peculiar 168 hz in tap response

You can ask questions here about Trevor and Gerard's exciting new book on Luthiery.

Moderators: kiwigeo, Jeremy D

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

peculiar 168 hz in tap response

Post by Jim watts » Thu Oct 24, 2013 2:20 pm

Hello all,
A real quick introduction: For starters I'm familiar with many of you at the OLF and some of you may be familiar with me as well. But for the many who aren't I've been building for about 20 years and study as much as I can. I have a very low output of a few instruments each year. I enjoy this and have a day job also. I recently purchased Trevor's fine set of books and am trying to absorb all I can through them. There's a lot there! Anyway enough of that and onto the subject at hand.

Below are three spectrograms of three different guitars. The first two are hybrid dual fan bracing a la Byers and the last one is a 9 fan Fleta using Coutnals plan. My question revolves around this peak at 168-ish Hz. It's there on all three and I'm not sure what it could be. At first I thought it was my tapping hammer (eraser on a stick), so I changed to a finger, still there; a pencil, still there; a different room, still there. It really surprises me it on all three guitars.
I used a Zoom H4 as the microphone. Any thoughts would be greatly appreciated.
I'm happy to answer any questions.
mb tap response 1200X553.jpg
Dual Fan
hr tap response 1200X554.jpg
Dual Fan
fleta tap response 1200X554.jpg
Fleta

User avatar
charangohabsburg
Blackwood
Posts: 1818
Joined: Sat Oct 16, 2010 1:25 am
Location: Switzerland

Re: peculiar 168 hz in tap response

Post by charangohabsburg » Thu Oct 24, 2013 10:19 pm

Hi Jim, welcome here at the ANZLF.

I think a Chladni pattern test would reveal the "secret" straight forward, respectively it would give the necessary hint for going to know the solution. This is the only reason I invested some time and money in a Chladni testing setup. I hate to use it because it is so noisy, but it really helps. Much better than my abilities in reading coffee grounds. ;)
Markus

To be stupid is like to be dead. Oneself will not be aware of it.
It's only the others who suffer.

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Fri Oct 25, 2013 3:17 pm

Thanks Markus, The Chladni test is a good idea for sure. I need to replace my speaker for it though before I can do it.
I was also wondering if the 168-ish is a harmonic of the main air, I suppose I can block off the sound hole easily enough and see if it goes away.

User avatar
kiwigeo
Admin
Posts: 10580
Joined: Sat Sep 29, 2007 5:57 pm
Location: Adelaide, Sth Australia

Re: peculiar 168 hz in tap response

Post by kiwigeo » Mon Oct 28, 2013 9:41 pm

Jim,

I actually had similar thing happening on a Hauser design classical. The guitar was built spanish method. Like on your spectrums I had quite a few extra small peaks in between the air and main top peaks.....amplitude not quite as high as on your plots. The peaks were all weak monopoles when I did some Chladni testing. Unfortunately the spectrum plots got lost during a recent computer OS reload. I thought it might be resonant peaks due to the neck being attached but I tried dampening the neck (clamped guitar by the neck close to 12th fret so it could vibrate) but peaks were still there.

Some questions:

1. How do you have the guitar mounted during tap tests? Are you holding the instruments in playing position?
2. Where are you holding the Mic?

Cheers Martin
Martin

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Tue Oct 29, 2013 6:08 am

Martin,
The guitar was sitting on an A frame stand, strings are on and damped. I'm holding the the guitars by the neck while tapping.
The microphone (zoom H4) was 1 meter out in front of the lower bout just off of center line and about bridge height, same set up for all three guitars.
I repeated the test with a piece of foam stuffed into sound hole and the the peak is still there.
Thanks,
Jim

User avatar
kiwigeo
Admin
Posts: 10580
Joined: Sat Sep 29, 2007 5:57 pm
Location: Adelaide, Sth Australia

Re: peculiar 168 hz in tap response

Post by kiwigeo » Tue Oct 29, 2013 12:42 pm

Jim watts wrote:Martin,
The guitar was sitting on an A frame stand, strings are on and damped. I'm holding the the guitars by the neck while tapping.
The microphone (zoom H4) was 1 meter out in front of the lower bout just off of center line and about bridge height, same set up for all three guitars.
I repeated the test with a piece of foam stuffed into sound hole and the the peak is still there.
Thanks,
Jim
Try holding the guitar in playing position and repeat the tests. I have my Mic hanging from my workshop ceiling so its about 100mm from the sound hole.
Martin

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Tue Oct 29, 2013 1:18 pm

Jim,

You seem to have a whole lot going on in your plots that I don't think is guitar related. So the first question is do you have VA set up correctly? I think some of the later versions (I use version 9 which is quite old now) are prone to loosing their settings, so even though you may have set it up once, it may not necessarily be so now. Worth a check. It is still possible to download version 9, BTW.

