Does the tap tone mean anything?

[koolimy]

Hi, I'm in the process of building an Engelmann/Tamarind Hauser style classical guitar from a kit from Blues Creek Guitars. I have finished the box and I'm in the process of cutting the binding channels. Yesterday I was changing the strings on my steel string guitar and had an opportunity to compare the tap tones of the two guitars. To my surprise, they were very different!

The steel string guitar, an Adirondack/Cherry J-185, had a very solid, low pitched tap tone. The classical guitar, on the other hand, had a mid pitched tap w/ a very hollow, airy sound. Of course, the J-185 has a bridge & finish on it, while the classical guitar does not.

Just out of curiosity, I was wondering if the tap tones had any meaning for you guys when you were building your guitars. I know that the pitch of the tap tone is relevant b/c you want the back and top to be close to each other in pitch to have a live back. But outside of that, did the characteristics of the tap translate to the final sound? It was just very surprising to hear such a stark difference in tap tones between the two guitars. Of course, the bridge and finish probably account for a lot of the sound so it's foolish to make any conclusions. Nonetheless, was just wondering if the tap tone will have an influence on the final sound. Would be interested in your comments!

[rick-slo]

Just curious. What's the Helmholtz frequency for each guitar?

[koolimy]

Originally Posted by rick-slo

Just curious. What's the Helmholtz frequency for each guitar?

Hi, I don't have measurements for my classical, but the last measurements I made for my J-185 were as follows:

Air resonance: 92Hz
Top resonance: 178Hz
Back Resonance: 234Hz

I'm guessing the classical is a bit higher in pitch for everything, although the classical has the top and back close together (just normal hearing, no measuring).

[Alan Carruth]

Usually the pitch you perceive when you tap the top is the 'main air' resonance: the Helmholtz mode as modified by the mass/compliance of the top and back. The higher pitched 'main top' mode is generally masked; subsumed as an overtone of the 'air' pitch. The usual way to isolate them is to cover the sound hole, which takes the Helmholtz resonance out of the mix and at least partly decouples the top and back.

The low range (say, up to 600-800 Hz?) tap tones seem to determine the overall 'character' of the sound of the guitar. The relative pitches and strengths vary in more or less systematic ways between, say small 'parlor' guitars and larger ones, and there can also be differences between large Dreadnought' and 'Jumbo' shapes that help us to pick up on those differences. 'Quality' differences, between ' very good' and 'average' guitars of the same general size and configuration seem to be more determined by what happens in the higher frequency range; particularly between 2000-4000 Hz.

A few years ago, as part of an ongoing effort, I made a couple of mahogany/ spruce OMs that were as similar as I could make them. Up to 1000 Hz the spectra were virtually 'identical'; well within the limits of measurement error. They were good instruments of their kind: with all the characteristics you'd expect of the breed, and nobody had any particular preference for one over the other. However, in 'blind' tests literally everybody could tell them apart. The high end spectra started to diverge above 1000 Hz, and were quite different in some way between 2-4 kHz. These differences could largely be tracked to small variations in local stiffness and density in the tops, which, despite being cut sequentially from the same plank, showed minor variations in grain.

[koolimy]

Originally Posted by Alan Carruth

Usually the pitch you perceive when you tap the top is the 'main air' resonance: the Helmholtz mode as modified by the mass/compliance of the top and back. The higher pitched 'main top' mode is generally masked; subsumed as an overtone of the 'air' pitch. The usual way to isolate them is to cover the sound hole, which takes the Helmholtz resonance out of the mix and at least partly decouples the top and back.

The low range (say, up to 600-800 Hz?) tap tones seem to determine the overall 'character' of the sound of the guitar. The relative pitches and strengths vary in more or less systematic ways between, say small 'parlor' guitars and larger ones, and there can also be differences between large Dreadnought' and 'Jumbo' shapes that help us to pick up on those differences. 'Quality' differences, between ' very good' and 'average' guitars of the same general size and configuration seem to be more determined by what happens in the higher frequency range; particularly between 2000-4000 Hz.

