Official Luthiers Forum!

Here I go on another of my journeys of thinking when I should be building...

Does anyone have experience/thoughts/opinions on different back bracing styles? Tonally, structurally, and practicality of constructing by hand. Standard ladder bracing, versus asterisk, lattice, carved arch, and so on? This is in the context of a lively back designed to increase sustain and color the tone, a'la Somogyi's weight on a rubber band analogy http://www.youtube.com/watch?feature=player_detailpage&v=fAEQKmyB5Nk#t=145s

As I understand, carved arch is generally very stiff and mostly just reflective on plucked instruments.

Lattice is likely capable of the best tone and structure, but kind of a pain to construct, especially for me without power tools to produce 10-20 little sticks of precisely equal dimension. However, a simple lattice of just 4 sticks (double X) would be easy and should still work well.

Asterisk type layouts seem like they have a lot of potential, with fairly easy construction, and stiffening in all directions.

Ladder bracing definitely has ease of construction on its side, but the limiting of side to side motion and lack of longitudinal structure both seem less than ideal to me. Like you're missing out on a lot of what the back can do tonally, by promoting mainly monopole motion and longitudinal waves at mid to high frequency between the braces, rather than at divisions of the full length of the back.
However, it's been done a zillion times and obviously works well... maybe the kind of motion it promotes is what sounds best. Or maybe the monopole is the only mode that really matters.

Here are a couple layouts I doodled up for potential construction on my re-top test guitar (maybe I'll make it a re-back test as well?).
The first is a pretty standard asterisk, but oriented so the third line of it follows the centerline (conveniently doubling as a seam reinforcement), rather than spanning the lower bout, and an upper transverse brace to support the nearby headblock foot. Laid out to divide the back into roughly equal areas.

Next is a layout based more on an A-type structure supporting against neck rotation, and not isolating the upper bout from the rest of the plate like the other layout. Headblock foot may not be necessary in this layout, as the braces converging there provide the same stiffening effect.

And of course there are the more common asterisk forms using one or two transverse braces in the upper region, and an X in the lower bout with optional third line spanning the lower bout. I'll probably go with either that or my longitudinal asterisk, but I figure it could make an interesting discussion before I start gluing things together

There are many different back bracing patterns . The back is not a major sound influence but it can be fun to experiment . As long as the bracing supports the back , that is the major contribution . Gibson built a few models without any back braces and used a lamination design to totally eliminate it .
There is a dampening effect of the back to the guitar but that is more influenced by it resting against your body .

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

Interesting.
I would like to hear what happens.
I think a loose back pumps air better,
but too loose might be bad.
I think I like the x meeting in the center of the lower bout,
so the back might flex there more?
I am using the ladder brace for now,
and I am not that experienced yet.
Good luck, and let us know!
Anybody tell me what the difference between a tight and loose back does?
Projection, focus, overtones?

That pic of the Reynolds back is about the prettiest pic I've seen of a guitar.

Thanks for the replies so far!

Filippo, that Reynolds layout is a good one, I may give something like that a try.

Todd, thanks for the analysis of my second layout. Pretty much tracks with what I was afraid it might do. The angled braces in the upper region would help spread the load of the neck to the waist, but being that the whole thing is kept fairly flexible and that the waist is so far away, the back itself would likely have taken much of the force by then, thus transferring to the fairly vulnerable upper bout curves. Probably best to sacrifice the upper bout to structure on both the top and back, as in the Reynolds layout... which brings us back to the common asterisk with two cross braces above it

Alan, I'm not experienced either, but I can't just go on following tradition unless I have good reason to believe that there's no improvement to be made I was reading Somogyi's books again and even he doesn't know that much about how the back actually behaves and influences tone, but says that classical people are more into research on it than steel strings. To me that says the tradition of ladder bracing is one of "if it ain't broke don't fix it", rather than something like the violin's current state of "nobody has figured out how to make it sound better yet", which I have greater respect for.

I dunno which would be better... X in the lower bout, or X centered between the UTB and tailblock. The first gives you a nice round drum-like area to pick up vibrations. The second gives you a larger area with a clear pivot point for long dipole/cross dipole/monopole motion. I think I like the second better. The funky shape may in fact help it to resonate at a wider variety of frequencies than a circle would. And the larger area/mass you get resonating, the more energy is maintained to be pumped back into the system.

And as I understand it, a tight back gives more projection, a loose back gives more sustain... but only if it resonates well with the top and air mass, otherwise it's more of an energy sink. And if the main tap frequency (monopole) of the back is about 2 or 7 semitones above that of the top, it will probably work pretty good. I think that's why it works well to try and get the longest echo you can from tapping the box before gluing the bridge. That sets the plates to equal frequency. Then the added mass of the bridge drops the top about 2 semitones, producing the proper interval.

My own guitar has a heavy and stiff back because it has a center wedge of African blackwood (it's a 3 piece). That stuff can stop a bullet. The top however is a bit too light. As a result it offers a very pleasant experience to the player, and overall in a not too big room, but in a larger venue the projection is just OK at best. I got better projection from stiffer tops, coupled with a normal not too light, not too heavy generic Indian rw back.

This was not some huge revelation, but just reminding that it still seems it is the top that matters in the first place.

_________________
Build log

2 to 7 semi tones?
Is that possible?
I recall John Hall saying 1 semitone.
Uh-oh, I'm geting confused.

There's some really good builders like Tim McKnight, Howard Klepper, Paul Woolson ect. who use alternate back designs and bracing with great results.......but, at this point I don't need another variable. Maybe someday it will be something I'll experiment with, but before I do I want to have a good understanding of what effects it will have.

