Tap tuning

[HoGo]

Not long ago there was a discussion about tuning. I posted my overview of chladni patterns I found on tops and backs (without f-holes and tonebars).
The thread is gone, so I'll try to post it again.

http://www.fpv.umb.sk/~minarov/F5_Modes.gif

[HoGo]

I also posted some pics of the top "in action" in other section. Here they are...

http://www.fpv.umb.sk/~minarov/m02.jpg
http://www.fpv.umb.sk/~minarov/m03.jpg

[HoGo]

The general shape of the mode lines is pretty much predefined by the shape and archings of the plate assuming it is evenly graduated. Lumps in the thickness tend to make the patterns look aslo "lumpy" and uneven. I don't know which modes should occur at which frequencies or which should be strong... These are just measurements that I took. I just tried to have the back at the same notes as the top. After I cut f-holes and tonebars the top was about one step under the back which I heard is even better than the exactly same notes.
You can hear the frequenties of almost all modes when holding and tapping the plate at certain points. I used PC with SB16 and 4Watt speakers and CoolEdit software. There is, however, some free software (and even more functional) for this available on the net. Look for sinewave generators and FFT's... or virtual stroboscopes...

[Brookside]

You can tap with anything you can come up with around the house. I use a wooden spoon with a sock wrapped around it. Experiment a bit and you will find the right tapping tool. Siminoff suggests tuning a F-style mandolin to A flat (G sharp) on the bass bar and A sharp (B flat) on the treble bar. He states the backboard is tuned to 3rd octave D. I used a Peterson VS II. I have mixed feelings about the whole tap tuning experience. I'm sure I have much more to learn on the subject.

It's worth mentioning that there are respected builders who do not tap tune at all. I read in an interview that Jim Triggs does not. I'm sure you could find strong arguments for doing it or not. As well you will find many opinions on how it should be done. Good luck.

Any of you professional builders care to chime in? Do you tap tune or not?

[Chris Baird]

The violin builders have identified the most tap relationships that correspond directly with improved tone. There is really little information as to the relationships that a mandolin should possess. I believe Dr. Cohen is working on identifying eigenmodes for mandolins and will publish soon. Even so, the known benificial relationships that can be understood and acted upon scientifically are few and represent a small portion of what is considered good tone. I believe it helps to understand basic acoustical physics and to conduct some scientific tests but ultimately if one hopes to build good mandolins consistently and in a decent time frame the luthier will have to become intuitively sensitive. In the same way that an experienced pro quaterback can take in all the complicated variables of a football game and deliver the ball in just the right spot using nothing but his experience and intuitive subconsious reasoning so can a good luthier detect just the right wood, flex, dampening characteristics etc.. Of course a physicist could figure out just how to throw the ball and calculate all the variables of a football game but certainly not in an effective time frame, likewise probably goes for a physicist building a mandolin; He could do all the tests and overtime build up a database of recorded information from which to work, but it would be very time consuming. I believe a good luthier will be able to make all the tests necessary in a fraction of the time using an intuitive approach and build a great mandolin never having conciously understood half of what he/she was doing, much the same way a good quaterback can figure out complicated trajectories without being able to pass a highschool math test.

[sunburst]

Chris,
Coming from one who is usually a man of few words, that was quite a post. That is the same "philosophy" I've developed concerning plate carving, but I couldn't have said it better.

