Bass side/treble side, or whole top?

[Brian Evans]

I was thinking about left handed guitars today. I think that a lot of builders have asymetrical bracing, but do you really think the top vibrates differently for treble strings on one side than the other, or does the whole top vibrate for all the strings? Would building a mirror image top be a requirement for a left handed guitar, or just change the nut and bridge? I personally think the whole top is engaged with all the strings, but that you could build one part of the top to emphasise trebles vs low notes. I have been thinking about this because of a post about parlor guitars - one person said he built one and made a "mistake" - braced it backwards - but that it turned out great anyway. I have always thought that the bridge drove the top symetrically, that it didn't favor one side of the top over the other in terms of transferring vibration to the top.

[Bob Gramann]

I have always thought that the unsymmetrical Martin style arrangement of the tone bars just made the sound a little more complex. When you think about the wavelengths involved at most fundamental frequencies produced by the strings, and the strings being at most 2 3/8 inches from each other, how much more does it vibrate on one side than on the other of the bridge area for a given note?

When I’ve built lefties, I have reversed the bracing, but only because I figured that’s what someone looking at it would expect. I make the area around the bridge very stiff on my guitars. Like you, I don’t think the side the tone bars connects to makes much difference. We hope the top vibrates as a whole except for smaller areas vibrating perhaps giving us some expression of harmonics. Making the stiffness asymmetric provides more of a chance for the smaller area to be unique and not part of a diopole movement that would cancel out the sound from that area.

[Freeman Keller]

Brian, I am the "one person" who build the bass-ackwards parlor. I understand that is fairly common with these plans, I think it is interesting that his other plans are drawn conventionally.

I'll leave it to those folks who understand bracing and vibrating modes and all that stuff to explain how it all works. At my level of beginning lutherie I just try to follow plans for something that is fairly well proven and I hope that when I'm done it doesn't sound too bad. I know that people move braces around, "tightening" or "loosening" the top, I have sat at the feet of some folks who I think know what they are doing as they explain voicings and tapping and glitter modes and all that stuff - but I'll be the first to admit I really don't understand it very well.

A few random thoughts, however. Here is picture of the "mistake" parlor

and here is one of the later ones

I have played these guitars side by side, one is mine, one is my daughters. They are slightly different woods (sitka over EIR, lutz over Mad rose) and they are different age so an apple to apple comparison isn't fair. You'll notice that one is scalloped, the other not. However, my carefully calibrated ears tell me that there isn't much difference. If anything, and this is a big IF, the lefty sounds slightly less complex and the word I like to use is more "boxy". Obviously these are small guitars and will not have huge bass but they are both full in the mids and highs and both are pretty loud.

Second, I have converted several guitars to left handed including a Martin OM-28 and have always stayed with the right handed bracing. I do the usual stuff, bridge, nut, neck dots, p/g but leave the insides alone. Left handed players seem to like them just fine.

Last comment, and this one is kind of interesting. I was commissioned to build a Weissenborn for a friend. I decided to make one for myself at the same time. I ordered enough wood and large enough pieces to literally make them out of the same wood (all koa) and they were made from the same plans in the same molds. I used the traditional plans from StewMac and LMI - traditional Weissenborns have X braces but ladder cross braces below the bridge plate. At the same time I was building, there was an article in AL about Weissenborns showing pictures of one with modern Martin style tone bars. So I built one that way.

When the whole process was over, my friend and I sat down one evening and played them back to back to back. We agreed that they had a very slight difference in sound - if anything the traditional bracing was a bit cleaner sounding, maybe less overtones. We agreed that while they both sounded "good", it was slightly better. My friend got the pick of the litter, thats the one that he chose.

Don't know what all this means, hoping that someone that really does will chime in. Meanwhile, I'm building a nice OM sized guitar with pretty much standard Martin angled tone bars and enough scalloping that the top seems to ring. With luck the finished guitar will be OK

[Bryan Bear]

I sure don't know enough to shed any real light on the question but my feeling is that there isn't really a bass side and treble side. The top works as a whole. With that in mind, I don't see a whole lot of reason why reversing the tone bar would make too much difference. The string spread is relatively narrow compared to the size of the top and the bridge area is the stiffest part of the top so I would expect the vibrations from the saddle to be fairly evenly distributed to the bridge/bridge-plate area. From there, I would guess, the asymmetry of the top bracing is its own factor no mater which way it is facing. Again, these are just my thoughts. I hear people talking about thinning a top behind the bridge for this or on the sides of the bridge for that; I have never heard anyone talk about thinning the bass side or treble side.

Freeman, I don't think you can reliably ascribe much of any of the difference in tone of those parlors to the direction of the tone bar. The two pictured are just so different. Scalloped vs. tapered, finger braces on only one, and it looks like the X layout is different on both with the angle/position of the tone bar being different too. As you say this is not apples and apples. Now the other two guitars are much more similar other than the tone bar. BUT, we don't have one right and one left, we just have a different orientation so we still don't know what effect reversing them would have.