A few years ago, as part of an ongoing effort, I made a couple of mahogany/ spruce OMs that were as similar as I could make them. Up to 1000 Hz the spectra were virtually 'identical'; well within the limits of measurement error. They were good instruments of their kind: with all the characteristics you'd expect of the breed, and nobody had any particular preference for one over the other. However, in 'blind' tests literally everybody could tell them apart. The high end spectra started to diverge above 1000 Hz, and were quite different in some way between 2-4 kHz. These differences could largely be tracked to small variations in local stiffness and density in the tops, which, despite being cut sequentially from the same plank, showed minor variations in grain.

Thank you very much for your reply! I am wondering regarding your experiment, is the 1000Hz range still in the range of normal guitar sounds, or is it in the range of overtones? If it's in the range of overtones, does that mean a lot of the differences b/w guitars come in the differences between overtones?

My tapping experiment was only done by ear, but it was interesting how different the tap tones could sound. The classical guitar had a lot more air/hollowness/reverb in the sound. I'm not sure what to make of it. Maybe it means that the higher pitched modes are more prominent? Nonetheless it was fun to observe. Hopefully it doesn't mean that there's a defect in my guitar, LOL.

[Alan Carruth]

koolimy asked:
"is the 1000Hz range still in the range of normal guitar sounds, or is it in the range of overtones?"

The fundamental pitch of the low E string is 82.4 Hz, and the 12th fret on the high E is 659.2 Hz, so 1000 Hz is up around C, 20th fret on the high E. The 'sound' of the guitar depends on what frequencies the player feeds in, and how the guitar responds to them. When you pluck a string you're driving the fundamental of the pitch and a whole bunch of overtones, out to 10,000 Hz and above, with the mix depending on where you pluck the string, mostly. The guitar takes some of that string vibration and turns it into sound. How effectively it does that depends on a lot of things, including how it's designed and built, as well as how you played the string.

The construction and material of the guitar determine the resonant pitches of the different parts, and how they will work together. Strings are simple structures and it's fairly simple to figure out the fundamental and overtones. Guitars are more complicated. At the lowest frequencies, down on the low E string and upwards to around the open G string pitch, it's possible to predict (more or less) what the resonances of a guitar will be. As you go higher there are more resonances, and they influence each other more strongly, so it gets more and more difficult to predict what's going to happen. Somewhere up around 600-800 Hz you get into a 'resonance continuum', where there is so much going on it's impossible to control it in advance. Small variations in things like wood properties, that you can't control, affect the sound in noticeable ways. So that 1000 Hz cutoff is not really arbitrary; it's simply the point beyond which the maker doesn't have detailed control any more. Guitars that behave the same way below 1000 Hz have the same 'character' of sound, but can still be easily distinguished from each other in 'blind' tests.

One of my customers designs satellites. He has software that he can use to look at the way they vibrate, to be sure nothing shakes off during the launch. He once tried entering the specs of a guitar into his software, but it was too complicated. This is harder than rocket science...

[H165]

It seems to make a difference to me, but not measured, just anecdotally.

I'm curious - Is anyone else here using ordinary clip-on tuners to try to measure random tapping results?

[koolimy]

Originally Posted by Alan Carruth

koolimy asked:
"is the 1000Hz range still in the range of normal guitar sounds, or is it in the range of overtones?"

The fundamental pitch of the low E string is 82.4 Hz, and the 12th fret on the high E is 659.2 Hz, so 1000 Hz is up around C, 20th fret on the high E. The 'sound' of the guitar depends on what frequencies the player feeds in, and how the guitar responds to them. When you pluck a string you're driving the fundamental of the pitch and a whole bunch of overtones, out to 10,000 Hz and above, with the mix depending on where you pluck the string, mostly. The guitar takes some of that string vibration and turns it into sound. How effectively it does that depends on a lot of things, including how it's designed and built, as well as how you played the string.