[HoGo]

I didn't know anything about tuning mandolin plates, but I read everything available about violins when I did it. Some great Czech luthiers said they do preliminary tuning with top without holes or bassbar since the change in frequencies of modes are almost constant (they drop about 20Hz for the lowest modes of violin and raise back almost to original freq when bassbar is glued in) I assumed it would be similar with mandolins. I don't remember exact numbers, but it drops about 40Hz and tone-bars raise it by about 30Hz (the lowest modes). I wouldn't go for exact notes. I went for elasticity and when I felt I'm close to final thickness I just fine-tuned top and back to match each other... When you tap the tone-bars you are actually hear a composite of some modes with one of them being dominant. I assume tapping the tone-bars will make the 3rd and 4th mode dominant, but it REALLY depends on where you hold the plate and where you tap the tone-bar. I don't tune to exact predefined frequencies as with some pieces of wood one would need to remove too much wood to get to exact frequency, however, there is some range that most plates will fall into. The relationship between modes in the plate and between top and back are much more important, IMO. There are some other variables, though, like the weight and stiffness of the rib structure and tuning of air-chamber and also the break angle of the strings, but I feel with standard model (let's say Loar F5) these are quite constant.
What I posted earlier was my own observation that was not influenced by anybody. I tried not to spend my youth experimenting. All it took me was one afternoon. I think it's reasonable time, for it may allow me build better mandos sometimes later.
I think that Dr. Cohen knows more about this than most other people together.
And yes, there are other ways. I believe folks at Gibson don't do any "inside tuning", most of their tuning is done after stringing the white instrument. This works, of course, but you have no idea what happened to the modes of the plates.

I wonder if there's anybody using so-called "Fuhr method" for measuring tap tones of the finished instrument?

[Brookside]

Which G#; which A#? He doesn't say. At least not exactly. I have both versions of his book. The old version says:

"The goal is to achieve either a full tone or a semi-tone difference between the two tone bars. Our optimum choice would be to adjust them to G# and A#."

"I learned that the starting points with this particular instrument were C5 for the treble tone bar, and A4 for the bass tone bar."

He goes on to describe how he alters the tone bar shapes to achieve the results he is satisfied with.

"...I had both bars completed and tuned at the fourth octave Ab and A#."

He finished one full tone apart and not at his "optimum choice." He is tuning the top, with sides attached, f-holes cut and tone bars attached. He strikes the tone bars with a felt covered hammer.

The new version of his book combines all elements of tuning the assembly into one chapter. This version does not walk through the process of tuning a particular instrument but provides more on tuning theory. All he has to say on the subject is worth a read in my opinion, but, it is more of a lay persons overview and may not satisfy the hard-core science junkie.
I personally have no desire to go to the extreme that HoGo goes to. I tend to lean towards Chris and Sunburst's way. (Use the force, Luke!) However, HoGo did produce one of the finest first builds this cafe has seen. It is generally said that one should spend some time on tap tuning and record the results of each instrument. Over time your records may offer some insight. I wouldn't agrue that Dr. Cohen isn't the worlds most premier authority on this subject but then again, I've never played one of his mandolins. http://www.mandolincafe.net/iB_html/non-cgi/emoticons/biggrin.gif

[Chris Baird]

I think that in the beggining of a luthier's development that doing as many scientific tests as possible will help to conceptualize what the mandolin is doing. But, spend just as much time feeling the wood with your own senses. I've done a lot of scientific tests and am starting to drift away from most of them as I feel my body's experience with the wood is surpassing my brain's ability to crunch the numbers I've accrued. The more I study musical acoustics the more complicated things get and the less I feel I know. I will no doubt continue to study and to test but not at the expense of building without my body's furthering knowledge. I have to say that from what I've read of Siminoff's tuning method that it is fairly incomplete and even inaccurate according to some of my own tests. The important relationship for treble response is strings to bridge to top. The important relationship for bass response is top to back to air chamber. Take note of how these things relate and take note of the kind of mandolin tone they produce. I don't think you will get any clear answers to just exactly what to tune to because that information just doesn't conclusively exist.

[sunburst]

It is generally said that one should spend some time on tap tuning and record the results of each instrument. Over time your records may offer some insight.
I would add to this: write down a LOT of stuff early on. You might not know now what will seem important to you later, so If it's all there, you might find the information you're looking for later on. I've been tapping and recording notes for nearly 20 years and I'm still waiting for all that information to show a pattern or correlation to the sound of the instruments.
I find that as I learn more, I want to know what I did on early mandolins only to look back through my notes and find out that I didn't write down anything about the aspect in question. The things I want to know have changed as I've learned. Some of the things I used to think were important I find don't matter much to me now.