This is fun, isn't it!?! I'm just a hobby guy so I get to enjoy building by feel and pretending like I might have an idea of what I am doing. It is always such a pleasant surprise when they are strung up and sound like guitars <g>

[Alan Carruth]

Aside from where the strings are, I don't think there is a 'bass' or 'treble' side. However, as with almost everything related to the guitar, it gets complicated.

Most of the energy from the strings that gets into the top and produces sound comes from them pulling the top up and down as they move 'vertically' with respect to the plane of the top. The bridge is reasonably massive and stiff, and moves as a unit, and the strings are all pretty close to the center line of the top. Thus the most important movement of the top is 'loudspeaker like', with the whole lower bout moving in and out, pumping sound off the top and air through the sound hole. In this sense the guitar is like a 'bass reflex' speaker cabinet.

Speakers are carefully designed so that they don't do much else: they move in and out like a piston for the most part, and when the don't it's a 'defect'. Unlike a good loudspeaker, though, the guitar top has a number of fairly strong resonances: 'modes' of vibration wherein it breaks up into smaller areas, each of which is moving counter to the ones next to it. The resonant mode that works like a loudspeaker is often called the 'monopole' or 'main top' mode. If it were not for the air moving through the sound hole and pushing on the top, it would tend to have it's maximum activity at around E on the D string, 165 Hz or so. Because of the air it gets pushed up in pitch to closer to the open G string, usually, although this varies a lot. Note that when we say that a resonance happens at 'X' pitch, that doesn't mean the thing only works at that frequency, and otherwise doesn't move. Strings work that way, sort of, but they're odd in some respects. Most things, like the guitar top, will move most at their resonant pitches, of course, but can still be moved more or less easily well off that frequency. Some researchers have said that the 'main top' resonant motion is actually the main sound produced on the guitar all the way up to 1000 Hz. It's the most effective way to make sound, and doesn't have to move much to produce usable power.

On most guitar tops the next higher resonant mode in pitch is the 'cross dipole', where the bridge rocks crosswise, with one end rising as the other falls. If the top is really symmetric the fulcrum line runs right down the center line of the top, between the D and G strings; with an asymmetric top it's off to one side or even tilted. For the middle strings, the D and G, driving this mode is like pushing on the middle of a see-saw; they don't get very far with it. Even the E string may not produce much sound from this, though; with one side rising as the other falls there's a lot of cancellation of the output. Some folks feel that a asymmetric top will work better with this, since the moving areas can be different sizes and so not cancel out. I'm not so sure: if the smaller area moves with greater amplitude it can still cancel out. What this mode does do is change the direction of the sound coming off the top.

Keep in mind that, even close to the dipole's resonant pitch (often around D or E, ~300 Hz more or less), the 'main top' loudspeaker motion is still at work. The string is pushing them both, but with one side of the bridge rising and the other falling from the dipole the motion on adds up differently on the two sides. One side of the top will be moving more than the monopole alone would allow for, and the other side less. The sound tends to be 'beamed' out more toward one side of the guitar or the other. Which side produces more sound depends on which string is doing the work, and at what pitch: things can change a lot with the shift of a semitone, even on the same string.

As you go up in pitch, the configurations of the resonant modes became more complicated, with a larger number of smaller areas at work, but the same rules still apply. And, of course, the top is not just producing one note at a time. What you get, then, is a complicated product of all of the different things the top is being asked to do. Being basically stupid the top doesn't know that it's supposed to get confused by all of this: it just does what it does. We, however, being smarter.....

I made an 'almost matched' pair of guitars some years back; one with 'normal' asymmetric bracing, and one with 'double X' , and had people at an ASIA Symposium try them out. They were pretty similar, but the double-X top tended to sound more 'modern', while the normal one was 'traditional'. Overall they slightly preferred the double-X sound, by a 2:1 margin (a lot of people liked it a little better).

I think that the symmetric top may be more efficient, but perhaps less 'interesting'. OTOH, there are lots of things that introduce asymmetry into a top; stuff like variations in the grain from one side to another ('bear claw' would be an extreme example). You can also restore some symmetry to a asymmetric pattern by the way you profile the braces. Dana Bourgeois makes a model where he does not scallop the X brace on the treble side, to 'stiffen up the trebles'. IMO it works more by reducing the asymmetry if the stiffness along the two diagonals that is introduced by having the tone bars parallel to the bass side of the lower end of the X. With the non-scalloped bar the tap tones are stronger, and the Chladni patterns are more symmetric (which is another way of saying the same thing).

The bottom line, as far as I can tell, is that the direction of any asymmetry probably matters a bit, but less than the fact that there is asymmetry in the first place. My opinion, FWIW.