Official Luthiers Forum!

Hey all,
Can someone please give me an easy way to determine the fundamental vibrating frequency of the top and back of a completed guitar? Tapping isn't doing it for me and the strobe won't pick it up fast enough.
I do have a signal generator, amp and speaker but preliminary tries with glitter failed to help.



I'm excited to try this and have some time on Saturday and hope someone can point me in a direction to find this info.

thanks,

Dave

why aren't taps working? are you doing an fft on the taps?

_________________
Jim Watts
http://jameswattsguitars.com

The taps aren't working for me because all I hear is the main resonance of the closed box.

I fyou record the taps and use a program like audacity (freeware) and perform a fft (fast fourier transform) on them it will break the sound down to its frequency components, and yes the lowest will be the air resonance. you probably want to damp out the back when you tap the top and damp the top when you tap the back.

_________________
Jim Watts
http://jameswattsguitars.com

Thanks Jim. I'm familiar with the fft, but how will I know the difference between the main air mode, it's natural overtones and the fundamental hz of the top?

Covering the soundhole will quiet the air modes, so you can see the main top freq clearly (typically between 160 and 200 Hz with closed hole). Then try opening the soundhole and tap again; the main top freq will shift higher by 5 to 20 Hz. Then you can pick out the main top freq from the spectrum with the soundhole open.

Same story for the main back freq, except I often find 3 back freqs closely spaced (e.g., 210, 225, and 250 Hz), depending where tapped. I find the lowest by tapping in the middle of the lower bout between the #3 and #4 braces. Tapping off-center or closer to the waist usually emphasizes the higher 2 freqs. Tapping directly over a brace will create more resonance and also emphasize the higher freqs.

Mic placement also helps differentiate: placing near the lower bout usually gives a stronger main top than main air. Tapping on a bridge wing with the mic off-center can find the cross-dipole (but this isn't very reliable just using taps, and the cross dp is near the long dp--this doesn't show which is which.) A live/real-time fft (e.g., wavesurfer) can be useful to see how the cross dp fades in and out by placing the mic off or on center... probably.

_________________
David Malicky

Hum over the top or back, placing your finger lightly on the surface to sense the vibration. Vary the note until you feel the strongest sympathetic vibration. This is a good way to train your ear, after which you should be able to hear it just by tapping.
.

_________________
John

Like David said the main air air mode is down around in the hundred range and the top fundamental is typically in the 200 range. I use the term range loosely.

_________________
Jim Watts
http://jameswattsguitars.com

Dave, I also had a hard time picking up the top and back main resonant mode when tapping with my finger. I made a hammer of sorts, using a very dense piece of foam for the head and just a top cut off for the handle. For the main top mode, I tap on the saddle in the center with the soundhole plugged and the back damped by it laying on a towel. I use a strobe tuner and can easily get consistent frequencies. For the back main mode, I lay the guitar top down with soundhole plugged on a towel and tap in the center of the lower bout.

Chuck

_________________
_________________

A piano hammer works real nice for tapping, I just added a longer and larger dowel for a handle instead of the short dowel on it. I find that Peterson Strobosoft works real well it the sound hole is blocked with some masking tape.

Fred

_________________
Fred Tellier
http://www.fetellierguitars.com
Facebook page http://www.facebook.com/pages/FE-Tellier-Guitars/163451547003866

This is very timely for me.... just started fooling with Audacity on a recent body to try & get a better grip on things.
FWIW, this is a shot of (clockwise) main air, main top & main back (before & after reworking to get it closer to the top). Very interesting learning experience re adjusting carved plates and various mode effects. BTW for a hammer, I use a white "drafting" eraser with a 3/16" dowel through it.

Feeling my way, I'm using spectrum algorithm, hanning window function, and log frequency axis. Any advice from the more experienced would be appreciated! (eg What means this "fast fourier transform" and how do I find/create/use it??)