I've played one of Dave's mandolins. It sounded good. It had a unique sound that I assume may be characteristic of his work, just enough different from the Loar model and it's variations to be slightly unfamiliar. I think it would take me a while with the instrument, hearing it mostly as I am not a good player, to really understand the sound and it's potential.

[Brookside]

I think what Siminoff was trying to do is provide a "blueprint" for tap tuning. I agree that it is a futile effort. The problem with targeting a specific note is that the amateur finds himself shaving, shaping, altering, sanding etc. in order to please the needle on the strobotuner. That does throw out any element of intuition.

The path to a great tone is to build, build and build again.

[Dave Cohen]

Couple of things.

First, my invited presentation (w/ Tom Rossing) to be given at the ASA meeting, May 27, will not be the first identification of mandolin eigenmodes. The first was our Catgut paper, CASJ of Nov, 2000, and the second was our paper in the January, 2003 issue of Acoustical Science & Technology. So eigenmodes of f-hole mandolins have been known since at least Nov, 2000. Further, as mandolin eigenmodes are quite similar to guitar eigenmodes, we have known what to expect since well before the Cohen & Rossing papers. What I will be talking about at ASA will be a comparison of the eigenmodes in different types of mandolins. We have looked at Neapolitans, a Vega cylinderback, several Gibson oval-hole As and Fs (some w/ Virzis, some without), and a '24 Loar, in addition to data for several of my mandolins published in the previous papers.

Second, anyone considering doing free plate tuning should read the CASJ papers by Schleske and Atwood (two different papers). You can find both of them in the May, 2000 issue of the Catgut Acoustical Society Journal. I should emphasize that the "free plate" modes and the modes of an assembled instrument are completely different. The modes of a "free plate" are what plate tuners are looking at, and they are closely related to the bending modes of a simply supported bar or beam. The modes of an assembled instrument are very different. Not only are they different to start with, but they are coupled to the string modes and cavity air modes, all of which can shift frequencies significantly.

[peter.coombe]

I have said this before and I will say it again. Plate tuning is a tool, but it is just one of a number of tools a Luthier can use. Can it make a difference - yes, but only as one part of a number of other factors that can affect tone. A lot of small improvements add up to a big improvement, and plate tuning can give you one of those small improvements, and can improve consistency. A mandolin is the sum of all it's parts, every one of which affects tone, so you need to think of it as a system. So many people have come to me asking about plate tuning, expecting it to give them the "secret" of better tone. It doesn't. Combine plate tuning with arching, choice of wood, graduations, neck, bracing, bridge, tailpiece, strings, varnish etc etc (it is a long list), then you start to understand what gives good tone. I built some fine sounding mandolins before I got into plate tuning, and have had some back for refrets so have been able to compare them with current instruments. So, build em, learn about the other factors that affect tone, and develop your Luthier skills first before stressing out about plate tuning. It may give you that small improvement that adds to the other small improvements that finally add up to a big improvement, but by itself will probably give you nothing. It takes a big investment in time to get it working so you think you understand (you probably don't) and can use it to predict the sound you will get from a set of tuned plates. It ain't easy, wood is a natural and highly variable material that usually refuses to tune how you want it tuned. That is the challenge, and is probably why many give up and use intuition instead.

[HoGo]

Brookside, I am a hard-core science junkie... math and computer science is my day job. these measurements are jus my intellectual exercise http://www.mandolincafe.net/iB_html/non-cgi/emoticons/smile.gif
Guys, I would like to hear your opinion on this. At a applied math conference I met with a prof. Bodenhoffer from Austria who did a lot of research in the field of data-driven machine learning. Simply said they are able to extract knowledge about a system behavior from large database of input and output measurements. They applied this to a paper production and with their model they were able to control about 200 input variables of the paper mill machine to produce 90% of top quality paper. I talked to him and he said the method could possibly enable #me to find a relation between input variables (properties of wood, thicknesses, frequencies...) and the resulting quality of instrument. The method is able to recognise important input factors from nonimportant and interprete results in the human-frienly way. The only thing you need is a database of enough records (they used 200 records each consisting of about 200 variables for the paper mill).

