Tonewood or not? Testing the myth!

neurotictim · #1 12-07-2010, 04:30 PM

So another thread has brought me to this point, out of curiosity more than anything.

I'd like input from some of you - there are many people here smarter than I, and I could make use of constructive criticism in this science project.

I want to take some of the equipment I have at work, and put it to use in a mostly pointless test on various types of wood, to see how they (and they alone) respond to various frequencies.

My plan is fairly primitive, but I think it's probably better to keep it simple, I just need help on a few points.

I'll need some sort of "control" material (I'm thinking acrylic) and criteria to measure - center frequency, amplitude, side lobe freq and amplitude, what else?

This is the basic idea -

I turn pens as a hobby, so I have lots and lots of "blanks", all (more or less) 1/2" x 1/2" x 4" - and I can trim them to be uniform, within a reasonable standard.

I want to take several different type of wood to work - maple, ash, poplar, wenge, purpleheart, etc - and test them at several different frequencies within the audio spectrum.

I figure I'll have some machinist friends make a couple of BNC-to-metal plate things, so that I can attach them to the wood, and try to conduct a signal through the wood. I'll connect one end of the wood to an audio sig gen, and the other end to a spectrum analyzer and oscilloscope.

I need to figure a way to attach the plates - I thought of clamping them, but I think the clamp would probably hamper the true purpose - it would also conduct sound, and not really give me the "clean" result I'm looking for. Screws, perhaps? I'd have to manufacture a jig that would let me hold the wood securely and transmit the tone through the wood.

I figured I'd try 5 or 6 frequencies, maybe 40Hz, 250Hz, 500Hz, 1kHz, 2.5kHz, and 5kHz, and just for kicks, 10kHz.

Input, please?

FWIW, this won't happen overnight - gotta figure out the jigs, and go from there - but I am willing to hypothesize we will get some real, empirical data from this series of tests that proves certain woods lend themselves to certain tonal characteristics.

BigOldHarry · #2 12-07-2010, 04:35 PM

Great idea...

I'd suggest that to get a sample that might have more meaning for an electric bass to use a longer piece of wood - perhaps thicker too. Heck, the ideal would be to be the distance from a neck joint to a bridge, and about 1 1/4" thick...

But you have the wood you have... that's a start!

I'd like to see a test like this with different bridges, too!

neurotictim · #3 12-07-2010, 04:36 PM

Oh, also - I realize this still won't "prove" that wood types have a serious effect on tone in a bass. There are so many variables that go into each and every bass, it's virtually impossible (or, at the very least, highly unfeasible) to test a large sample of basses with the ONLY difference being body wood.

That being said, whichever way this pans out, it will either prove or disprove that differing woods DO have differing tonal characteristics, and will hopefully (and more importantly) do some good in clarifying exactly what those characteristics are.

neurotictim · #4 12-07-2010, 04:42 PM

Quote:

Originally Posted by BigOldHarry

Great idea...

I'd suggest that to get a sample that might have more meaning for an electric bass to use a longer piece of wood - perhaps thicker too. Heck, the ideal would be to be the distance from a neck joint to a bridge, and about 1 1/4" thick...

But you have the wood you have... that's a start!

I'd like to see a test like this with different bridges, too!

Yeah, as awesome as that would be, the money involved with buying the various species of wood in those sizes would be prohibitive for me.

Like you said, I'll go along with what I've got. Should be able to tell us something.

BigOldHarry · #5 12-07-2010, 04:43 PM

Quote:

Originally Posted by neurotictim

Oh, also - I realize this still won't "prove" that wood types have a serious effect on tone in a bass. There are so many variables that go into each and every bass, it's virtually impossible (or, at the very least, highly unfeasible) to test a large sample of basses with the ONLY difference being body wood.

That being said, whichever way this pans out, it will either prove or disprove that differing woods DO have differing tonal characteristics, and will hopefully (and more importantly) do some good in clarifying exactly what those characteristics are.

I have to think that the bigger companies have done work like this... Heck, if *I* started making basses for a living, I'd put together a jig with this for different bodies/necks - Even strings / bridges. I don't know that the data collected would be all that useful, but it would be interesting.

John D · #6 12-07-2010, 04:43 PM

"I figure I'll have some machinist friends make a couple of BNC-to-metal plate things, so that I can attach them to the wood, and try to conduct a signal through the wood."

This sounds like an interesting project. However, rather than an electrical signal, wouldn't it be better to try to conduct the mechanical vibrations through the wood? Maybe you could mount a piezoelectric transducer on each end of the wood sample.

ggunn · #7 12-07-2010, 06:00 PM

Quote:

Originally Posted by neurotictim

I figure I'll have some machinist friends make a couple of BNC-to-metal plate things, so that I can attach them to the wood, and try to conduct a signal through the wood. I'll connect one end of the wood to an audio sig gen, and the other end to a spectrum analyzer and oscilloscope.

