Official Luthiers Forum!

O.K. Alan , thanks Jody

I had a guitar made by a very fine English maker about a dozen years ago
(very beautiful instrument, and great guy) which had the the problem Alan
mentioned initially, imho: main resonance (don't know whether air or top)
seemed to be too low. Also, this was a smaller- than-normal classical,
35cm lower bout. The upshot was a sixth string with a strong fundamental,
(very nice), but a thin fifth string, and a downright *stringy* fourth string.
I *really like small guitars*, for various reasons, but this instrument made
me wonder if more new problems are created in their making than are solved.
I'm still going in that direction (34cm lb) in my own making in an effort to find out. I think the back was quite stiff in the instrument I commissioned, and
will be looking in that area, among others, to see if this is a solvable problem.

-I should have mentioned in my post above, for reference, that the instrument I
mentioned was a classical guitar, pretty much in the Torres style, quite thin top,
more substantial back, Rio rosewood and European spruce (apparently:)).

Normally the 'main ar' resonance on small guitars is higher pitched than on larger ones. Sometimes the sixth string sounds really strong because all of the overtones are strong, and that masks the fact that there's little energy in the fundamental. The ear can do funny things with sound, which is why actual measurements are so nice to have. Do you have recordings of the sound of that guitar that could be analyzed to figure out exactly what the problem is?

Hi all,

I'm new here, and am very glad to find this forum -- so much to learn and a looks like a great group of people!

I've tried some tap testing and am attempting to understand the results. Here's a tap test of the top of a Martin D28, done in a store:



(I used my fingertip, tapping on the bridge as quickly as possible; the higher freq dropoff may be due to the finger method. Soundhole was completely open and then completely covered. Mic was an AT2021 about 10" away from bridge. FFT by WaveSurfer.)

So, it looks like the main air mode (A0) is the 104-107 Hz peak. Not sure why, but for the D28, this peak varied between 104 and 107 Hz from tap to tap (other guitars I've tested are +/- 1 Hz).

Some things I'm wondering about:

Why does the main top mode decrease from 206 Hz to 192 Hz when the soundhole is closed up? I would have thought the top mode would be higher freq with a closed soundhole, since the top sees a stiffer package of air inside. Maybe the top mode gets a coupling effect from the 1st harmonic of the main air mode (2x105 = 210 Hz)?

What are the 380, 581, and 610 Hz peaks?

Is it generally ok to have such a deep valley between the 105 Hz and 206 Hz peaks?

Any other guidance on general interpretation on these kind of graphs, and how to help them guide guitar development?

I've also tapped these guitars, fwiw (last 2 done at home, the rest in a store). (not sure how to do tables here, thus the goofy format.)

Soundhole: _________________Open_______Open________Closed____________Closed
Guitar_______________________A0________Peak________Main_Top_________Main_Back
------------------------------------------------------------------------------------------
Martin_D28_________________104-107______206_________193______ ______189,_247
Martin_DX1___________________93________189__________164_______________272
Taylor_DN5___________________92________190__________170_______________199
Taylor_614ce_________________99_________180__________175_______________209
Takamine_TF360S______________98________180__________169____________no_data
Seagull_Maritime_6_Cedar______104________192__________188________________255
Yamaha_FG-300A_____________103________188________no_data____ _______no_peak

Looks like the Main Top freq drops for every guitar after the soundhole is closed.
I thought it was interesting that the Taylor DN5 had a low A0, given the "Taylor Sound". Maybe they brace for the upper freqs and let the A0 fall where it may?
I found the Back freqs more difficult to measure... sometimes I couldn't find 1 distinct and repeatable peak.

All insights appreciated!

Thanks, David

_________________
David Malicky

-No, I didn't make any recordings at the time- it didn't occur to me. The guitar had
a strong resonance between F and F#, despite its small size- I know that's unusual.
The maker told me while he was making it that "it taps quite low, which *can* be
good". Wish I had more info to share. -Carey

David:
Some good information there. Might I suggest that in the future you use a light, hard 'hammer', and tap on the saddle of the bridge somewhere off center. I use a 5/8" or 3/4" diameter hardwood or plastic bead that is rigged up to swing freely, and have it hity between the 4th and 5th strings. This activates a nice selection of top modes. It makes a sharp 'whack' when it hits, but there's actually very little energy involved, and no chance of damaging anything so long as you hit the saddle.

