Official Luthiers Forum!

So what effect does that have on the bridge saddle? Does it rock the bridge back and forth forcefully, does it tension and release the bridge? How does it impart energy to the guitar top?

_________________
Old growth, shmold growth!

This is the simple explanation, but it'll serve the purpose: The force in the direction of the string rocks the saddle, driving a long dipole type mode of vibration. The force transverse to the string drives the cross dipole when a component of that force is parallel to the long axis of the bridge and drives the monopole (the most efficient sound radiator by far) when a component of the transverse force is perpendicular to the plane of the top. There is always a perpendicular component to the transverse force, usually from initiation of the pluck, but the plane of vibration rotates regardless, so you always get some, which couples very well to the monopole mode. As the transverse force is always larger than the tension change force and the monopole is most easily driven and is the most efficient sound radiator, most of the sound from a guitar is radiated as a result of the transverse force driving the monopole. Don't let anyone convince you otherwise!

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

What's the difference in top compliance to a transverse force v.s. a longitudinal force? Violins are designed to very efficiently turn transverse string forces into a pumping action perpendicular to the top. Guitars must be much more responsive to longitudinal string forces than transverse forces. Surely the difference is greater than the 4/1 string force proportion measured on steel string guitars, maybe greater than the 40/1 ratio of classicals?

So is this still based on tension and release?

_________________
Old growth, shmold growth!

Okay, we're using different terms. I'm with you on the efficiency of the transverse waves perpendicular to the top. And I'll buy that being the dominant force. It was my understanding that the initial transverse component parallel to the top quickly rotated, and became perpendicular. Maybe that's bogus, but it seemed to make sense when I read it: the force moving in the direction of compliance. I think something like that happens on a sail boat tacking to windward.

The question about compliance in my last post was about the difference in top sensitivity to transverse forces parallel to the top v.s. longitudinal forces.

Here's a bonus question: Is ignoring the higher vibrational modes of the guitar justified, or merely convenient? With the sensitivity of human hearing centered around 1,000 Hz, aren't the higher modes (the modes that nobody tunes) crucial in determining a guitar's timbre?

Only the tension change force is based on "tension and release". The dominant transverse force isn't. It only depends on the change in angle of the string at the termination. The whole string tension is applied, which comprises the static string tension plus the relatively tiny bit of tension change due to plucking. If you had a constant tension device (e.g. a pneumatic cylinder) applying the tension so that as the string vibrates you got no tension change, you'd still get close to the same magnitude of transverse force and still a good sound out of the guitar, but there wouldn't be any tension change force at all.

Merely convenient. They're very difficult to tune, as you say, but add significantly to timbre. I think it's something like 90% of the sound energy radiated by a guitar is below 1000Hz (don't quote me on that one, I've not checked it), but still the rest is significant. Try listening to an otherwise "dry" recording of a guitar and set up a parametric shelving filter above 1000Hz and listen to what you get. Design Section 3.2.2 goes into the detail of all of this, relating frequency ranges to types of sound (ringy, nasal etc.) and to vibrational modes and eventually to the woodwork of the guitar (for the low order modes). The higher frequencies you have to control using things like bridge mass and the damping inherent in the components. At least, that's the only way I've found to do it.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

So I've been watching these loaded string simulations, so I can see how the wave moves back and forth across the string, and also how the string vibrates perpendicular to the top at the bridge as the wave strikes the bridge and rebounds. Those modes certainly look more active in activating the saddle. It seems to me that the minimally adequate construction principle would still apply in sensitizing the top to respond to those vibrations.

Not trying to be contentious.

_________________
Old growth, shmold growth!

OK, I'll be contentious then!

How are you going to determine what "minimally adequate construction" is? Over what period of time will your structure be minimally adequate? What, if any, principles of structural engineering and materials science are you going to invoke to figure this out?

Most current X-braced guitars can hardly be called sensitive or responsive (I use monopole mobility as a measure of this - what do you use?), yet many of them, over the next 30-50 years, some much less, will have their tops cave in, as I'm sure you've seen. So how are you going to build a truly sensitive top that isn't prone to collapse???

Just thought I'd ask!

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

I had to get my brain wrapped around the forces.... I ended up doing some back of the envelope static force balances.... 100% of the string tension force is pulling on the guitar long-ways in the middle of the guitar - between the nut and saddle.... At the bridge - because of the break angle.. it resolves to something like 25% of the string tension force pushing straight down into the saddle - balanced out by the same force pulling up at the string ball ends... and very little in the up/down direction because the strings are pulling long-ways.... Maybe something in the range of 50 in-lbs trying to yank the bridge off the top because of torque depending on how the bridge is made....