_________________
Dave
Milton, ON

I spent the day out in the shop testing and tapping and humming and blowing and huffing and puffing, plugging and opening the soundhole...
Eventually I found a method that worked.

Using a piece of software like Audacity, I ran a signal into the laptop from a mixer and a condensor mic. Like Dave Stewart, I found the Hamming view most useful.

Through trial and error it was discovered that damping the other side of the guitar is crucial to getting the right readings. By doing so a fundamental, like those in the previous post by Dave Steward showed up.

My goal was to first explore and collect data and then see what results I could acheive through manipulating the braces.
I found that my favorite guitar had a top frequency of a flat F# and the back was at an in tune G. That is close to the range reccomended by many who talk about voicing their instruments.

My least favorite guitar had a top at G and a back at a sharp B!
By reaching in the shound hole and shaving with a finger plane I was able to get that back down to a half step away from the top.
It now has a different tone and is downright acceptable to listen to.

Most interesting to me was that by loosening up the back braces, I also dropped the pitch of the top at the same time. While not as significant, without touching the top, it's pitch dropped 9 hz! This was a reminder to me of just how much the closed box functions as a whole unit. Affecting one area of it, does so to the rest.

Next time, I'll test both the top and back as I voice the other parts.

And the next time I build a new instrument I'll be sure to do this testing BEFORE I close the box up so I don't get these nasty bruises on by forearms.

Dave

Dave,

Glad you found what you were after.

Top, back, and main air resonant frequencies all seem to affect one another, so it's not uncommon to see one plate affecting the other. If you had checked the main air before shaving the back braces, and compared it to after, you might have seen a change there too.

Damping the plate not being tested would be required to get any kind of accuracy—if you can call any of this stuff accurate. Seems like one affects the other, which affects the other, which affects........

The half-step difference espoused by Alan Carruth works for me too. He prefers to have the top lower than the back. The top often drops over time, so if you start with the top higher, it might drop down to the frequency of the back, and then you have both plates responding strongly at the same frequency, possibly causing an uneven response, or wolf note. Carruth starts with both plates at the same resonant frequency before the bridge is glued on, then the top often drops a half-step when the bridge goes on, owing to the increase in mass. That's been my experience too.

If you do the work before gluing the plates onto the sides, you might get closer, but once the plates are glued on, they change and you still might have to do some brace shaving, though maybe not as much.

Pat

_________________
formerly known around here as burbank
_________________

http://www.patfosterguitars.com

Glad you got that worked out Dave.

_________________
Jim Watts
http://jameswattsguitars.com

Thanks for the help Jim!

I did another guitar today and had similar results.
Where fine tone was experienced before, I now have another guitar I can't put down. Sounds great, projects great. Even gave me an excuse to put it on the bench and deal with the year-old action. Now it REALLY is something.

The two instruments were a side/side build I did a couple years ago. One Sitka/Sapele, the other Redwood/Walnut.
I had meticulously measured and recorded every detail on the builds. Deflection testing, weighing etc... I was hoping to see what the wood choices would make for difference in tone.

Problem was, neither turned out sounding all that great.

Now they sound Fabulous, and I have a basis of comparison for the differences in wood.

Dave

Dave, I think this relates well to a series of post by Todd Lunneborg (sp? and maybe not the right author, apologies in advance) on voicing the back and the way it made the guitar come alive. It's probably worth a re-read now that you've gone through it.

Dave Stewart- The fft transforms signals from the time domain to the frequency domain, basically it's 1/time, for example if a wave has a time of .02 seconds peak to peak it would be a 50 Hz signal. The speed of computers allows us to look at very complicated signals and break them down into their components. I believe this is basically the key to the digiatl age we live in. What we need is an electrical engineer here to come and set us both straight. As far as doing it, you've done it! This what the analyze function in Audacity does.

_________________
Jim Watts
http://jameswattsguitars.com

Good enough for me! Thanks Jim.

_________________
Dave
Milton, ON