The construction and material of the guitar determine the resonant pitches of the different parts, and how they will work together. Strings are simple structures and it's fairly simple to figure out the fundamental and overtones. Guitars are more complicated. At the lowest frequencies, down on the low E string and upwards to around the open G string pitch, it's possible to predict (more or less) what the resonances of a guitar will be. As you go higher there are more resonances, and they influence each other more strongly, so it gets more and more difficult to predict what's going to happen. Somewhere up around 600-800 Hz you get into a 'resonance continuum', where there is so much going on it's impossible to control it in advance. Small variations in things like wood properties, that you can't control, affect the sound in noticeable ways. So that 1000 Hz cutoff is not really arbitrary; it's simply the point beyond which the maker doesn't have detailed control any more. Guitars that behave the same way below 1000 Hz have the same 'character' of sound, but can still be easily distinguished from each other in 'blind' tests.

One of my customers designs satellites. He has software that he can use to look at the way they vibrate, to be sure nothing shakes off during the launch. He once tried entering the specs of a guitar into his software, but it was too complicated. This is harder than rocket science...

This is quite fascinating! So basically we are up to the mercy of the wood for a large part of the guitar's sound . I read online that top selection is one of the most important parts of making the guitar, and I can see why.

One of the things I most appreciate after I took up this hobby is learning about the actual stuff that makes a guitar tick. And you have probably been one of the biggest sources of knowledge! I really appreciate your willingness to teach luthiers, hobbyists, and consumers, stuff that usually amounts to trade secrets in other professions. At least for me, you have expanded my understanding of the acoustic guitar greatly. So thank you.

[koolimy]

Originally Posted by H165

It seems to make a difference to me, but not measured, just anecdotally.

I'm curious - Is anyone else here using ordinary clip-on tuners to try to measure random tapping results?

Anecdotes are the start of data! I was actually looking for anecdotal experiences, LOL. I guess that our ears and fingers are data gathering devices, even though the device that interprets the data might be imperfect. Whatever we hear from the taps must be telling us something about the wood and construction, which should translate to the final sound.

I don't have a clip on tuner, unfortunately, so I can't say anything about using it to measure tapping results.

[Alan Carruth]

koolimy wrote:
" I really appreciate your willingness to teach luthiers, hobbyists, and consumers, stuff that usually amounts to trade secrets in other professions. "

It's been said that what passes for normal conversation at a luthier's meeting would be 'industrial espionage' in any other business. Older luthiers like me remember when nobody knew very much, and the only way to learn was to pool what we did know (or thought we knew). Hence organizations like the Catgut Acoustical Society, the Guild of American Luthiers, and the Associated String Instrument Artisans. Thanks to organizations like these we've gone from being a global backwater to setting standards in many respects. It's a hard habit to get out of.

[redir]

Originally Posted by H165

It seems to make a difference to me, but not measured, just anecdotally.

I'm curious - Is anyone else here using ordinary clip-on tuners to try to measure random tapping results?

You need a spectrum analyzer. Something that will record and freeze the graph so you can look at it. The clip on is good enough for getting the natural frequency of the guitar by humming into the sound hole till you hit that pitch. But for taps which are quick and punctuated you need to be more precise.

[Alan Carruth]

Thanks to the computer revolution spectrum analysis is cheap and easy to do these days. There are any number of programs that run on all sorts of platforms that can plot out the spectrum of a tap easily. One I can recommend is 'Luthier Lab', written for Android devices by a friend of mine and his wife. He's an amateur luthier, so that app has all sorts of 'builder' tools, including a tone generator and a spectrum analysis module. Best of all, it's free.

[redir]

Yup that is what I use and it works very well, the Luthiers Lab. I got a little external mic for my Android phone to try and make it even better but the native mic on the phone works good enough too.

[viento]

Is there any similiar app for Apple ?