensional, with vibrational motion in the other two spatial dimensions. When you talk about reflection of standing waves at plate edges, that is exactly what the normal modes of the body are! You need a basic understanding of plate motion, e.g., from a textbook. I would suggest the Fletcher & Rossing text (Chapter 3), or, you could go back to Morse ("Vibration & Sound - 4th edition (1981) was the last edition), chapter V. Or, you could go all the way back to Rayleigh, "The Theory of Sound", Vol I., published in 1894, reissued by Dover in 1945. Chapters IX and X in that one. Dover reissues are cheap. If you are gonna throw around terminology like "boundary layer", manipulation of the relative energy into that nice singer's formant region", & etc., you might as well get it right, or else not use the teminology at all.

http://www.Cohenmando.com

Pete Jenner

Oct-21-2012, 10:46pm

I won't touch the stuff about "manipulation of the relative energy into that nice singer's formant region,...", as it is too much like magic for me to know where to start.

...and I was about to make a mandolin in effigy and wave a dead chicken over it at midnight.

Mark Seale

Oct-22-2012, 10:08am

Neither the top or the back "fires" like a speaker. The physics of mandolin plate movement has been studied enough, and the findings published for all of us to read and at least try to understand. The f-holes seem to make no difference in the plate modes and the recurve doesn't seem to make any difference either.

I didn't mention or indicate plate modes changing. The QUESTION was whether the recurve on the top was meaningful once the f-holes are introduced into the plate. I'm talking about the area of the top that actually gets set to vibrating when a note is sounded. As to the back firing like a speaker, that may be an oversimplification, the idea being that the back moves more uniformly and benefits more from having recurve than the top.

Pete Jenner

Oct-22-2012, 10:31am

I'm talking about the area of the top that actually gets set to vibrating when a note is sounded.

Mark, why not have a look at this (http://http://www.mandolincafe.com/forum/showthread.php?89383-Building-an-A-Style-Mandolin-F-hole-vs-Oval-Hole/page2) thread. Especially post #36 where Dave Cohen explains the 'global' nature of plate modes.

Another very informative thread.

Cheers,

sunburst

Oct-22-2012, 10:38am

The whole top is set in motion any time it is set an motion. There is no specific "area of the top that actually gets set to vibrating when a note is sounded", it is the whole thing. Neither f-holes or recurve change that. Same with the back; the full complex series of plate modes, involving the whole back, is set in motion any time there is motion regardless of recurve.
Graduated carving does not make top and/or back plates behave like an ideal speaker cone, with a stiff center and a flexible edge, and the recurve area doesn't do that either. Why graduate and why carve a recurve? I've asked myself that question, and I don't really have an answer. Graduated carving allows us to make areas of the top and back stiffer or less stiff, heavier or lighter, and that can affect the frequencies of plate modes (and thereby the rest of the instrument in complex ways). We can improve the match between the top and back we're working with through selective carving and improve coupling. In other words, the top and the back must work together for good sound, and carving them can improve the way they work together, but the speaker cone analogy and comparing the recurve area with the flexible edge of a speaker cone is "barking up the wrong tree", so to speak.

Mark Seale

Oct-22-2012, 11:10am

I agree with everything you're saying John, except I guess my simplified view is that the top and back work together a bit like a sound pump, the top and back (and everything else) do work in concert. In looking at the physical studies done by Cohen, et al, it doesn't look like there is much movement outside of the f-holes on a top plate, thus the inference that perhaps the recurve on the top is either not necessary or not uniformly necessary. Again, I'm not saying the modes are differing or that we're affecting them. The original question was how much the recurve has to do with the quality of the sound of a particular instrument or whether it was an attribute to notice when looking for a quality instrument. What I'm asking is whether the recurve has more use on the back than the top.

Dave Cohen

Oct-22-2012, 11:41am

Mark, the areas of the plate near the edges do move, including the areas outside the ff-holes. On of the things we do to get good images for publication is cut down on the amplitude so that the interference fringes don't crowd together and become a big blurry blob. Each fringe repesents an amplitude of vibration equal to 1/4 of the wavelength of the laser light used. In our case, that is frequency-doubled Neodymium, so the wavelength at the camera is ~531 nm. That means each fringe represents about 133 nm of amplitude. For a good picture, that means that the amplitude near the edge is smaller than 133 nm, so no fringe shows up out there. The amplitude of a vibrating plate clamped at its' edges is naturally smallest near the edge, but not zero by any means. Many times, I've pushed the amplitude and seen the fringes crowd right out to the edge. They go right through the soundhole area as if the soudhole(s) weren't there. There should be some images like that in the Cohen & Rossing papers, iirc. Also, keep in mind that the forces from a plucked string drive the top several times harder than what I do to get good images. So, you can be sure that when you play, the edges of the plates are in motion, even though the amplitude there is smaller than at the center of the plate.

http://www.Cohenmando.com

sunburst

Oct-22-2012, 11:44am

The top and back do work as an air pump, if not a sound pump, and since moving air is what we hear as sound it could be said that they are a sound pump, but only in a couple of the lower frequency modes of motion (things are different for the higher frequency modes).
As to whether recurve has more "use" on the back plate, I don't know. If you find out, please let me know! ;)

Kevin K

Oct-22-2012, 12:57pm

Should the back be really stiff or have some flex to it for the better tone and volume character?

DougC

Oct-22-2012, 2:18pm

I'm simply awed by the talent on these pages. This 'monitor' has got enough for about ten years study.

One issue somewhat related to this topic I've been thinking about is that the rib height, often taken for granted, has a
sweet spot in terms of the distance between plates and their inherent arching (and re-curves...) influences the tone balance across the strings.

Hope that makes sense. It is hard to be real specific off the top of my head.

sunburst

Oct-22-2012, 3:56pm

Should the back be really stiff or have some flex to it for the better tone and volume character?

The back needs to work with the top (and the rest of the instrument). It is more the relationship between the two than the stiffness of the back itself that effects loudness. There is some optimum weight and stiffness for each back depending on the top it is paired with and the sound desired.

As for rim height/body depth, that influences the air modes which influence the top modes, which influence everything else to one degree or another. There would be some optimum body depth for the sound one wants, deeper tends more toward "oval hole sound", thinner tends toward more "projection" or "f-hole sound", for want of better terms.