Where do I go from here? (Spectrum Analysis)

[chrisbaer]

Hello everyone! I just finished construction of a Spalted Maple Jumbo with X-bracing and a live back. It's my second build, so I'm kind of fumbling through at this point. I was fortunate enough to take a class with Trevor in Denver so I've gotten the bug to try and add some Gore-like touches (live back, side weights, spectrum analysis). I had the top panel done before the class, so I asked Trevor if he had any advice on building a Jumbo. He said, "Don't. There are too many big guitars already!" So, I'm already in the hole here....

Having done the spectrum analysis (I'm using Audacity and exporting the Spectrum Analysis text file to Numbers), I'm at a bit of a loss how to proceed. It seems to me that there's a lot more going on than I expected! Looking at these two graphs (one with 350g weights and one without), how would you start to analyze the results? What jumps out?

Thanks so much for any assistance!

Chris Baer

[kiwigeo]

My thoughts pending Trevor chiming in with his learned words.

Before doing any testing with side mass added you need to work out what your targets are for the two important top peaks - the Air resonance peak and the main top peak. For a "large" guitar The Books recommend targets of 90/160Hz, 90/170 or 95/170Hz.

My interpretation of the two plots:
1. Air resonance peak hasn't changed at all (84. with added weight. You're already below 90Hz which is the lowest I go with any of my steel strings.
2. Main top frequency has dropped from 175 to 174.6Hz...not a huge drop. If 170Hz is your target then it should be attainable with added side mass but keep in mind your air resonance peak will also drop if you drop top resonance peak. How are you attaching the weights to the sides of your guitar? Unless they're securely bolted in they won't work that well to drop top frequency. How much of a drop you get also depends on the responsiveness of the top.

[Trevor Gore]

Hello Chris,

First thing to do is to figure out what's what.

I'll start by assuming that you have the spectrum analyzer and signal levels set up correctly.

It is pretty certain that the 84.8Hz peak is the main air resonance, the T(1,1)1. Being that low suggests that both the top and the back are pretty "soft". A T(1,1)2 at ~175Hz is not particularly soft, which sort of hints that the ~150Hz peak may be the T(1,1)2 (main top), especially as it responds to added side mass, but it could also be a very soft back giving a T(1, 1)3 lower in frequency than the T(1,1)2, hinted at by the amplitude being lower. Hence the first thing to do is to figure out which peak really is the T(1,1)2 and what the other peak in that area is. The best way of doing that is to use Chladni patterns. If you don't have a Chladni kit available (a DIY one here), just take a tap response of the back, tapping only in the centre of the back's lower bout. Take the measurement in the same way as you would the top, but with the back facing the mic. Don't plug the sound hole or anything. The main peak you see in the frequency response when you tap the back like that is usually a little lower in frequency than the T(1,1)3, which in normal circumstances (which we may not have here!) should allow you to identify the T(1,1)3. Have a go at that and let us know what you find. A few more details about the guitar would also help; e.g. top and back thickness, bracing styles for top and back, nature of any scalloping etc..

[chrisbaer]

Hi Trevor, so great of you to respond, thanks so much!

I don't have a Chladni rig built, but I'm looking into that. Attached is the back tapping that I did. Also some pictures of the bracing. The top is standard X bracing, but the lower braces are short of the sides (not glued in). Looks like I don't have a good picture of the top.

Top thickness is roughly 2.85mm, back thickness about 2.9mm. Sadly, I did not pretest my top or back, so I didn't really have any target thickness... just winging it!

So, this may just be a Frankenstein Jumbo... It may not really be possible to analyze satisfactorily! Any attempts are appreciated.

Chris

[chrisbaer]

Hi Martin, thanks so much for your response! The side weights are attached per Gore/Gilet - I think they're pretty solid. I'm guessing my top is not very responsive - it's a standard X-brace and if I had to guess, I'd say it's a little over-built. I didn't lock in any of the lower braces, so that might help.

Ultimately, I think it sounds pretty good! I'm just hoping to move slowly down the path towards a more open and responsive instrument! Maybe guitar #4 will have falcate bracing... we shall see!

Chris

My thoughts pending Trevor chiming in with his learned words.

Before doing any testing with side mass added you need to work out what your targets are for the two important top peaks - the Air resonance peak and the main top peak. For a "large" guitar The Books recommend targets of 90/160Hz, 90/170 or 95/170Hz.

My interpretation of the two plots:
1. Air resonance peak hasn't changed at all (84. with added weight. You're already below 90Hz which is the lowest I go with any of my steel strings.
2. Main top frequency has dropped from 175 to 174.6Hz...not a huge drop. If 170Hz is your target then it should be attainable with added side mass but keep in mind your air resonance peak will also drop if you drop top resonance peak. How are you attaching the weights to the sides of your guitar? Unless they're securely bolted in they won't work that well to drop top frequency. How much of a drop you get also depends on the responsiveness of the top.

[Trevor Gore]

Chris, that seems to indicate that the back is tuned lower than the top (usually not a good idea) as the 154Hz peak has increased in amplitude for the back tap. However, I'm still suspicious! A back tap normally looks like this:

...sound hole NOT plugged, pretty much a singular peak, with the residue of an air resonance at ~100Hz. This suggests that your spectrum analyser might not be set up right. The quickest way to test that is to tap another guitar's top and see if it shows a more normal response that you can replicate consistently.

I've never liked guitars with backs tuned lower than the top. I found that out before I got into all the spectrum analysis work, so I don't have any data on how they behave with added side mass, etc. . The pertinent question, though, is do you like how it sounds? If so, leave it as it is. If not, the first thing I'd be doing would be to stiffen the back, if you can confirm that the T(1,1)3 really is at 154Hz.

[kiwigeo]

Ultimately, I think it sounds pretty good! I'm just hoping to move slowly down the path towards a more open and responsive instrument! Maybe guitar #4 will have falcate bracing... we shall see!

Chris

If you're happy with the sound then there's nothing more to do!!