Official Luthiers Forum!

I have been having a discussion with a Brazilian builder and he asked me about the pitch of the closed box. I had no answer for him and said I would ask the experts.
BTW, we are talking classical guitars in his case.
He is under the impression that he should try and get the resonating frequency of his closed box as low as possible. He usually ends up around "G". Recently he has been experimenting with a tornavoz and has been able to get it down to "F" Is the goal of trying to get as low a resonating frequency as possible a good one? Also, you can measure the pitch of the top, back, sides etc. but once they are attached to the rims all bets are off.
What about tuning the top and back to different pitches? what are most folks doing in this area? I have heard a semitone apart is good.
This person has made almost 600 guitars and since he doesn't speak English it has been a long, hard road for him. He doesn't have access to information on forums like this one. I learned to build from him and help him gain access to information anytime I can. In this case I would hate to steer him in the wrong direction so any help or insight others can offer is greatly appreciated.

Hey Robbie you bring up some good points and some very good questions, I know I read you can cover the sound hole and get a reading of the top and back plates of a closed box separatly ! however I did match the top and back as closely as i could before assembly, now when I cover the sound hole with masking tape, and measure the top and back frequencies by tapping , I am getting exactly the same pitch reading . I would like to think i was that accurate but I realy dont think so ! and I am not so sure what is happening , so I have ordered another sound card to perhaps get a better reading . I am very curious as to what some of the pros will say about this ! Jody

The goal is not to get the lowest frequency main air resonance; it is to get the most support for the guitar's bass range. Toward that goal, lower resonant frequency is good up to a point. I'm happy when it's around G, or about a semitone below. Thing is, there are trade-offs (surprise!). Resonance isn't all at one frequency. It has a bandwidth in the shape of a bell curve. A conventional size and location of the soundhole strikes a pretty good balance between the bandwidth and the strength of the resonance. You can do things that would lower the pitch of the resonance (actually, since there is a bandwidth, there isn't a single resonant frequency; what we call the resonant frequency is the top of the curve, where the amplitude is greatest and easiest to hear). For example, making the soundhole smaller and moving it closer to one end of the body will lower the frequency, as will attaching a tube to it. These things all make the box more like a true Helmholz resonator. Unfortunately, at the same time as they lower the frequency and increase the amplitude of the resonance in the center of the curve, they also narrow the bandwidth. So, for example, if you did all these things and got the pitch down to E, your low E would sound great, but you would hear a steep drop in volume as you went up a tone or two. G turns out to be a good compromise. A box that has its maximum amplitude at G doesn't usually get very strong support of the low E, but it gets enough bandwidth to be reasonably supportive of the bottom half-octave or so.

You can also lower resonance by thinning the top and back; again, there are tradeoffs, both structural and tonal. To my ear, this tends not to narrow the bandwidth so much, but if taken too far it makes the bass boomy, with what seem to be undesirable peaks in the resonance. I'd speculate that this is from uneven coupling between the top, back, and air. It also can dull the midrange. No free lunch.

Where's Al??

_________________
Howard Klepper
http://www.klepperguitars.com

When all else fails, clean the shop.

I think he heard the commonly quoted saying that Bream nagged Romanillos and others to build him a guitar with the air reso in E.

To pick up from where Howard left it, I find that large and thin-plate classical guitars, while being loud to the player and open, and with big bass, they have less interesting treble and overtone content. I enjoy more the sound of a small guitar with more restrained but fast bass and "crystalline" treble. I guess that is what a Romanillos tends to be with its rather thick top and small size.

_________________
Build log

I'm with Howard!
G is a good compromise!
But as guitars age and loosen up sometimes I go G# for those crystalline trebles Alex mentioned .
The bass loosens up also with age !
G# (the third of E major)also is not used in music as much as G -so if I were to have a very loud or(wolf) dead note G# is a good choice.
My clients love a good bass responce BUT it's the trebles that need to cut through at a concert!

Mike

_________________
Mike Collins

On steel string guitars, I generally like the body resonance, (main air mode) to be between G and G#. Putting it between notes spreads out the "wolf" note" (which tends to make the note sound thuddy) and makes it less noticeable.

I'm not yet sure what is best on nylon string instruments, but I suspect between F# and G would be a good target.

Mark

First, to answer the question: usually getting the low resonance to come in around F# or G is good. As has been pointed out, this helps to reinforce the fundamental on most of the lowest notes of the guitar in normal tuning. If you put the resonance around E then the strength of it is dropping off quickly as you go up, and you've wasted about half of the peak.

Howard Klepper wrote:
"For example, making the soundhole smaller and moving it closer to one end of the body will lower the frequency, as will attaching a tube to it. These things all make the box more like a true Helmholz resonator. Unfortunately, at the same time as they lower the frequency and increase the amplitude of the resonance in the center of the curve, they also narrow the bandwidth."

I don't know abou moving the hole toward the end of the body, but my experiments show that making the hole smaller or putting in a tube will _lower_ the height of the resonant peak, and make the effective band width greater (lower the Q value). Both of these things increase the drag of air moving through the hole: the smaller hole because the ratio of edge to area is greater, and the tube because the surface area in the tube is increased. There may be other things at work, too.