[peter.coombe]

Well I am also a hard core science junkie (PhD). This is probably why I got into Chladni plate tuning. What you say is fine in theory, but there is one large snag - how do you objectively measure the quality of a musical instrument? This is subjective, and different people have different tastes and different opinions. Double bind tests have not got a good record for evaluation of violins. Thus your measurements of output is going to be unreliable, and thus the correlations will be unreliable. We all do this sort of analysis in our brain, it happens through experience in building lots of instruments, but even then it is difficult to keep the output measurements reasonably reliable because sound memory is short. The same problem occurs when evaluating Hi Fi equipment, something I have done a fair bit of. I try to do the quality measurements by using a reference mandolin that I am very well familiar with. Unfortunately, the reference needs to change at regular intervals because eventually it becomes not useful because everthing I am making then sounds much better than the reference. However, by doing this I can be sure that progress is actually occurring, and it is not wishful thinking. It also helps to get the occasional instrument back for a re-fret, especially if it is a former reference.

There is no known way of objectively measuring the quality of a musical instrument. Physical measurements are relatively crude when compared to the human brain. Also, making lots of objective measurements is very time consuming, and the next mandolin needs to get out the door soon or else the bank will cancel my credit card.

[HoGo]

Peter, the output can be as simple as classification of overall quality Let's say tone quality is 1-10 (subjective), volume 1-10... But I work in a field of fuzzy math, where you can describe it linguistically like "very bassy", "Trebley"... and assign value of membership (0-1) to tell the instrument has that property at some level. But you can simply sort the resulting mandolins into sample categories as "Loar sounding" etc. The reason of this "output" is that the resulting model will then tell you that if you use material A, thicknesses B, tuning C,... your result will most likely be .... And it's easily seen what you have to adjust to get into the category you want to get.
You can look at some publications of prof. Bodenhoffer at his personal page. (http://www.scch.at/index.jsp?menu=alphabetical&link=/staff/person.jsp&id=126)

[Brookside]

I wouldn't discourage anyone from running with such an idea. I like having you science types around to share your thoughts on building. Still, I am skeptical that you can gain any sort of edge with this idea. I think that an experienced builder can beat your machine in his ability to evaluate material A, thickness B, tuning C = your desirable result. Experienced builders have developed their brains into such a machine.

We've had some discussions in the past regarding blind taste-testing of mandolins. Most don't believe that a person can tell a Loar from a copy, a model A from F etc. And obviously, everyones opinion of what sounds "good" varies. However, most people, when they hear that particular sound that is pure magic to them, react to it in an instant. I think it takes a human brain to zero in on that. It will always be a far more capable machine.

It appears as though you are combining your passions (science and mando building) and having fun with it. The value in that is indisputable. It may even make you a more capable builder. Just don't be surprised if every once in a while some hillbilly builder comes along and beats the pants off of you. http://www.mandolincafe.net/iB_html/non-cgi/emoticons/laugh.gif

[Dave Cohen]

Peter, as another practising "luthier-scientist", I have a few thoughts.

One concerns the violin criteria of Dunnewald and a few others. Those are reputedly pretty good quality criteria. similarly, Meyer's 1983 paper was able to identify some quality factors in the classical guitar. Are you familiar with them? And if so, what do you make of them? I guess my point is that they are admittedly not perfect, but by dismissing them completely, we seem to say that we are abandoning any hope of finding objective criteria for instrument quality. I think that they are at least a start.