My solution was to try to think of the top of an instrument like an empty beer can with some drunken party fellow standing on it.... The paper thin aluminum cylinder can support his weight just fine - so long as the side of the can doesn't crease... so my bracing is there to help the top retain it's strength between the bridge and the nut.... The bracing isn't there to provide long ways load support against string tension - it's not really in the right plane to do this..... Really - I want to keep the top from creasing.... but that's my idea, anyway...

A side effect of this is that the entire structure is stiffest in the tension-change signal direction.... In effect - you have a wooden block that is 2' thick in between the nut and the saddle.... and another wooden block that is 10" thick behind the saddle....

If you wanted to make a guitar so that the top was most supple to the tension change signal - you would end up with an accordion.... Mushy in the push/pull direction.... but then, how would you tension the strings?

Thanks

Ah, now ain't that the thousand dollar question? (don't think we ever get paid enough to have a million dollar question )

I think this pretty much sums it up for me. http://www.youtube.com/watch?v=u0-oinyjsk0
If the answer is yes, then I shave a little more off the X brace.

And on the topic of the original post, I do tap, but I don't tune Well, not to specific frequencies. But I do aim to get the top and back "echo" to ring for as long as possible when the box is tapped, before gluing the bridge on. If nothing else, it's a way to get a somewhat consistent interval between the top and the back.

Tapping is one of those things where it's hard to tell if it does any good or not. But I figure between that, flexing the plates, and just looking at the bracing, I'll eventually build up some idea of what will sound good as a completed instrument. No sense NOT tapping on the thing... the more sensory input you can get, the better. And at least to me, it does seem to convey useful information, especially in locating specific "tight" spots where I need to shave the bracing down a little more. And tapping on the raw woods, you can get a pretty good reading on the damping properties of each piece, so that's definitely useful.

After reading Siminoff's book (I assumed he knew what he was talking about), I too thought most of the sound was produced by the bridge rocking forward and aft. I recall discussing this with Al Carruth and his thoughts were identical to Trevors.......the bridge bouncing up and down in monopole mode was the primary sound producer influencing frequencies fairly high up the spectrum.

On a different topic, I'm guessing one could build a guitar top that was just on the verge of collapsing that didn't sound good........so surely there is more to it than that. Conversely, you can build a nice stiff top (stiff in the right places) that sounds very good.

_________________
Formerly known as Adaboy.......

Well, experience, that's simple. Reduce the structure to failing, and then increase it. Take careful measurements. Trial and error will be your guide. All structures age and collapse over time, no matter what. Ultimately, the final tone must please the ears of the luthier and the the clientel. How long should a guitar last? They can always be repaired. If the top fails after 30 years, it can be replaced, most Stradivarious violins have been repaired multiple times, and even had components replaced, including their soundboards. It seems to me that a guitar that sounds phenomenal for 30 years is better than a guitar that sounds extremely mediocre for 100 years.
Even the best built guitars require maintenance. However, most acoustics need a neck reset before they need a top replaced, even lightly built ones. All the most responsive guitars I've seen have had top distortions after a few years, but still sounded and played phenomenally, so I think tops can be made as lightly as can hold the string tension.

I'm not criticising what you are doing, I'm just asking questions. How does striking a top with a mallet or vibrating it with a speaker give you useful information about how a top responds to the action of a tensioned vibrating string coupled to the top of a guitar wrapped over a thin piece of bone and plucked?

The point of the OP was to ask if top tuning is necessary, and the answer seems to me to be no. That may change in the future. Top luthiers build phenomenal sounding guitars with and without component tuning. However, if it helps you obtain a tone that you find pleasing, to you and your clientelle, then it should be done. I'm still exploring all possibilities and have gained some useful information from this thread, so thanks for the responses.

_________________
Old growth, shmold growth!

If you are truly aiming for a top to be close to failure then those careful measurements should include modulus of elasticity measurements of your top and bracestock, rather than just dimensional measurements. 10 or 14 gPa for your spruce makes quite a difference.
Now that you mention neck resets, they should not be considered as a separate issue to top failure. Neck resets are required due to top collapse around the soundhole and pullup around the bridge.

