Study result - Effect of back wood choice on the perceived quality of acoustic guitars

m_c · January 2019

A mate just sent me a link to this video - https://www.facebook.com/lancasteruniversity/videos/1184000751753057/

The video only provides a brief summary, but the full details can be found at https://doi.org/10.1121/1.5084735

It certainly doesn't rule out the tone wood debate, but it certainly casts doubt on the actual difference it makes.

fishfingers · January 2019

Thanks for posting that. I recently played the maple test guitar; someone local was selling it and he had taken part in the study. It was a good guitar and sounded much like other Fyldes of the same size that I've tried over the years.

Andy79 · January 2019

Notoriously difficult to control for. Subjective outcome
was it blinded?
i saw the luthier was also the co author...

mgaw · January 2019

its the builder who makes the difference surely and of course how much the player has paid for the guitar

barry2tone · January 2019

Andy79 said:

was it blinded?

Summary;
https://www.alphagalileo.org/en-gb/Item-Display/ItemId/173207/fbclid/IwAR2B9ZnJw3gJAjfpXDHlSVReC-y13x0QHU_4PvO3tOGDNk45Fxb79Ms-3NE?returnurl=https%3A%2F%2Fwww.alphagalileo.org%2Fen-gb%2FItem-Display%2FItemId%2F173207

Paper;
https://asa.scitation.org/doi/10.1121/1.5084735?fbclid=IwAR0Vfnb3IG3V27U2ECKpu8sZGTyuiwjsMTsQtGaMGuyD0e_GStDGaHiQgXc&

The paper provides some answers. (clues - welders goggles, dim light, retuning for each pass, pink noise.)

I've not seen another study with such a large number of subjects.

It's brave of RB to have taken part. It'll be interesting to see if it changes his approach to building.

The guitars

Image: https://www.alphagalileo.org/DTControls/DocumentsUploaderControl/DocumentsController.aspx?action=download&document_id=109826&hit=true

ToneControl · January 2019

this link is a bit easier to read:
http://www.psych.lancs.ac.uk/hearing/the-guitar-experiment/

if you click on the graph, you can mess about with it to see elements of the data in more detail

ToneControl · January 2019

I notice that the pro players could tell the walnut from the others

That's interesting for me since I once had to sell on a walnut back+sides guitar since it was just too bright sounding for me as a player who uses nails a bit

ToneControl · January 2019

I think these guys missed a massive opportunity here:
They have looked mostly at the perception of the player.

I think they should have teamed up with an acoustics or physics dept, and measured and analysed the sound output
They could have engineered a repeatable plectrum pluck with a robot, and analysed the audio
(Also playing it back to listeners to rate)

Then we'd know if it was "all in the heads" of the players

ToneControl · January 2019

Why did they use the Elixir 80/20? They are very bright jangly strings, not the normal choice for most people
I think these would make it a little harder to hear the different between guitars. I only use them on a baritone

crunchman · January 2019

Interesting that nothing scored above about 3.8 overall. That's probably partly down to the Elixir strings as @ToneControl said but it's also the fact that it is Fylde guitars. If I was looking for a new guitar, they wouldn't be on my shortlist based on the ones I've tried. I've never found the ones I've tried to be particularly responsive - and the guitarists in this experiment seem to agree based on the scores. I'm not sure that they are the best test bed for this kind of experiment. You would want a more responsive guitar that isn't deadened by the contruction and bracing.

The other thing to note is that the construction is not identical. If you read the full paper, he compensates for the different woods with different builds. They say in the paper that a stiffer wood like rosewood can't be bent for the sides if it's too thick, and some of the other woods need to be thicker to provide the required stiffness. That's how he would build the guitars in real life, so those constructional differences would be what people perceive as pasrt of the wood differences. In some ways that is fair enough, but does mean that the wood is not the only variable.

It's interesting that the playability scores for the different guitars are different. If they are of genuinely identical construction, then the playability should be identical. This does suggest construction and/or set up differences between the guitars.