Second, I am not a plate tuner, and, Al Carruth notwithstanding, I would tend to discourage luthiers who are starting out from bothering with the plate tuning. The reason is the Schleske paper and the Atwood paper which I mentioned above. For the benefit of most of the folks in this thread with the exception of Peter, those authors attempted to follow "tuned" plate modes into assembled violins. Bottom line was that they couldn't. What they found instead was that whether the plates were "tuned" or not, the mode frequencies in the assembled violins were very similar. So plate tuning can certainly give you consistency, but if you are interested in breaking new ground, tuning is not much help. I am interested in being able to tweak and measure the asssembled mandolins in the white. I have a few ideas about how to go about it, and if I can ever get off my duff, I will get around to doing it.

[sunburst]

A couple of thoughts.
While an in-depth understanding of the state of the art in modes etc. is interesting, I feel that a basic knowledge is all I will likely benefit from. In other words, if I can develope a background knowledge of how plates and air modes and so forth behave in a mandolin, that knowledge will influence the developement of my "intuition". Another way to say it is, scientific knowledge can't hurt. It's one of the influences that is guiding me in my progress as a builder.

Oh and by the way, Brookside. I'm a "hillbilly" builder http://www.mandolincafe.net/iB_html/non-cgi/emoticons/smile.gif. That may not have been the term I would have chosen, but you're right. It's suprising, sometimes, what results a builder can achieve without applying much apparent sophistication to his or her craft.

[Chris Baird]

A good way to gather experience is to build a couple test mandolins. I built a fairly heavy duty rim and built 3 backs and 4 tops for it. I glued the plates on with slightly diluted liquid hide glue so that they could be easily removed. I started off thick with each plate and then regraduated them in different ways until each one failed due to being too thin. I found that plate thickness isn't as finicky as I'd thought. It takes alot of wood removal to alter the tone signifigantly. Altering the arching seemed to be the biggest change one can give a mandolin of the same shape. My arching heights ran from 1/2" - 3/4". Unfortunately I wasn't into observing any modal measurements when I built these test mandolins. However, I've learned much more from these test mandos than I have using chladni, fourier, tap tuning, etc.. It isn't all that time consuming because you don't need to worry about asthetics and isn't expensive because you can use plain wood. Eventually my rim broke while going through the thickness sander so I decided it was time to start building real mandolins. I certainly am glad I conducted those tests because they changed my current designs significantly.

[sunburst]

That's like what Pete Langdell does at Rigel. He can put the mando together and take it apart as many times as he wants. I talked with him about it and he said similar things about graduations and tone bars. They don't make a lot of difference at that point. He said aperature size was his best tool for tweeking the sound.

[Brookside]

Sunburst,
No offense intended and none taken I assume. I got that impression from you when I inquired about spray guns. You strike me as the type who could produce a quality finish using only a windex spray bottle. There may be no apparent sophistication to your method but I suspect that if we could crack your brain open and look inside we would find endless micro-judgements at work. In short, there is great sophistication to your methods. I assume that only came through experience. Practice, practice, practice....

[Chris Baird]

"He said aperature size was his best tool for tweeking the sound."

I have to say that I haven't been able to do much with aperature size. Usually my mandos come out somewhere around D - D#. I've often heard that the cavity should be tuned to some whole tone but I haven't found that to make any difference. I do all my air chamber tuning before I even start to build by designing the mando to have the right internal volume. I haven't been able to do squat by messing with the f-holes.

[sunburst]

Brookside,
No offence taken. I didn't assume any was meant. Yes, practical knowledge does accumulate through practice.

[sunburst]

Chris, that's interesting to hear. Mine come out to be between D and D# also. I've been thinking of experimenting with the F-holes since talking to Pete.
I'm not surprised, tho. Seems that every time I learn about something that works for one builder, another comes along that has a different experience.