It sounds simple. However, upon examining loaded string models, I agree that the transverse mode perpendicular to the string imparts the most energy, but it still does it by tensioning and releasing the saddle, although it appears that the tension oscillates from resting tension to vibrating tension. However, it appears the saddle doesn't really bounce up and down, it is still being leveraged back and forth in relation to the top. What I'm looking at is the action of the string as it oscillates up and down at the saddle at the end the wave cycle, It appears to release tension on the up wave and pull forward on the down wave. Consider that the saddle is under constant tension and torsioning throughout the cycle, and you will see that the saddle itself does no "bouncing up and down", although the overall induced top movement includes "bouncy" modes, caused by sympathetic string vibration.

Here's the one I'm looking at.
http://www.falstad.com/loadedstring/directions.html



If what you say is true, tops should be replaced every time a neck reset is performed. Most of the time when a neck is reset, the structural integrity of the top is still accepteable. I have nothing against modulous of elasticity measurements.
This thread has been informative, but I think it's descending to the realm of "pissing contest", so I will bow out.

_________________
Old growth, shmold growth!

I’ll give it one more go explaining how string forces work, seeing as you’ve found Paul Falstad’s excellent applets (then maybe Al Curruth might explain things better!)

Open the applet and pluck the (virtual) weighted string in the middle and then look at the string ends for the first few cycles. Simplifying things, the angle the string makes with the termination has just two values, plus alpha and minus alpha from the “at rest” position. The forces on the termination can be resolved in two directions; parallel to the string axis at rest and perpendicular (up the page) to the string at rest. The force perpendicular to the string’s axis (the transverse force) consequently has just two values, the tension in the string multiplied by sine(+alpha) and the tension in the string multiplied by sine(- alpha). The tension in a guitar string is large, the angle change is small, but the net result is an oscillating (switching) component of force (a string of pulses) that drives the perpendicular motion of the soundboard at the string oscillation frequency.

Now we need to look at the axial component of force, which is the string tension multiplied by cosine(+alpha) and the string tension multiplied by cosine(-alpha). Now for small angles cosine(+alpha)=cosine(-alpha) so there is no oscillating force. However, (BIG however), because the string has to stretch to take up its deflected shape, it also increases in tension (Hooke’s law) and it does this TWICE every period of oscillation, producing pulses at twice the string oscillation frequency that tug on the terminations. If you work the numbers, this tension change force is only a fraction of the transverse force.

Don’t believe this? Well, the dead giveaway is that when you pluck a string on a guitar the sound you hear is largely at the string fundamental frequency, not twice that frequency, which infers that it is the string transverse force that drives the top to produce the majority of sound.

I can’t explain it any better than that, so that’s my last go!


Striking with a mallet imparts a single impulse to a guitar top, (not too unlike a string imparting a series of pulses as described above). An impulse contains many frequencies and is a standard way in engineering of eliciting response functions from structures. You record the guitar’s response and you can see how it responds to many different frequencies. The high peaks are frequencies where the energy from the string (or mallet) easily transfers to the top. If the peaks are too high, the energy drains from the strings too quickly and you get a clunky note (loud, but no sustain) at that frequency peak, the guitar wolf note. If the guitar is not very sensitive, none of the peaks are very high, so you don’t get a problem. If you make a very sensitive guitar, some peaks are very high, so you have to manage those heights and you have to place then in the right place so as not to get wolf notes, frequency shifting and all the other problems that go with building sensitive guitars. You can build guitars that aren’t very sensitive and you don’t run into these problems and that is the way the majority of guitars are built. Guitars can be a lot better than that, but you have to manage all these other issues. Some makers claim their guitars are “responsive”, but compared to what? Rarely, if ever, is a measure provided for comparison.


No, the OP’s point was “is tap tuning necessary”. I’ve not found a way of tap tuning free soundboards that is in any way useful. And the thought of tuning specific braces on free soundboards to specific notes makes my teeth curl. So the answer for me is “No”. However, MODAL tuning of the finished instrument is absolutely necessary if you want to make responsive guitars that aren’t plagued with wolf note, frequency shifts (playing out of tune) etc. etc. which are all problems that arrive when you make responsive guitars and that is why it is so hard to make really great guitars.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Thanx for being nice.

"So I was hoping to get some thoughts on the subject, is this all it's cracked up to be, I know there are some schools of thought that its all "voodoo" and others who swear by it.."

From the original post, "is this all it's cracked up to be",sounds like an inquiry to the necessity, to me. I may be wrong. There are luthiers who produce phenomenal sounding guitars without tap tuning, there are others who tap tune and believe they are doing the necessary final steps. Quite frankly, I've heard some horrendous examples of guitars made by masters of tuning, and I've heard some great examples as well. Same thing with non-tuned guitars. It just seems to me that if tuning (modal or component or whatever) is the step that made the difference, that would not be true. It also highlights to me how subjective the whole thing may be. The guy who paid $15,000 for his guitar that I thought was fairly mediocre believed it was the most phenomenal guitar he had ever heard.