It won't work. You want to test acoustic properties, not electrical properties. I'd start with a consistent hammer tap and a piezo pickup routed to an FFT frequency analyzer. Look for resonance peaks.

Hoover · #8 12-07-2010, 06:09 PM

Quote:

Originally Posted by ggunn

It won't work. You want to test acoustic properties, not electrical properties. I'd start with a consistent hammer tap and a piezo pickup routed to an FFT frequency analyzer. Look for resonance peaks.

Use an accelerometer and impulse response test tones. That's how some speaker manufacturer's document resonance peaks in their cabinets.

Rob22315 · #9 12-07-2010, 06:12 PM

There's no doubt different types of wood have different acoustic properties. We don't need an experiment to establish that. Regarding 'tonewood', I don't think your setup will provide useful insights as you've proposed.

The question is whether or not the varying acoustic properties of wood have a tangible effect in solid body electric bass guitars with electric pickups. (Variations on that theme might include piezoelectric pickups and hollow-body guitars.)

At the extremes,there is a connection between the string and what it's vibrating on - sustain is the biggest example - a string will vibrate forever (according to Les Paul) if it's attached to a train rail. Obviously it will vibrate less as the material to which it's attached allows energy to dissipate from the string. It also follows that if the material's acoustic properties vary with frequency, the material's interaction with the string will also vary. This means it's POSSIBLE for the two to affect one another. Whether or not that interaction has any noticeable effect in a real bass guitar is another matter entirely. Can you set up an experiment to test that?

neurotictim · #10 12-07-2010, 07:00 PM

Hrm... Good points, all. I'm trying to figure out how to do this, and what results to expect. The pointlessness of this experiment has already been pointed out, by the OP - who I know personally.

I just have to figure out exactly what to do to facilitate this test, within my limited scope and equipment. Anyone got any specific, helpful ideas? I don't have an accelerometer or impulse response test tones. I do have a bunch of communications and navigation test equipment at my disposal.

Rob22315 - the points you make I made just a few hours ago in the other thread, which prompted me to do this. The question, the idea of this whole project, is the acoustic differences in wood, and just wood - not in that "we know that..." sense, but with some sort of data to back up that knowledge.

Not whole basses, just the wood - I'll leave that to more educated, experienced people with more motivation.

Rob22315 · #11 12-08-2010, 02:23 AM

Good on you for pressing with your experiment.

Your setup looks good for understanding frequency response but you won't be able to see what happens to the speed of sound through the wood (phase shifts) nor will you be able to see what happens at interfaces between different types of wood. Assuming you want to limit this to just one type of wood, you'll need to connect both the input signal and the output signal to the oscilloscope. I haven't dealt with o-scopes in a while but there's a trigger input that might give you the ability to measure the time delay between the input and output (absolute speed) or see the phase shift as you vary frequency (relative speed). Understanding the speed of sound through the wood might help characterize the natural response of the wood so you can better anticipate where a specific size piece of wood will resonate. It should also help characterize what might happen at the interface between different types of wood. I expect density of the wood and moisture content to play a big factor here. I don't know if it's possible to measure scattering or other effects due to the grain structure of the wood. The wavelength of bass frequencies relative to the size of those features may be large enough that they don't matter, I don't know.

I've seen an increased use of fiber optics as a sensing mechanism, do you have access to any fiber? Simply wrapping the fiber around the wood then looking for phase shifts in the resulting light signal is a very effective sensing mechanism.

Just some musings . . .

kosmicwizard · #12 12-08-2010, 02:44 AM

Lilacs smell purple

Rob22315 · #13 12-08-2010, 02:57 AM

Quote:

Originally Posted by kosmicwizard

Lilacs smell purple

Carrots smell orange?

neurotictim · #14 12-08-2010, 04:09 AM

No fiber optics - that stuff is expensive, and the Navy wouldn't let me play with it for an science experiment anyway.

So, let's think of what I should measure. It occurred to me last night that I should have a set idea of what I want to measure, and then define how that data would change tone.

Obviously, frequency response is on the list. You think the speed of sound through the wood is important? I think I can manage to measure input vs. output on that point, looking for phase shifts. I would imagine that attenuation (in pure terms of vRMS) would be important, as well - it would correlate with frequency response, of course.

I would probably have to increase amplitude as I went lower in frequency, wouldn't I?

Still thinking on this...