[Dave Cohen]

Sunburst, if you had a little better grasp of the physics involved, you would have been able to exercise some scepticism about tweaking the sound with the aperture size. The Helmholtz air resonance frequency is proportional to the square root of the radius of a round sound hole, which means it is proportional to the 4th root of the area of the sound hole. While the relationship changes some with the shape of the soundhole, the differences are subtle compared to the effect of sound hole area. That means that if you increase the total area of the sound hole(s) by a whopping 25%, you only get an approximately 5.7% increase in the Helmholtz air resonance frequency. In other words shaving the edges of the soundholes a little bit as suggested by such as Siminoff, as well as by some who frequent this board, will do essentially nothing. Now, I don't know about you touchy-feely guys, but this science nerd doesn't even have to do a small part of one experiment to figure that out. The calculation can literally be done on the back of an envelope in less than or approximately 10 seconds. So much for how it takes the physicist such a long time to figure something out. I have conversed w/ Pete Langdell a couple of times now, and I am generally favorably impressed with him, but I have no idea how he has fooled himself into that one, if indeed he actually has.

[peter.coombe]

Dave, off hand I can't recall the Dunnewald et al papers or the work of Meyer, but I may have come across them when reading about violins and have forgotten. My PhD is in Entomology, so all this physics stuff is fairly heavy going for me. I would also agree with you in discouraging new makers to take up free plate tuning. I have been doing the same thing myself. There are so many other more important things that a budding Luthier needs to learn.

At one time I did start scoring my mandolins on a scale of 1 to 10 and then plot the score against various things I measured. Result was pretty much a random pattern, i.e. no correlation, which probably should tell me that plate tuning is a waste of time. However, there are problems with this sort of scoring because things change, and my perception of what is fine tone changes because the instruments I make improve. For example, I made an exceptional mandolin I could not fault the tone and gave it a score of 10. Then I made another mandolin that sounded even better so the score then becomes more then 10 (better then perfect?). I should have gone into more detail and scored the various aspects of sound, but there is still the problem of the passage of time and how can you be consistent over a long time period. It is a lot easier if you have all the instruments in a single room all at once.

I agree that arching and the tuning of the soundbox do have a profound impact on tone, but all these things are interrelated, e.g, increase the height of the arch and the free plate mode frequencies increase. There is no doubt the biggest improvements I have managed to get recently have not come through measuring Chladni patterns. These have come by changing the arch, modifiying the bridge, and changing the varnish. Red Henry's experiments have shown that big changes can be observed by changing the bridge. Personally I think he is barking up the wrong tree by insisting on using Maple, but he is certainly on to something in terms of bridge design.

[John Jesseph]

I am reposting this from Maestronet.com from a long thread on breaking in and tone. It is written by "jmasters", and not me. I thought some of it was relevant here. I only have an M.D. so I only know about snake oil and voodoo, but I find this all very interesting.

"For Michael and others who may identify with his postings:

Here is a change in the appeal I have been trying to make. There are diffences amongst technology, engineereing, and science. The actual relevant science to understand violin break-in (the original subject of the thread) is likely not to be made into a technology for a long time.

The CAS was not scientific, they were trying to develop an engineering technique based on normal modes. At this period in history normal modes are very old hat. Today is much later than the develpment of the (linear) mechanics of Lagrange (1736 - 1816). You can see that he was a rough contemporary of Joe Haydn.

The CAS had a few good ideas based mostly on Arthur Benade's success with winds, but likely not relevant to the most important violin questions I would pose. In particular, there is no room in Lagrangian analysis for mechanisms related to breaking in.

There has not been an "enormous amount" of research related to violins. There has been practially none. I have read a couple of things by Schleske and find him at the front door of a bit of science. He is doing what I have recommended for some time. That is, to statistically categorize a lot of instruments and try to develop a language for describing sound in violins. Later, one can try to correlate this with some kinds of measurements. I wish him and the others all the luck possible.