_________________
I didn't mean to say it, but I meant what I said.
http://www.brackettinstruments.com/

Nobody ever said it would sound bad if it wasn't 2 or 7
Actually Somogyi mentions there are a few schools. One says 1 or 2 or 3 semitones, another says 2 or 7, and some Kasha bracing styles go for 2 or 7 below the frequency of the top... so maybe what's more important than the specific interval is just that they are not too close to eachother, as Todd L. discovered and fixed in this thread http://www.luthiersforum.com/forum/viewtopic.php?f=10117&t=30241

In my opinion, it's a mistake to even talk about the tonal advantages of one bracing system over another. There are so many variables that a builder has to decide on within any scheme that it really becomes a question of the builder rather than the pattern. To say that system A sounds this way and system B sounds that way just doesn't work.

My belief is that most good builders have chosen a system and continuously tweaked every element until they have figured it out. It's not a matter of the system, it's a matter of the tweaking.

One thing I will say about back bracing is that I believe a lattice (more bars than a double X) should do a better job of preventing splits than other systems. And for a new system, if you have doubts, err on the stiff side and you can always shave braces after the fact.

_________________
http://www.chassonguitars.com

Well my approach hasn't been based on mathematical models or scientific studies, it's all been very much "hunch" & ideas methodology but I've built my last three guitars using two backs, an inner & an outer. The outer back is braced stiffer to take the external loads imparted on it during normal use, this inturn allows me to brace the inner back lighter (both are ladder braced) & because the outer back takes all of the structural loads, I don't run the inner braces right out to the linings, instead I carve a recurve similar to an archtop in this 'band' between the braces & the linings & 'tune' the back as you would an archtop.

The bracing on the inner back is on it's back thereby presenting a 'clean' ,smooth & unhindered face in the active soundbox.

I decided to try this after thinking about the very same problem that John brought up, that of the player dampening the back during playing (I'm in the camp that thinks that the back has a major contribution to make in the overall sound of the guitar, tonally & projective wise) & also eliminate any possible 'bad' interaction the raised braces may have with the soundwaves in the box. As I say, this has had no scientific research or figures to back it up but more importantly to me, it seems to work. There is more sustain when played & I've compared my guitars with 'factory' produced guitars of the same 'size' & there is definitely a small volume increase in the double back, these points have also been noted by the people buying or trying them.

_________________
Nick Oliver

http://www.oliver-guitars.com

Can anyone point to a definitive comparison of the relative tonal merits of parallel guitar builds where the main difference is difference in the back bracing pattern?

There may be good examples, I just don't know of them.

I've always used plain old ladder bracing on all my guitar backs, and the main thing that leads me to think that a more longitudinally controlled pattern may be advantageous is the problem of reaction to humidity changes and trying to control the development of a waviness to the back. This is totally a structural consideration and is not motivated by tonal concerns at all.

The picture of Randy's guitar is just yummy. No set of wood that nice has ever crossed the doorstep of my shop yet.

_________________
Jim Kirby
kirby@udel.edu

I've used a number of different back brace systems over the years, and seen several others used by good makers. They all seem to work pretty well acoustically if you've got them 'tuned' right. At the moment, I'm running with ladder bracing, simply because...

Right now, the best model I can come up with for how the back efects the sound is this:
The top is the only part of the guitar that can produce much sound and which is also directly driven by the strings. Everything else gets whatever energy it gets from the top somehow or other.

The top is also lighter in weight than the back, and both less stiff and less heavy than the sides. In other words, energy going into the top is more likely to produce sound than energy going someplace else.

OTOH, the back can help pump a significant amount of air through the soundhole and thus produce sound at the 'main air' pitch. It does this most strongly if the 'main back' (monopole) resonant pitch is matched with that of the 'main top' mode. This, however, is also very likely to produce a strong 'wolf' note at the 'main air' pitch, at very least. The 'safe minumum' pitch separation seems to be about a semitone.

I've found that the 'main top' resonant mode tends to drop about 1/2 semitone in pitch with the first month's 'playing in' on my guitars (YMMV). I've also typically seen a drop of about 1/2 semitone in the 'main top' pitch when gluing on the bridge, although this obviously depends on the bridge weight, outline, and so on. Starting with the 'top' and 'back' monopoles at the same pitch before gluing on the bridge thus usually works out well. Sometimes you get a bit of a 'wolf' on the attack initially, but this usually plays out over a few days or a week as the top pitch drops.

Above the 'bass reflex' range, roughly from the open G string upward in pitch, back resonant modes tend to be 'losers': they typically show up as dips in the spectrum chart, and seldom as peaks. Minimizing the depth and width of these dips makes some sense. You don't want to eliminate them altogether, since they seem to contribute to 'tone color'(Ovations lack these dips, and tone color too), but you want to cut the losses down. This calls for a back that is fairly heavy and stiff (since it can't absorb sound in damping when it's not moving much), and that has low damping (so that the dips are as narrow as possible). Sounds like Brazilian rosewood to me...

Both of the Knutsen guitars I have use a slanted ladder bracing - someone can probably tell me what the angle he used was. I'm pretty sure he did it because he thought it looked cool; but then again, most of his guitars have the same back bracing, so maybe he thought there was some benefit to it.

Hawaiian guitar


Harp guitar (in need of lots of restoration, in case you're wondering about the nails and the lack of frets)

The neck block on that Reynolds guitar picture is interesting. I wonder how people feel about the advantages of a truss rod that adjusts at the nut end.

_________________
"Preoccupation with an effect gives it power and enhances the error"
from "Your Owner's Manual" by Burt Hotchkiss.