"Effect of back wood choice on the perceived quality of steel-string acoustic guitars"

[Larry Mal]

This is from the The Journal of the Acoustical Society of America.

Haven't quite parsed it all or anything, so I'm not really discussing the findings made here in anything except the most general terms, but according to this study, back and side woods of acoustic guitars makes a negligible difference.

I thought some folks here might find it interesting.

[Embenny]

Not a shock. Over on the Acoustic Guitar Forum, they have a number of very skilled luthiers who will tell you that construction method and soundboard wood are the primary determinants of an acoustic's tone.

There are a number of experiments people have conducted. David Weber, a master luthier here in Canada nearing the end of his career, built a guitar using particle board for the backs and sides and firewood-grade mahogany for the neck (riddled with wormholes), but a standard grade spruce top.

I've played it at his shop. It's probably about 70% as nice sounding as his real guitars, of which I own several.

I've had the pleasure of visiting his shop several times, and trying all the instruments he had on hand. Typically that was 6-12 of them. Gave a lot of opportunities for back to back A/B with only one main variable changing (body shape, topwood, back/side wood).

They were all built with his particular preferences for bracing etc, so I found the difference makers to be topwood>body shape>>>back/side wood.

That is to say, a cedar and spruce guitar of the same shape and size, with the same back and sides, sounded the most different. Two guitars with the same woods but different size/body shape were the next biggest difference. The same shape guitar with the same topwood but different back and sides were extremely similar but still with noticeable differences in the case of maple vs mahogany vs rosewood...but negligible differences between types of rosewood (like cocobolo vs Indian rosewood).

Having said that, I played his personal guitar built with Brazilian rosewood, and own both a steel string and classical guitar made by him with that wood...and they are magical, just one small step beyond the tone and responsiveness of all 40 or so of his guitars that I have played.

I'm realistic about the factors affecting a guitar, and a factory instrument made in a shape I don't like (e.g. dreadnaught) won't be a dream guitar just because it has BRW back and sides, but knowing David's instruments as well as I do, I can pick out a BRW guitar of his with my eyes closed, easily. There's a ring to the harmonics that no other wood can replicate, though interestingly my second favourite choice was a ziricote guitar of his that I foolishly let go when I had a chance to make a profitable flip on it.

[Telemnemonics]

Well the guitar industry certainly needs players to get used to the idea of alternate woods!

I still find it funny though, despite it seeming obvious that in the case of music, the listener should be able to judge the quality of the tools used by the resulting music; IMO that assumption is just plain wrong.

It is certainly useful, but we can find examples of high quality music made with low quality gear, which may tell us that the music is not a representation of the quality of the gear.
Being able to find situations where higher quality gear directly compared to lower quality gear resulted in different sound qualities, is not IMO enough to conclude that our hearing of musical sound is always capable of determining much of anything about the quality of the tools used to make the sounds.

Reverse engineering those findings to conclude that we can judge the quality of the gear, or the qualities of all the possible tones produced the gear, simply by listening; just isn't a functional scientific method, largely because the player makes the music, and the gear is only one of the tools the player is using. Since a skilled player can make many different better and worse sounds with the same gear, a legit test would need to measure the player input along with the resulting sound. Might even need to measure the players and listeners ability to judge quality vs tone vs amplitude etc before making statements like "has only marginal impact". AKA what does "perceived sound" have to do with scientific experiments?
Science is largely to circumvent the limitations of our perception.
The earth appears flat, and we need science to illuminate our perceptive limitations for us.

Here we have tests that might well determine that the earth is indeed flat, because 52 guitarists and even 52 geologists plus 52 astronomers cannot see the curvature.

But back to the tools vs the user: when I play a darker toned guitar I strike the strings in ways I know instinctually will brighten the resultant tone.
Likewise I automatically play to darken the tone of an overly bright guitar.
Rather than the end result being different, I find my effort is what is different.

Although there are certainly guitars that cannot do what other guitars do, no matter how much effort I make.

[Kent]

I haven’r Read the specific article. I’ll agree with Larry’s summary. The top is, by far, more critical than anything else. The back is pretty low especially as it is very much dampened by the player. The sides have zero sonic attribute as long as they are stiff, suited to the job and aren’t overly porous.

We ran these experiments when I worked at Yairi/Álvarez. Same general conclusions.