I erased many of my postings because I could not make the above points no matter what I said. I am now trying to restate a few things. You were right to throw out the CAS ideas based on Chladni patterns as models for anything at all. I threw them out when I was about 18 or 19. (First year or two of college) Traditional Chladni plates are flat. The results of what you saw in the comparison of outlines should be that nothing can be said about tone based on observing Chladni plates of different outlines. The different outlines may place some constraints on the sorts of "nice" archings makers are likely to use, and this may makes some differences in tone types. If one uses arched plates and calls them Chladni plates, then one may learn few things, very few. Prestrssing of the structure would not be considered even by arched plates in the Chladni model, of course.

One thing you non-science people need to look into is the nature of the person doing the research. LaFolia said this and he has shown himself to me to be a scientific person. Chladni may have been only an amateur scientist plus he lived a LONG time ago. I never heard of any other thing he did.

Acousticians: These are not people who look into music. I posted before what modern acoustics was, in part. They also look at all kinds of strange mechanical motions in solids and whatever that may mean for mathematical physics. Materials science too, and that means applied mathematics big time, because these are very difficult things to model. No academic who values his reputation is going to dealve into violins.

As for me, this "science type" has interest in questions which may or may not eventually lead to some understandings. I do not make violins with any "science" except the following:

I find an effective stiffness at the 5-mode, which is the one which is produced if you hold in the approximate center of one half of the upper bout and strike it with a knuckle. From this I compare to the mass of the plate and make a number which seems to guide me in the graduation of a plate. The idea is to compensate for varying stiffness and densities of different billets of wood. That is all. On the other hand, I likely would not have thought of this simple excercise if I did not think like a "science type."

I do no engineering technology which is what you and the others seem be confusing with science.

Hopefully, this will clairify things. As for me, I will not loose my interest in possible scientific aspects which might help to understand nearly anything about the violin or its sound perception by players. No more than I plan to stop playing violins or violas, looking through telescopes, or riding motorcycles.

So I ask only that you stop using the term "science type" because you are not making good distinctions as to what science may be. It is a kind of put-down that is not warranted. I am a "science type." But I do not try to use science-derived engineering or technology in making violins; simply because things are too complicated for a decent investigation with no money or workers. However, some intuitions may have guided my thinking. You are right to be sceptical of science claims, but please ask yourself in the future if a particular model or idea is the result of actual science.

This seems to be a reasonable request. "

[sunburst]

Dave, I follow you on the limits on ones ability to change the Helmholtz frequency by trimming the F holes. I used to try to do that on my early mandolins, a la Siminoff, and found that you're pretty restricted in how much you can do.
Are you saying that the Helmholtz frequency is the only thing that changes with F hole size?

[Dave Cohen]

Peter, having said disparaging words about plate tuning, I will offer what is on the other hand. Carleen Hutchins is aware of Schleske's and Atwood's papers, and she still plate tunes at 90 something yrs old. Al Carruth still plate tunes. Plate tuning may very well have some merit; I just can't make a connection with what that merit is in terms of Newton & Hamilton & LaGrange. So I certainly didn't mean to imply that plate tuning is all wet.

John J., that was an interesting post. I agree with "jmasters" on some points, but not on others. For one thing, Lagrange developed a powerful formulation of classical mechanics. It is as relevant today as it was in Haydn's time, as is Newton's formulation, and as is Hamilton's formulation. The three formulations are just different ways of saying the same thing. Classical mechanics is still completely valid in the macroscopic domain, i.e., for anything bigger than small to moderate sized molecules. The same goes for normal modes. I had this conversation with a physicist at the 2001 GAL convention. He felt that normal modes were only the linear part, and when an instrument body is driven hard by strings, things get nonlinear fast. He was right, but I countered that one has to find out the linear part before proceeding on to the nonlinear part. A lot of non-scientists, and apparently even a few scientists, seem to think that just because we can't explain the entire instrument immediately with science, the whole scientific approach is invalid and will never have any success. I don't buy that. Science is just too much fun; it is just too exciting. And for me, it was the fastest way to get to where I am. With only the traditional right-brained intuition, I wouldn't yet be making anything but toothpicks and fiddlesticks.