User avatar
kiwigeo
Admin
Posts: 10580
Joined: Sat Sep 29, 2007 5:57 pm
Location: Adelaide, Sth Australia

Re: peculiar 168 hz in tap response

Post by kiwigeo » Tue Oct 29, 2013 2:18 pm

Trevor Gore wrote:Jim,

You seem to have a whole lot going on in your plots that I don't think is guitar related. So the first question is do you have VA set up correctly? I think some of the later versions (I use version 9 which is quite old now) are prone to loosing their settings, so even though you may have set it up once, it may not necessarily be so now. Worth a check. It is still possible to download version 9, BTW.
Trevor,

I recall my plots looking similar to Jim's and I had my guitar sitting on foam blocks.....I went to holding the guitar in playing position at your advice and got much better results.
Martin

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Tue Oct 29, 2013 4:59 pm

OK! Maybe that's the issue.

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Thu Oct 31, 2013 12:24 pm

Ok, so I tried this again a couple of different ways.
Concerning Visual Analyser it's labeled 2011 and I have the FFT sample size set to 16384 with a sampling rate of 11025 hz.
This guitar that I'm re-tapping is the first graph shown in my original post.

In this graph the guitar is tapped with the sound hole plugged, the microphone is in the original position (1 meter out in front of the lower bout) and the guitar is on a stand, strings of course are well damped
I believe what is T1,1 dropped to 200Hz and the 168 Hz is still there.
Malaysian blackwood sound hole blocked 1200X554.jpg
This next graph is the guitar tapped in playing position and the mic is about 1/2 meter off of the lower bout. The graphs a little different still.
Malaysian blackwood tapped in playing position 213 peak 1200X553.jpg
Thanks for the input guys.
Jim

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Thu Oct 31, 2013 8:11 pm

Jim, it seems to me that you are still getting spurious results. If you have a stable measurement method and a stable guitar, you should get the first 3 peaks (at least) within ~0.5 Hz whenever you tap. So I think there is something still awry with your measurement method.

Here's a few things to try:

1) Hold the guitar as you would when playing, strings damped, mic about 1m away, pointing at the bridge
2) Make sure you get only one tap per buffer update (you should be able to see the screen refresh each update). Be sure there are no double bounces of the hammer
3) Try a different mic if you have one. A dynamic mic straight into my PC works well for me. I'm not sure what your Zoom stereo mic will be doing, but if you can use just one capsule you might find it better
5) Stay away from anything else that "rings", e.g. other guitars, your bench top, rooms that are particularly live, etc., and things that make noise like your computer fan, air con, etc. and things that induce electrical noise; magnetic fields, motors, fluorescent lights etc.

Your plots should be coming out as "clean" as those shown in the book. Those plots are just as it happened, no post production or anything.

Keep us informed on how you go.

Dreadnuffin'
Gidgee
Posts: 1
Joined: Sat May 07, 2011 10:07 am

Re: peculiar 168 hz in tap response

Post by Dreadnuffin' » Fri Nov 01, 2013 12:24 pm

Well Jim, here's my take on this. I have a 50 Hz peak show up on all of my plots, taken on two different guitars. So I simply took a background (silent) reading of my test equipment/area, (without tapping the guitar) and low and behold there was the 50 Hz peak, large as life. Seems likely that this could just be an artifact of your Zoom H4 mic/computer setup. It'll only take you 30 seconds to find out?

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Fri Nov 01, 2013 3:44 pm

I agree, something is not stable. I can get different reading depending on mic position. I'll see if I can borrow a mic from someone and get my same readings. I did check my background and there a lot of general noise, lights, computer ect... pretty quiet, but they're still there.

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Fri Nov 01, 2013 7:19 pm

Dreadnuffin' wrote:I have a 50 Hz peak show up on all of my plots, taken on two different guitars...
Jim, I think, is in the USA. So he'll be looking at 60Hz (and harmonics) as background electrical noise if it's coming through.