Jim Nazium · #15 12-08-2010, 04:34 AM

You know, there was a site from a bass builder - The Groove Shoppe I think - that had recordings of three different basses that were identical except for the woods. Unfortunately I can no longer find the link; maybe someone else has it. One bass was ash/maple, one was alder/rosewood, and one was ?/ebony. All three had the same pickups, strings, and electronics, and the same bass line was recorded on the same equipment with all three instruments. The difference in tone was immediately apparent even on my crappy computer speakers.

Don't get me wrong, I'm all in favor of using science to bust myths, and there are a lot of them surrounding music equipment. But to me there is no question that wood has an audible effect.

T-Bird · #16 12-08-2010, 07:19 AM

Hi.

Didn't read all the replies carefully, but essentially what You're trying to do will work. But as several posters have stated, not without a some sort of a "speaker" & "pickup".

Also, You won't find what You're looking for even though you'll find quite a few differences between the test pieces.

If I was interested on the so called "tonewoods" a some sort of tap method, as John D suggested, would be a good signal transferring method. A piezo element won't work well unless it's stressed, which in turn is very hard to repeat since the wood itself shouldn't be under a load.

Like mentioned, knowing the phase difference is pretty important, and taking a scope reading from both send and recieving ends takes care of that problem.

If You just want to fiddle around, a piezo disk on both ends of the stick, clamped together with a C-clamp, will get You readings. The easiest reference material would be steel, easy availability and magnitudes more stable than wood.

Good luck.

Regards
Sam

i_got_a_mohawk · #17 12-08-2010, 07:53 AM

Quote:

Originally Posted by Jim Nazium

You know, there was a site from a bass builder - The Groove Shoppe I think - that had recordings of three different basses that were identical except for the woods. Unfortunately I can no longer find the link; maybe someone else has it. One bass was ash/maple, one was alder/rosewood, and one was ?/ebony. All three had the same pickups, strings, and electronics, and the same bass line was recorded on the same equipment with all three instruments. The difference in tone was immediately apparent even on my crappy computer speakers.

There are two (with a possible 3rd) problems that are immediately apparent with that.

1) The player, I know this isn't really something one can replicate, but as we all know, the effect of the tone from the way the bass is played has a massive impact. Ideally you'd want to remove the player as a variable (robo-bassist?).

2) Repeats. The characteristics, the difference in tones could have been characterised by individual properties of that lump of wood. For isntance, they could have recorded two ash bodied basses and found the difference to be equivical to that heard between the ash and alder bass.

3) This is only a possible problem, but I've raised it before in these discussions. The electronics generally used in musical instruments have rather wide tolerance ranges. If the same electronics were taken out and plugged into each body, that would eliminate this issue, but it's difficult to confirm the electronics as being identical, just because they are from the same manufacturer (ie most pots are +/- 20%, common ceramic caps can have tolerances from -20% to +80%).

Granted more accurate electronics are avalible, but it depends on if they were used (ie Mica caps generally have a tolerance of between +/-5% to +/-1%, depending on manufacturer), or if the same harness was moved from one bass to the other, as said, that's why #3 is more of a possible (because I'm not familiar with how it was set up).

Just food for thought

(I do think wood has an effect, IMO more related to density and the sustain you'll get, but I do think that technique, strings and electronics have a far more substancial effect)

**Coolhandjjl** · #18 12-08-2010, 07:57 AM

You'd probably want a waterfall plot instead of a plan spectrum analyzer frequency plot. Most RTA software can do this anyway.

Thing about the graphs, yes, they will all be different, but what constitutes 'good' sound vs 'poor' sound? Some folks prefer maple. Some prefer swamp ash., etc, etc. It's almost as if the graphs will be meaningless.

i_got_a_mohawk · #19 12-08-2010, 07:59 AM

Quote:

Originally Posted by Coolhandjjl

You'd probably want a waterfall plot instead of a plan spectrum analyzer frequency plot. Most RTA software can do this anyway.

Thing about the graphs, yes, they will all be different, but what constitutes 'good' sound vs 'poor' sound? Some folks prefer maple. Some prefer swamp ash., etc, etc. It's almost as if the graphs will be meaningless.

I don't think he is talking about using this experiment for good vs bad, but to try and quantitatively show there is a difference.

lomo · #20 12-08-2010, 08:05 AM

The simplest way is also the most conclusive. Take 4 identical basses differing only in body wood. Put the same new strings on all of them and have a blindfolded believer listen to a blindfolded player pick the same simple riff on all 4 basses through the flattest rig you can gather, in random order, for 100 samples (25 from each bass). You will be able to accurately test whether the subject can differentiate, without any rational reason to doubt the result. There will be irrational doubts, of course :-). You also need 3 or 4 people and a willing, believing test subject.