As far as the bridge action goes, I don't think we are too far apart, and I believe your calculations of string energy, and the directions of the waves. That's not in dispute. I think what we are not together on is the action of the saddle. I don't believe the saddle bounces up and down, the way it would on a violin. If you look at the applet, the wave travels to the saddle parallell to the top, swings up strongly, then small energy ripples occur, the string swings down strongly, small energy ripples occur, then the energy wave travels back the opposite direction. It does this many times per second. So what it looks like is that there is a very strong set of lower frequency tugs, mixed in with higher frequency weaker tugs. Add them up over the course of one second and you get the fundamental frequency (the large tugs) followed by the higher frequency harmonics (the smaller weaker tugs of much higher incidence). Add them together you get the full tone of the string. The larger tugs impart far more energy and are the loudest frequncies, hence the fundamental being the primary tone, as you say. The direction of the energy pulses are perpendicular to the guitar top, so yes, the transverse frequencies are perpendicular to the top and impart the most energy to the top. The visual depiction in the applet correlates well with what you say in these sentences:
"Now we need to look at the axial component of force, which is the string tension multiplied by cosine(+alpha) and the string tension multiplied by cosine(-alpha). Now for small angles cosine(+alpha)=cosine(-alpha) so there is no oscillating force. However, (BIG however), because the string has to stretch to take up its deflected shape, it also increases in tension (Hooke’s law) and it does this TWICE every period of oscillation, producing pulses at twice the string oscillation frequency that tug on the terminations. If you work the numbers, this tension change force is only a fraction of the transverse force."
It makes sense, we can see the string stretching to take up it's deflected shape, and visualize how the tension increases very sharply in those moments compared to the movements in between those cycles.

However, the actual saddle itself is under tension and being pulled forward, that doesn't change, so any increase in tension pulls it further forward, and when the string relaxes to the original tension, it relaxes back as well. The net movement of the saddle is still to be leveraged back and forth on a pivot point as the string vibrates to dissipate the energy imparted to the string by the motion of plucking. So the principle of minimally adequate construction applies, if you can make the guitar strong enough to hold the static tension of the strings and no more, then the saddle will respond to the smallest tugs of the string on the saddle. If tuning helps you achieve that point then it's a good thing. Usually when people tune an instrument, they start overbuilt and gradually reduce the structure until it is "right". At that point, it is minimally adequate, to suit the ears and sensibilities of the luthier, and his clientelle. Others don't need to tune to achieve the same result. The final determinant is to please the ears of the luthier, and his clientel.

_________________
Old growth, shmold growth!

OK, well I tried and obviously failed to convince you. All I can recommend is that you read the physics/acoustics books. A good place to start is "Physics of Musical Instruments" by Fletcher and Rossing (2nd edition p241) or my book, which is more oriented towards guitars. Or talk to Alan Curruth who may be able to tell it in an easier-to-understand way.

How a string drives a guitar top is not opinion; it is researched, peer reviewed and published fact. You don't have to come up with "new" models or rely on "kitchen sink science" from people with no training in the discipline. All you need to do is look it up! However, you're far from being alone in not appreciating the subtleties of how a string drives a guitar top and some of these people are quite vocal in spreading misinformation. One thing is for certain though: there's little chance of improving the performance of a guitar if how the top is driven by the strings is not understood; which might explain a few things!

Your view on how a violin bridge works may require some revision also!

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

I am looking it up. I'm not relying on kitchen sink science from someone with no training in the field. I have eyes as well, and if you simply look at at how a string physically vibrates, the string is always being pulled to the side, leveraged across a pivot point. Yes, the energies are moving the string perpendicular and parallel to the guitar top, but if you look at where the string is actually anchored, the string never leaves the saddle, it is always in contact, and as the wave approaches the saddle and the energies are directing the string upward, the string is STILL being pulled to the side. There's no way the saddle itself can "bounce up and down", as you are saying. In order for that to occur, the string would have to pull directly straight up, and it never does that, the string always remains at and angle to the saddle, so the tensions are always pulling the saddle to the side, even as the forces are directing the string upward. Look at the string model, anybody can see that. Now, the wood may be vibrating sympathetically, and resonating in an up and down motion, but the string action directly on the saddle is rocking back and forth. All you have to do is get off your paper and look at how the mechanism is physically moving. Maybe you're too smart and educated to do that, and I'm just a stupid idiot. The fact is, I'm listening to what you are saying, and changing my opinions as I go, despite your arrogant and demeaning attitude, because I think you might know something, and my understanding of string action might be enhanced, but I'm not convinced that EVERYTHING there is to know about guitars and string science is understood.

