Official Luthiers Forum!

Hello,

I just got strings on another one the other day. The sound seems to be pretty nice, but gets a little wolfie on the "A" string around the F# area (9th fret). It gets better when move in either direction. The wolf tone is not just isolated at one fret, it has a range of 2 or 3 frets. It also gets a little wolfie around the 4th fret on the "D" string. I posted the frequency response below. This is a completed guitar, with the sound hole open, guitar in playing position, with the strings dampened with a rag, and the body size is an OM. The sides have mass, but I am not quit sure how much.

I think the plot is
Air: 100 Hz
Top: 180 Hz
Back: 217-222 Hz? (there are two little peaks there)

Thank you,
Rob

Does it tell you exactly what the peak is, or do you 'think' it is 180? Cause if you thought it was 180 but it was really 185, an f#, that would go a long way towards explaining your f# problem. A mere 5hz...

And you need to find out what your back monopole actually is as well. 217 and 222 are right either side of A 220.

Does the Audacity software give an exact freq with cents, or are you left guessing where the image lies between graph lines on the bottom?

The program locks onto the peaks. The peaks that it locked onto are the ones that I gave. I I do however feel that back of the guitar moving more when playing in those wolf note areas. I suppose that I could either try to stiffen up the back or loosen up the back. I can't really stiffen things up since the back is lattice braced and would be a serious pain. If I were to loosen up the back would that shift my the wolf notes towards the lower end of the neck or would it reduce the interaction between the top and back at those notes? As far as the numbers that I am getting am I in the ballpark at where I should be?

I just another plot and I got
Air:100 Hz
Top:178Hz
Back:211 Hz and 222 Hz... there are two peaks right here. This time they were a little further apart from each other. The peaks are of equal height as well.

Can't help past saying the numbers look about right. Really, you want about 215 for the back. Sometimes backs don't like to play nice in their readings.

You might try generating a plot while you have the back of the guitar facing the microphone. Granted, thats not
playing position, but I find it helps nail down the back frequencies because they are more dominant since they
are closer to the mic. I do both front and back plots on most guitars to try and understand the resonances
better.

_________________
Gene

Politicians and diapers must be changed often, and for the same reason- Mark Twain

Very good point.

With my admittedly limited understanding of this, I would try adding washers in a stack to the inside lower bout of the back (held in place with a magnet on the outside) to try and incrementally lower the reading to the 213/214 range.

I redid the plot with the guitar on my knee, back facing the microphone, strings dampened, tapping on the bridge, and sound hole open. This is what I got, Air: 100 Hz, Top 179 Hz, and Back 217 Hz. It seems that the back is a little high. I am going to try to add some tacky tape to the back to see if I can get it to drop some. Aside of reducing or eliminating the slight wolf tone, will this have any other impact? Will it make it louder?(not to imply that it isnt loud or doesnt sound nice)

Thanks again for help

I am by no means an expert. I follow Trevor's of guidance of having the back 4 semitones higher than the top. By that standard 217hz is not too high really a little low but it would be hard to raise the frequency of the back. On the first diagram It looks like the cross dipole may be right around 222 as it looks like you are getting two peaks 217 and 222. They look to be combining and putting a strong peak centered on 220. Moving the back down a touch might help though and separate the two peaks. Although I was surprised looking at the graph that it responds to F# as a wolf note.

_________________
http://www.Harvestmoonguitars.com

For some reason I have never gotten good readings with audacity. I have cross referenced readings with audacity against readings taken with Adobe audition and spectra plus, which both agreed with each other, and I could not get the same reading with audacity.
I have no explanation why, but that is why I do not use audacity. If you could use another fft to verify your reading it would confirm your data.

I also do not trust Audacity. I think part of the problem is that it forces you, as I remember, to use the full scale for 1 to 20K Hz. Most of that range you are not interested in. I feel better with a program that allows you to see more detail in the range of interest.

_________________
_____________

EddieLee