Anecdote:

I have two Toshiba lap tops; an old one and a new one (ha, make that a very old one and an old one). Fortunately, both have the same power requirements and power sockets in the computer. However, the old one has a 2 pin mains lead, the new one a 3 pin mains lead (so has the ground pin there, too). With audio gear I have to use the old power supply with the new computer or I get all sorts of weird digital electrical noise (not just 50 Hz) due, I presume, to a ground loop of some sort forming. If I didn't have the old power supply, I'd probably still be looking for the source of that particular problem.

jeffhigh
Blackwood
Posts: 1536
Joined: Thu Feb 07, 2008 5:50 am
Location: Caves Beach, NSW
Contact:

Re: peculiar 168 hz in tap response

Post by jeffhigh » Sun Nov 03, 2013 9:56 am

How are you damping the strings, Jim?
Not right on the 12th fret location?

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Sun Nov 17, 2013 3:29 pm

All right I'd thought I'd try to update this.
I was able to grab a speaker for checking the Chladni patterns at our local thrift store today. A KEF model 102 reference speaker for $3 !
I'm reattaching the spectrograph from the first post that is from this guitar just to make it easier for people if they're interested in comparing the the spectrograph with the patterns.
When I checked the 167 hz pattern there was nothing there, which I suspected. The confusing thing for me now is all the patterns look the same! Aghhh!
I made recording of the speaker to check the frequency being put out and it's spot on with some background noise about 40 db down.
In order to get the patters I just held the speaker box up over the top, maybe that was a mistake, I donno.
A quick comment about the guitar. It sounds really good, it's responsive, it's loud and a little punchy. A lot of people like it. I'll try to attach a recording too.
Well, here goes;
201 hz1066X800.jpg
235 hz 1066X800.jpg
266 hz 1066X800.jpg
335 hz1066X800.jpg
450 hz1066X800.jpg
Milan.mp3
(1.83 MiB) Downloaded 535 times

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Sun Nov 17, 2013 3:41 pm

This was suppose to be in the above post but I got kicked out of the editing function. Too slow I suppose.
Here's another piece, a modern one this time by my guitar instructor Greg Schneider. Both piece are played by Greg BTW.
Night in the ghetto.mp3
(1.51 MiB) Downloaded 470 times
And I almost forgot the spectrograph
hr tap response 1200X554.jpg

User avatar
kiwigeo
Admin
Posts: 10580
Joined: Sat Sep 29, 2007 5:57 pm
Location: Adelaide, Sth Australia

Re: peculiar 168 hz in tap response

Post by kiwigeo » Sun Nov 17, 2013 9:11 pm

Jim,

The Chladni patterns all look like monopoles except for the last one. Does the instrument have standard fan bracing?
Martin

Jim watts
Blackwood
Posts: 103
Joined: Thu Oct 10, 2013 3:37 pm

Re: peculiar 168 hz in tap response

Post by Jim watts » Mon Nov 18, 2013 10:15 am

Martin, This guitar is based on a Byers dual fan system, a type of lattice.
I wonder if this is part of what accounts for the volume in lattice guitar?

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Tue Nov 19, 2013 8:36 pm

Right, Jim. I still think you have a bunch of weird stuff happening here, so let's see what we can figure out.

First up, a KEF 102 Ref (if you got the whole cabinet) for $3 is a real steal. Add a few zeros for what a working pair of cabs would normally go for!

Looking at your Chladni patterns, I would be saying that you have a T(1,1)2 at 201Hz, a T(1,1)3 at 235Hz, a cross dipole at 266Hz and probably a cross tripole at 335Hz. The mode at 450 Hz could be something like a T(4,2). So why do I say that? Well, I think the T(1,1)2 and T(1,1)3 are self explanatory and in the right frequency range. I don't think the 266Hz is a monopole. The reason I say that is because the outboard node line is pretty close to the edge, suggesting momentum equilibrium within the top, and a cross dipole is in the right frequency range. You may have shaken all the tea leaves off the central node line going through the monopole frequency with the sweep, but if you drag your finger widthwise over the top driving at 266Hz you should be able to feel the node on the centreline, and confirm it by sprinkling some more leaves there. I'd put up a similar argument for the the cross tri at 335Hz. There is nothing in your pics to suggest a long mode in either case.

The nodes of the multi-pole are evident in the lower bout of the 450 Hz mode. I think it has a node line going across the guitar, pretty much through the bridge. If you're really careful, you should be able to get quite a clear pic of it.