A better test would be to get a large number of factory produced guitars that are consistently made the same way. When it's only one guitar of each wood, it's not enough to be significant. Working with the UK Martin distributor to get 10 D18s and 10 D28s would be a far better test.

Even if they did get 20 Martins for the test, they would need to change the methodology. A D18 sounds different from a D28, but if I had to give them an overall rating, they would probably both get a 4 to 4.5 with decent strings (not Elixirs). The overall rating is mostly meaningless.

Even with the methodology they used, the ability to correctly pick the right guitar out of a pair was "credibly above chance" to use their woods. With a guitar that is more responsive than a Fylde, I suspect that the differences would be more pronounced.

I'm not particularly a wood snob. I've got a really nice high end Martin Dreadnought, but I've also got an OOOX1AE with HPL back and sides and a Stratobond neck. For a while I had a Martin OOO28EC. There was a limited edition of that with Brazilian Back and sides that I tried. That particular guitar didn't sound that great to my ears. Two or thee of the better Indian Rosewood guitars I played sounded better to me.

There are a lot of factors that affect tone of an acoustic guitar, and body style, bracing and other factors are more important, but there is a definite difference between a D18 and D28 to use the most obvious example.

xscaramanga · January 2019

ToneControl said:

I notice that the pro players could tell the walnut from the others

Not particularly well:

Lancaster University said:
[A]s a rule of thumb a d’ of one represents poor discrimination performance, a d’ of two correspond to a mediocre performance level, a d’ of three to a good performance level

The pro players were managing a d' of about 1.5, with the error bars topping out at 2. So no one was consistently or easily telling the guitars apart.

brojangles · January 2019

I don't need persuading that a bunch of guitar players probably can't tell different woods from one another in a blind test, but surely using Elixir strings to make them all sound like shit weakens the point being made?

ToneControl · January 2019

brojangles said:

I don't need persuading that a bunch of guitar players probably can't tell different woods from one another in a blind test, but surely using Elixir strings to make them all sound like shit weakens the point being made?

Plenty of people think that normal Elixirs sound great. It would be way more consistent for this test to use them, rather than putting a set of non-coated strings on that are way too bright for one day, then way too dull after a few days. However, the 80/20 strings are unusually bright sounding, I have no idea why they would choose them

TimmyO · January 2019

No one study can cover every angle, but this is impressively done - I really hope it's used as the basis for future studies covering other aspects.

Other things it would be cool to see them do:
Get him to make 6 "identical" guitars and see how well or not those correlated.
Repeat the whole thing with a more typical lower-cost production line manufacture method.

LastMantra · January 2019

TimmyO said:

No one study can cover every angle, but this is impressively done - I really hope it's used as the basis for future studies covering other aspects.

Other things it would be cool to see them do:
Get him to make 6 "identical" guitars and see how well or not those correlated.
Repeat the whole thing with a more typical lower-cost production line manufacture method.

I thought they would have done that. Even a blind test with ONE guitar as a control.

Maybe they did...I haven't read it all.

TimmyO · January 2019

they seem to have (for the purposes of this) been comfortable with the assumption that all 6 were identical enough having been hand built by the same experienced builder to the same measurable tolerances.

As I say though - they've gone FAR further than anything else I've ever seen and for hand-built guitars at least it does throw up some very interesting questions.

Of course it is also possible to conclude that a good luthier can build out any effect of material - remains to be seen if the same is true for factory items.

bugileman · February 2019

Not a fan of elixirs considering how they break down so easily, when examined under a microscope. Never got on with them, always thought they didn't bring the best out of a guitar.

stickyfiddle · February 2019

bugileman said:

Not a fan of elixirs considering how they break down so easily, when examined under a microscope. Never got on with them, always thought they didn't bring the best out of a guitar.

I don’t care how they look under a microscope, but I stopped using them on my Martin because they sounded shite!

NelsonP · February 2019

Paging @Three-ColourSunburst.....

Howdy, Stranger!

Categories

In this Discussion

Become a Subscriber!

Study result - Effect of back wood choice on the perceived quality of acoustic guitars

Comments