By the way, science is not exact, it relies on perception and interpretation, and is always changing as new information comes to light.
Also, just because a group of people agree on something and believe it, and peer review each other, doesn't mean the final word is said.

Explain to me how the string is bouncing the saddle up and down, when the tension is pulling the string sideways at all times, and therefore pulling the saddle sideways at all times.

I've been nice up to this point, despite your increasingly negative tone! You're not explaining it very well, maybe you're not as smart as you think. Or maybe the problem you have is that I am not simply accepting what you say as fact (which it may be) because of something I think I'm observing that you can't explain.

_________________
Old growth, shmold growth!

I’m sorry if my “attitude” offended you. Certainly unintentional, but maybe some frustration coming through.

OK, I’ll give this one more go, despite some of your comments, and then I’m outa here!

That is wrong! All you need is a component of the string force directed “straight up from the soundboard”.

Looking at the string model is giving you a view of the motion of the string, but not a view of the forces the string exerts on the terminations (one of which is on the soundboard).

A guitar top will only move if a force is exerted upon it. (Newton’s 2nd law). An alternating force is exerted by the strings that are attached to it as they vibrate. As they vibrate, the angle the strings make with the top at the saddle varies and so there is a component of force perpendicular to the soundboard as well as a component parallel to it. The perpendicular component of the force drives the soundboard up and down. If you run the numbers you’ll find it’s significantly larger than the alternating “tugging” force.

I Googled “resolving forces” and came up with these:

http://www.physicsclassroom.com/class/vectors/u3l3b.cfm

and

http://www.examsolutions.co.uk/maths-re ... rial-1.php

Have a look at them, then check back over my earlier posts.

I hope that helps.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

Okay, that makes more sense. The whole soundboard is moving at once, including the bridge and saddle, kinda like the the string is flinging it around. There IS a tugging force, but it is overpowered by the directional up and down force. That's all you had to say. No need for frustration, just give a better explanation.
You know, I might consider buying your book, but I'm not so sure that I would understand it without asking it the exact right questions to elicit the informational components I need to understand the concept.
Is there a layman's version?

_________________
Old growth, shmold growth!

The other thing to consider is that the applet you have been looking at is a 2 dimensional representation of the 3 dimensional motion of the string.
and is based on fixed end points(rather than having a saddle which is able to move vertically)
It's a great tool for understanding string motion in response to different pluck locations
You might perceive the action of plucking the string as parallel to the soundboard but there is a perpendicular component too and even if you were able to initiate the motion in a pure parallel direction, the string motion quicky resolves to a 3 dimensional action.

http://youtu.be/6sgI7S_G-XI

Slow motion vibrating strings.

_________________
Old growth, shmold growth!

No worries. Spread the word!

There's a couple of chapters in the Design volume which are very mathematical. I wrote it like that because I'd read too much unsubstantiated opinion and didn't want to add to it. It was basically an invitation for people to challenge what I was saying, but to do so would mean them supplying an alternative with substantial and quantifiable argument behind it. If someone does, well and good! Those sections are summarised in layman's terms so you don't need to be a maths major to understand and use the concepts. My co-author is not mathematically inclined, so it had to pass his "understandability" test.

Here's some more slow waves:

http://www.acoustics.salford.ac.uk/fesc ... rvideo.htm

Lot's of interesting stuff on that site.

_________________
Trevor Gore, Luthier. Australian hand made acoustic guitars, classical guitars; custom guitar design and build; guitar design instruction.

http://www.goreguitars.com.au

So I've modified my perception such that if anyone asks my opinion on how the strings interract with top I will tell them that the directional forces of the string move the top of the guitar by pulling the bridge/saddle assembly in whatever the direction the directional forces of the vibration move it, resolving to an up down motion of the entire top as the dominant driver of sound.

Interestingly enough, I still think that if the structure is just strong enough to hold the static tension of the string and no more, it will be responsive to the slightest tug of the string on the bridge/saddle assembly. To me, the strength to weight ratio of the top materials has become even more important, especially for nylon strings. Not so sure about the tapping thing.

_________________
Old growth, shmold growth!

Yes, the bridge bounces up and down when the top is vibrating in the monopole mode.......the virating mode that is the largest sound producer.

_________________
Formerly known as Adaboy.......