The 'main air' and 'main top' resonant modes are actually the two halves of a 'bass reflex couple'. There is a lot of top motion at the 'main air' pitch, and lot of air moving through the soundhole at the 'main top' frequency. In any such system of coupled oscillators the pitches will be changed from their 'true' values simply by the fact that they're coupled. One way to think of it is that, at the 'main air' pitch the air is moving out of the soundhole as the top is moving 'inward'. Not only does the air have to move itself in and out of the hole, but has to drag along the top as well, so the pitch of the 'main air' mode is lower than the 'real Helmholtz' pitch of the box with rigid walls would be. At the 'main top' pitch the air movement through the hole is such that the pressure in the box is rising as the top moves inward, so the 'air spring' pushing on the top adds to the restoring force of top stiffness, and raises the pitch. With the hole closed you lose that mass flow, and the pitch of the 'main top' mode is lower.

"What are the 380, 581, and 610 Hz peaks? "

That's a good question. We, as Americans, have the right to ask these questions, and nobody can stop us. It's a right that Americans have fought and died for all over the world, and one that many people don't enjoy. We need to cling to that right, and proudly defend the right of others to ask questions, even when we disagree with them. And blah blah blah....

(In other words, I haven't a clue without some more data)

"Is it generally ok to have such a deep valley between the 105 Hz and 206 Hz peaks? "

Well, it's normal.... I've never seen a guitar that didn't have that dip, so I can't say whether it would sound better or not. It's interesting to think about what you might do to fill it in; I suspect it won't be easy.

"Any other guidance on general interpretation on these kind of graphs, and how to help them guide guitar development? "

Some other sort of 'ground truth', such as Chladni patterns of top and back modes, can really help in figuring out those 'other' peaks. In general, the more you can standardize your test, the more comparable the results will be. The room has quite an effect on the relationship of the peak heights; although it won't alter the pitches of the guitar's output, it can enhance or reduce the heights of peaks, or even eliminate them, or add peaks of it's own. Things like microphone location are important as well; guitars can be surprisingly directional, even at fairly low frequencies, and you can pick up minor resonances strongly, or totally miss major ones, depending on how you place the mic. Remember that mics and computers are not nearly as smart about this stuff as the human ear and brain: they haven't been around as long! Interestingly, my acoustics guru tells me that those little electret mics that are so common are also quite good measurement mics, particularly at the low frequencies we're most interested in. They often have a surprisingly 'flat' response up to about 6kHz.

The most important thing, though, is to just keep making measurements, and save the results. After a while you'll figure out what you'd like to do better, depending on how you think the guitar works, and you'll also begin to see some trends in the guitars tyou like, and the ones you don't like, too. In that light, test everything you can; without a 'standard of badness' you can't know what's good.

Finally, keep in mind that the two most time-consuming activities in the world are Art and Science. In both cases you're doing something that nobody has done before, and that means that you're not going to be very good at it. By the time you do get good, it's time to move on to something new.

Thanks so much, Alan, for your helpful thoughts. I see the hard-ball-on-a-string makes a lot of sense. And I'm guessing this also gives you a repeatable tap by simply swinging the ball from a set distance to the saddle.

Oh! I didn't know that about the 'main air' and 'main top' modes. Thanks for your clear explanation--I can see now why the main top freq is higher with the hole open, due to the the pressure rise adding 'stiffness'.

>We, as Americans, have the right to ask these questions...
You crack me up. Right, I see I could excite the top at these freqs and read the tea leaves. So, I'm working on a signal generator setup, based on some of your prior posts.

I checked the plots for the height of the A-0 peak, relative to my eyeballling a nearby 'baseline'. All of them were in the 10-20 dB range, with the Seagull the highest and the Taylors the lowest. But I see those dBs probably mean little since my room acoustics weren't controlled.

I didn't realize the room could also eliminate peaks, yikes. On mics, I wonder if averaging the signal from 2 or more mics would be a good way to make mic placement less critical? Or maybe that would just add more source for error. Yes, that sounds very useful to test 'bad' guitars, too, and better science.

Thanks again. As I learn more, I learn there is a lot more I don't know. But that makes it interesting.

David

_________________
David Malicky