[Telemnemonics]

Maybe the writers saying "has only marginal impact on...perceived sound" is a somewhat manipulative way of saying "the luthier can mitigate most of the tonal difference of back & side woods"?

"Six steel string acoustic guitars were built for this study to the same design and material specifications except for the back/ side plates".

We presume that a skilled luthier takes the qualities of the wood into consideration, and adjusts accordingly to get the desired sound.
This is a big part of the guitar industry faced with wood availability problems argument, that you don't need the wood we think we need, to get the sounds we want, because the luthier can build for a sound regardless of the wood used.
Yet in this test designed to show us how mahogany guitars sound pretty much the same as rosewood guitars, we are to suppose that the luthiers making guitars for this test, did nothing to mitigate any differences in the materials?

The acoustic guitar industry wouldn't have to be intentionally misleading us by suggesting that they can make guitars we will like without the lumber we crave. But it would be crazy for "them" to not at least manipulate the instruments and the info for the purpose of calming buyers, so that we will be able to enjoy new guitars made from woods we have come to view as lower quality.

Even within the same spruce from the same growth area, there will be harder and softer pieces.
Even from the same tree, these variations will exist.
Harder should give brighter tone while softer should give warmer tone.
All the luthiers would need to do in building the six test guitars would be to mate the harder spruce with the softer back/ sides, and the softer spruce with the harder back/ sides.
Might they have attempted to test a large quantity of spruce and chosen six sets of close to identical hardness?
Didn't say that in the story.
Can a luthier adjust the tonal potential of a top with minor bracing (and other) changes within "the same design"?

They are trying to help us out here, we need to be happy with our new guitars!

[sookwinder]

I have serious issues with the paper. And for the record I am an acoustician.
Let me mention some immediate issues that sprang to mind as I read this paper.

1.The object of the paper is stated as being:
“does the back wood choice of an acoustic guitar, in itself, produce repeatable and recognisable differences of sound quality?”
Valid question to try to answer .. so why then did they get the players to play the instruments?
If the desire is to determine if there is a discernible difference between back wood construction material, surely it is just a "listening assessment", preferably using the same "neutral" player (not involved in the assessment him/herself) for every "performance".
In fact the way to do it would be to record each instrument once using equipment that simulates hearing/recording at the listeners ears (Bruel & Kjaer have had this type of equipment available for decades) and then getting each assessor to rate each of the performances.
This is what we do in the auto industry when attempting to determine the positive or negative affects of acoustic treatments in vehicles. Record each scenario and then have it played back for assessors in a controllable environment.

Having the participants play the instruments to make the assessment (with all that BS about wearing welding goggles) just introduces biasing factors that are just impossible to eliminate. . When developing a series of experiments to attempt to show or not show something, it is fundamental that the method of observation does not influence the observation.

2. Empirical data is ALWAYS KING.
I can disprove there findings immediately.
1960s Gibson Hummingbird is exactly the same as a 1960s Epiphone Frontier except that the epi has Maple back and sides.
I have a 1965 Gibson Hummingbird and a 1965 Epiphone Frontier.
Strung with exactly the same strings (and someone else playing them) I can tell every time which has the maple back and which does not. I can tell the difference blind folded , I can probably tell the difference between the two guitars when I am blind drunk

3. Just because it has been published in The Journal of the Acoustical Society of America does not mean it is correct. I haven't read the journal for some years now (stopped doing acoustics for my job some year back) but I have read some absolute shite in that journal over the years, especially where the subject matter is so niche' that those reviewing the subject do not have the ability to do the "sniff test" i.e. is this real or is it a stinker.

I remember years ago in some other accredited journal there was a paper about how someone had analysed the 50s Tweed Deluxe amp design and the paper proposed fixes to remove the distortion that was inherent in the design. There was no understanding that the Tweed Deluxe is regarded as being perfect as it is, inspite of its "technical design faults".
Yet again, some one writing a paper and others reviewing a paper with none of the "niche'" knowledge that it takes to understand the question being asked.


I think I shall still be looking for Mahogany, RW or Maple backed/sided acoustic guitars in the foreseeable future.

[Telemnemonics]

I haven’r Read the specific article. I’ll agree with Larry’s summary. The top is, by far, more critical than anything else. The back is pretty low especially as it is very much dampened by the player. The sides have zero sonic attribute as long as they are stiff, suited to the job and aren’t overly porous.