John H., when you change one thing in an instrument, you also make a small perturbation in everything else in the instrument. Nevertheless, the predominant effect of increasing the soundhole size is an increase in the Helmholtz air resonance frequency. So the answer to your question is 'mostly, yes".

[John Bertotti]

Science guys keep doing what your doing. I tend to believe that everything discussed here has an effect on the instruments being made. Maybe what works for one builder and not another isn't just a matter of materials used, but the processes involved. Any # of other variables, that with our level of sophistication at this time, may seem meaningless could be discovered to be tantamount to the construction process. Who knows, we are just infants learning as we go, and grow. It may be a instinctual leap a luthier makes one day that will unravel the mysteries of our mandolins or it may be the data accumulated by the science guys. Either way read a lot, play a lot, let your strengths guide you and always ask why. Why and what next. Just my thoughts. John

[Chris Baird]

Certainly if one is inclined towards math, science, physics, etc.. than it is natural and logical to build using the precepts found in those disiplines. My point I was trying to make ealier is that if one isn't predisposed to science or math that all is not lost for your future in building. I believe that all relevent information that one needs to have to build a good instrument can be obtained using empirical methods devoid of measurement equipment. Of course some of this information will come as cognitive realizations and some will be internalized and never make a run through the cognitive process. I think the human body and mind are capable of much more than what is easily observable in the conscience state. If I were to not allow myself to internalize many aspects of building by isolating my body from the measurements I just wouldn't enjoy it as much. I don't like to have my conscience brain "on" all the time. Life is better when I allow myself to be "touchy feely" and I don't think I'm less effective for it. The ultimate goal is music and music will always dwell in the realm of feeling.

[peter.coombe]

Thanks Dave. I still plate tune and don't intend to stop because I know I can improve how responsive the instrument is by doing it. However, more recently I am thinking that maybe other factors are more powerful when it comes to the tone of a mandolin. Plate tuning still has it's place, but you can still build fine sounding mandolins without it. Most of what we do is basically educated guessing when trying to improve tone.

[Rave On]

I put together a spreadsheet to calculate Helmholtz frequency
from the formula I found in the book the 'Physics of Mucical Instruments'. The Helmholtz frequency (If I did this correctly) is fairly insensitive to changes in the F hole size. #It is very sensitive to changes in the depth of the F holes. #Three things: 1. The depth of the F holes is not constant because the top is graduated. 2. Changes in the depth of the F holes are small absolute changes (.12 inch to .14 inch = .02 inch) but large relative changes (.02/.12 = 18%) 3. Changes in the thickness of the top will affect tap tuning because it is also changing distribution of stiffness and mass. #This is all very complicated. #Has anyone built finite element models of mandolin or violin tops? #Also I would be interested in conjectures: If tap tuning works...why does it work? #What is the goal that tap tuning helps us achieve?

[Dave Cohen]

Rave On, the formula for Helmholtz resonance frequency which you want to use is the one for a "neckless" Helmholtz resonator. I asked tom Rossing about that one on my first trip to DeKalb. He dismissed the effect of top plate thickness at the soundhole as being relatively unimportant, unless, of course, you build some baffles or walls around the inside of the sound hole(s). Recall that Tom Rossing is the author of the book to which you referrred. If you compare the results from the formula for a "neckless" Helmholtz resonator to those from the formula for a typical long-neck Helmholtz resonator, you will see that there isn't much difference if you substitute in various values for the top plate thickness at the sound hole.

If I can borrow the use of some FEM software near my location, I will do some modeling, time permitting. It is on my list.

I have no idea why plate tuning works. Its' proponents advocate it strongly. Schleske and Atwood seem to think that it doesn't work, or at least that its results are fortuitous. It is certainly a way to be consistent. So if you find the tuning that works in a 'good' instrument, you can repeat it and thereby produce consistently 'good' instruments. That is, if the wood allows you to do so.