The next issue is that the spectrogram looks nothing like I would expect! I've tapped hundreds of guitars and not seen anything like you've got. However, the sound recording suggests a guitar that has the sort of frequency response that you've shown. It sounds somewhat weak in the bass and a bit on the nasal side. Whether it actually sounds like that in real life, only you can tell, but I'm guessing that it doesn't. So, presuming you used the Zoom mic to record both the taps and the playing, the finger of guilt points towards there being something odd about the Zoom mic.

The easiest was to test that is to find another mic, something like a Shure PG 57 or a similar respectable (not necessarily exceptional) dynamic mic. Do some tap tests with that and see what happens. I've used the internal mic on my lap top, external electret mics and external dynamic mics, all plugged directly into my laptop. For whatever reason (impedance matching, automatic gain control or ???) the dynamic mics (of whatever brand) come up with the cleanest and most repeatable plots. Most of the book plots were done with a Sennheiser dynamic mic, which died at age ~30 yrs) but the PG57 I use now produces very similar results, whereas other mics of other types don't, when plugged directly in to my laptop.

So Jim, that's as much as I can tell you and a lot of that is based on some pretty shaky assumptions. Let us know how you get on and maybe we can figure it out.

dshaker
Myrtle
Posts: 77
Joined: Sat Aug 31, 2013 6:38 am
Location: Palo Alto, California

Re: peculiar 168 hz in tap response

Post by dshaker » Sat Dec 14, 2013 11:40 am

When I was trying to debug my spectrum recordings, I found two spurious peaks, one at 60 Hertz,
from my fluorescent lights (US electricity) and another at about 160 Hertz, from my laptop's fan.

The 60 Hertz was reduced, but not eliminated, when I turned out most of my lights.
The 160 Hertz went away when I piled pillows on top of the laptop I was using to make the recording.

I also put my Shure 57 microphone in a pile of pillows with a passageway in the pillows pointing outwards at the guitar.

Doug Shaker
-Doug Shaker

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Sat Dec 14, 2013 7:11 pm

Hello Doug!

Welcome to the underside! :D

dshaker
Myrtle
Posts: 77
Joined: Sat Aug 31, 2013 6:38 am
Location: Palo Alto, California

Re: peculiar 168 hz in tap response

Post by dshaker » Tue Dec 17, 2013 6:09 am

Trevor Gore wrote:Hello Doug!

Welcome to the underside! :D
Well, this is where the action is, isn't it?

Thanks!
-Doug Shaker

User avatar
kiwigeo
Admin
Posts: 10580
Joined: Sat Sep 29, 2007 5:57 pm
Location: Adelaide, Sth Australia

Re: peculiar 168 hz in tap response

Post by kiwigeo » Wed Feb 19, 2014 11:59 am

Ok I just switched to doing my tap testing with VA running native on an HP laptop rather than via virtual Windows on a Mac.

I'd revert back to running VA on the Mac but the new MacAir doesn't have a bldi mic input!!! :evil:

All that has changed in my setup is the computer and the fact Im running VA with Windows 8 on PC rather than Windows 7/Parallels on the Mac.

Its got to be fan noise right? Ive turned off fluro lights and any other electrical device in the workshop except the computer and pre-amp power supply. Ive got both of the latter sitting 5m away from the mic.
Martin

User avatar
Trevor Gore
Blackwood
Posts: 1605
Joined: Mon Jun 20, 2011 8:11 pm

Re: peculiar 168 hz in tap response

Post by Trevor Gore » Wed Feb 19, 2014 1:04 pm

So Martin, what's the problem?

If it's noise that looks weird (i.e. not 50 Hz, fairly random etc.) here's a coupe of things to look at:

1) Digital "Earth loops". These don't look or sound like analogue earth loops. I had this happen when using an M-box with my Toshiba laptop via the USB port. Drove me crackers. The current lap top has a 3 pin power plug on the power supply. I had an old power supply from the previous Tosh. That was two pin. Swapped the two power supplies, which were otherwise compatible. Problem gone.

2) Overloaded input. You might be over-driving the input (one problem) and/or interacting with some automatic gain control, likely embedded somewhere in the computer rather than in the pre-amp.

Neither of these problems looks or sounds like their analogue equivalents.

Post Reply

Who is online

Users browsing this forum: No registered users and 21 guests