You have to understand that the 'main air' resonant pitch that we're talking about here is not an independant variable. It's actually the lower output peak in a 'bass reflex couple' that involves the air and (at least) the top. If you bury a guitar body in sand so that the hole is open but the wall can't move the 'Helmholtz' air mode will be up around 110-130 Hz ( A-C#). The 'main top' resonant mode on a backless guitar would be somewhere around 150-170 ( Eb-E+). On the completed body the two have to work together: the top moves air through the soundhole and the air going in and out pushes on the top.

In a system like that there wil be two stable states; in phase and out of phase. In other words, at one pitch the top will be moving 'out' as the air rushes into the soundhole, and in the other the top will be moving 'in' when the air is flowing inward. In the first case the mass of the top tends to slow the air flow down, and the pitch drops from it's isolated (pure Helmholtz) value, to about G on many guitars. This is what we call the 'main air' resonance. In the second case the air pressure acts as a restoring force on the top, in effect stiffening it up, and the pitch rises fto around 196, the open G string.

Lots of things effrect this! If the back is flexible enough it will get into the act, dropping the 'air' mode pitch, and possibly effecting the 'top' as well. Reducing the depth of the body will make the coupling stronger, increasing the seperation between the 'air' and 'top' mode pitches, but it also raises the 'real' Helmholtz pitch. The end result can be that the _'top'_ resonant pitch rises, while the 'air' pitch remains unchanged; just the opposite of what you'd expect.

The point is that, with all that's going on, it's difficult to predict exactly where things will end up, but given a certain outcome you can figure out ways to move them in the direction you want next time. Fortunately, the traditional designs are such that things come out pretty close to right if you do a good job, and you can usually get what you want with a few tweaks.

Alan,
Thank you. I was hoping you would comment on this post. Everyone has given great advice and here is what I have gained from this. Please tell me if I am going in the right direction.

The general consensus is that The pitch of "F" that my friend is going for is a bit extreme and that shooting for a "G" or even a "G#" or "F#" would be better. Anything below an F# and you begin to lose the fundamental of the bass notes. I assume this would get worse in D tuning. Also, if using a tornavoz would the resonating pitch need to be a skoshe higher due to the fact that the tornavoz will lower the pitch once it is installed.
Also, should the top and back be a semitone apart or do we want them to resonate at the same pitch?
Here is another thought. If the tornavoz lowers the resonant pitch thus affecting the bass, will a soundport affect the treble of the instrument? Has anyone done a guitar with both a tornavoz and a soundport? What would be the expected result of having both a tornavoz and soundport other than some purists turning up their noses at the guitar?
Thanks everyone for your comments and I look forward to more.

Robbie O'Brien wrote:
" Anything below an F# and you begin to lose the fundamental of the bass notes. I assume this would get worse in D tuning."

A lower pitched 'main air' resonance would probably help the lowest notes more in drop D tuning. I made a 'sorta' baritone for Ken Bonfield recently on my small Jumbo platform. I made the sondhole smaller to drop the Helmholtzt pitch so that the resulting 'main air' would be low enough for the intended D-D tuning. The actual 'main ar' resonance came in at 91 Hz, a little below F#, and he normaly uses it tuned C-C. There is still enough fundamental on the low notes, in part because the so-called 'neck' mode ended up at 80 Hz, close enough to couple with the 'main air' and give a lower pitched peak in the output spectrum down around E, four frets up with the low string tuned to C if I'm counting it right. BTW, that one turned out really well, and he's played it capoed as high as the seventh fret without a loss of tone. Some days you get the bear.....

"Also, if using a tornavoz would the resonating pitch need to be a skoshe higher due to the fact that the tornavoz will lower the pitch once it is installed."

More than a 'skoshe' I'd say. A tornavoz as deep as Torres used will drop the 'main air' pitch down to about 80 Hz (from 97 on the same guitar without it), and the amplitude drops to about 1/4 of what it was without the tornavoz. The low end pretty well just goes away.

"Also, should the top and back be a semitone apart or do we want them to resonate at the same pitch? "

I like to have the back end up about a semitone higher then the top when all is said and done. This is close enough to get a lot of enhancement in the bass and far enough apart to avoid the 'wolf' effect, at least on my guitars. YMMV!

"Here is another thought. If the tornavoz lowers the resonant pitch thus affecting the bass, will a soundport affect the treble of the instrument?"

Everything effects everything. Have you read my article in the most recent 'American Lutherie'? It gives some measurements of how a port effects the output in various ranges, although a lot more could be said. In fact, I did say a lot more in the orignal submission, but we cut it down and made it a bit simpler in the final edit.

"Has anyone done a guitar with both a tornavoz and a soundport?"

LOL! Properly done they could cancel each other out in terms of the overall timbre of the guitar, and pretty much just change the radiation pattern. Actually, that's more or less what I do when I make a ported guitar: I reduce the sze of the 'main' soundhole to get the 'main air' to come out at the 'normal' pitch. The port does alter the sound output at higher frequencies some, but only locally, and the overall timbre is not all that much different from my usual guitar; it's just pointed in a different direction!

Thank you Alan for sharing your insight and for taking the time to answer my questions. This whole discussion has cleared up a lot of things for me. I hope others have benefited as well.
I had misplaced the American Lutherie magazine with your article and just found it again yesterday. I am reading your article tonight.