We ran these experiments when I worked at Yairi/Álvarez. Same general conclusions.

Well certainly the top is the biggest materials influence, this is not really in question today as the industry is forced to make expensive guitars with what the market may view as cheap wood.

But for your own experience, would you use your words: "the back is pretty low (tonal influence)" interchangeably with their words: "the back... has only marginal impact"?

Or would you call the (traditionally seen) differing tonal influence between Brazilian rosewood and mahogany back and sides, marginal/ negligible/ unimportant/ insignificant?

While the current debate focuses on the top wood being the big difference, we have already seen convincing arguments from luthiers that even the top wood differences can be mitigated by the builder, where sitka and adirondack spruce has been traditionally viewed as sounding different, guitar players have been fooled by guitars made of the two woods yet not sounding different in listening tests.

[sookwinder]

Just did a 10 sec search and found a B&K binaural head set up being used to set up a mix.
No welding goggles in sight at all!

[Larry Mal]

I want everyone to be clear that I am not endorsing the findings of this study, I haven't even read the whole thing. I'm just throwing this out there for good discussion, which is going great.

[DeathJag]

Fascinating. But more fascinating is the debate here!

My thoughts are also that it wasn’t very scientific. Oscilliscopes and wave form monitors should have been used, man THAT would be some interesting numbers!

[jvin248]

.

If the back wood is important, then the beer gut pressed against the back of the guitar should be more important -- what brand of beer was consumed? Tone Beer! Now a few can get behind that testing program

If you spend some time driving down the road and notice rail and car bridges, you may observe those I-beam thingies under the overpass. In fancy Engineering equations, the web depth is a cubed factor in the strength vs the top and bottom flange widths.

A 3mm guitar top has little influence on supporting the strings, or allowing vibration, compared to the bracing under it. What bit does a talented guitar builder shave away while trying to tap tune the top? They spend all their time on the bracing.

The problem is no buyer/player can really 'see' the bracing -- and that is a critical component for what the marketing department focuses on because they need something to photograph for the advertising copy to tell players what they need to hear and why spend so much on a particular guitar model. Yes, a little industry cynicism.

Get the focus off the magic wood and align it with the actual guitar builder. On the electric guitar side there are certain pickup winders at the Fender custom shop that sign what they built and collectors chase pickups with those specific signatures.

.

[mgeek]

2. Empirical data is ALWAYS KING.
I can disprove there findings immediately.
1960s Gibson Hummingbird is exactly the same as a 1960s Epiphone Frontier except that the epi has Maple back and sides.
I have a 1965 Gibson Hummingbird and a 1965 Epiphone Frontier.
Strung with exactly the same strings (and someone else playing them) I can tell every time which has the maple back and which does not. I can tell the difference blind folded , I can probably tell the difference between the two guitars when I am blind drunk

are you hearing the difference in the back/sides wood, or the difference between those particular guitars though? Two guitars of absolutely identical spec frequently sound totally different.

I'm sure it *all* makes a difference, but having recently got and been blown away by an all laminate Yamaha FG160 from the early seventies, I'm deffo a design-believer!

[sookwinder]

the difference I hear is the maple back and sides, not "just those two guitars".
Difference in attack, tonality and to a lesser degree timbre.
I have heard it in every other version of those 60s guitars be they in real life or in recordings.

[timtam]

The litmus test of good science is replication - in this case similar results have been found before. eg the Leonardo project has been testing non-endangered woods to see if they are judged by players as similar sounding to traditional acoustic guitar tonewoods. To do that, you definitely have to blind the study, to remove strong beliefs that the traditional woods are 'better'. When blinded, differences in judgements favouring traditional woods often disappear ...
Summary here ..
https://hazeguitars.com/blog/listening-tests-tone-woods
More details here ...
http://www.leonardo-guitar-research.com/
https://sites.google.com/site/leonardog ... ine-test-1

[sookwinder]

we have to be careful here not to conflate two distinctly different questions...

- does the back wood choice of an acoustic guitar, in itself, produce repeatable and recognisable differences of sound quality (taken from the paper in the OP)
- can other types of woods (that are not endangered) be used to produce an instrument that is equivalent and acceptable to existing manufacturing materials and player expectations

for me the answer to both is yes.