Long-term scientific "blind" acoustic string test

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
Hi folks.

As I mentioned in a couple of other threads, I'm about to start a "blind" string test of phosphor bronze lights (12s) from a variety of different brands. (I'm putting "blind" in quotes here because a true blind test would have more controls than I'm able to bring to bear, as you'll see in a moment.) I could have continued the earlier thread to do this, but I wanted a fresh start for the "official" test, and also I wanted to get away from the lingering stench of the Orphee strings.

Here is how the test will work. I've obtained several various string sets. Each time I'm ready to put a new set on, my wife will randomly select a set in another room and bring them to me unpackaged. (This is the "blind" part.) I'll install the set on the guitar (a 2021 Oxnard-built D-55), play them, and evaluate them at three intervals: immediately, after two days' break-in, and after two weeks. Typically I leave strings on for a month at a time, so any test set I leave on for a full month will receive a final evaluation; but if I really don't like a set, I'm not going to play them any longer than two weeks. I'll be making and posting personal assessments of tone, volume, construction quality, and ease of play. (I know those things are partly or fully subjective, but it's the best I can do.)

After the two-week evaluation, my wife will show me the packaging for the set I've been evaluating, and I'll finally know.

For those who are of a more scientific mindset, I do plan to collect some objective data on each set as well. I'll be maintaining a downloadable spreadsheet containing my evaluations but also some physical descriptions and measurements. I'll check the gauges of each string to see how closely they conform to what's stated on the package. I'll measure the strings for length. And I will do a rough check of flexibility by means of a testing jig I built:

20211227_145246.jpg

Basically a string sample gets clamped into the jig:

20211227_144733.jpg

And then a weight (the binder clip, with a hole drilled in it) is attached to measure how far the weight deflects the string:

20211227_144915.jpg

It's not a precision instrument, but it should still provide some useful comparisons. I'm still working on where along the sample to attach the weight for best data.

The strings being tested are:

MakerNameModelTested
Black DiamondPhosphor Bronze Light 12-53N600LSet E
D'AddarioPhosphor Bronze Light GaugeEJ16
D'AddarioXTXTABP1253
DarcoAcoustic Lights 12'sD220Set H
DRSunbeamRCA-12Set F
DRVeritasVTA-12
DunlopPhosphor Bronze 12/54DAP1254Set B
Ernie BallEarthwood2146
GHSAmericana SeriesS425 Light
John PearsePhosphor Bronze Wound Light Gauge600LSet A
La BellaPhosphor Bronze Light7GPS
MagmaPhosphor Bronze Med/Lt 12-54GA140PBSet G
MartinAuthentic Acoustic SPMA540Set D
RotosoundJumbo KingJK12
SIT StringsRoyal Bronze Fusion WoundRL1254
WebstringsPortland AcousticLightSet C

Now, I'm sure some of you have seen a flaw in the procedure here. "Dwasifar," you cry plaintively, "you'll know the D'Addarios because of the color coding." And that's true. It's true of the Rotosounds, too; and it's true of the DR Sunbeams because they're round core. This is one reason why I included a set of XT, the only coated string in this test. I added them (at the last minute) so that when I get a set with D'Addario color coding I won't know whether it's the XT or the EJ16; and when that happens I will ask my wife not to show me the packaging at the end of the two weeks, but rather bring out the other D'Addario set next, so I can review them back to back. It's not totally blind in that I'll know they're D'Addario, but I won't know which D'Addario until I've played them both. As for the Rotosounds and Sunbeams, well, all I can do is try to pretend I don't know and be as objective as I can.

To further "blind" it up, I have included three sets of Martin MA540. Why? Well, just to add a little more uncertainty about what I'm playing. Right now the guitar has Martins on it, and they sound great, so I figure it wouldn't hurt to have them come up more than once, to make things a bit more random and also to see if I rate them the same each time. I'll bet I don't, which would sort of invalidate some of the more subjective elements of the test, but on the other hand it's valuable information in itself. Why Martin and not any of the others? Because right now, on sale for $4.99, they're cheapest, and doing multiple sets of multiple brands would take years to complete. As the test group stands now, it will probably take the better part of a year to finish.

That's it. Test begins on Jan 1, 2022. If anyone has any suggestions, tips, or pointers before I start, I'm all ears.

Downloadable spreadsheet of results here.

Edited 01/08/2022 to add Black Diamond and Magma sets.
Edited 03/17/2022 to add indexing links to the start of each tested string's evaluation period.
 
Last edited:

walrus

Reverential Member
Gold Supporting
Joined
Dec 23, 2006
Messages
24,019
Reaction score
8,101
Location
Massachusetts
You have much more patience than me - kudos! Looking forward to the results!

Random observation - as soon as I saw the photo of the "bent" string on the jig, I saw a butterfly and a grasshopper in the shadows. I know, I know, it's not a Rorschach test, it's a string test!

walrus
 
Last edited:

MacGuild

Member
Joined
Oct 1, 2021
Messages
250
Reaction score
354
Location
Account abandoned.
Guild Total
14
You didn't mention recording. If you don't already have a plan, dwasifar, consider recording your various string specimens at each of those three string-life intervals you cited. Same recording in the same conditions every time; open this string, open that string, open strum, etc, whatever standard you like. Take advantage, you probably won't do this experiment twice. It would be pretty cool to have recordings of, for comparison, an open A-string ringing from strike to decay, from a dozen different sets. And so on. From a smartphone to a DAW, lots of ways to document your experiment.
 

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
You didn't mention recording. If you don't already have a plan, dwasifar, consider recording your various string specimens at each of those three string-life intervals you cited. Same recording in the same conditions every time; open this string, open that string, open strum, etc, whatever standard you like. Take advantage, you probably won't do this experiment twice. It would be pretty cool to have recordings of, for comparison, an open A-string ringing from strike to decay, from a dozen different sets. And so on. From a smartphone to a DAW, lots of ways to document your experiment.
I didn't, you're right. There's a reason, though.

It is actually the second time I've done a test like this. (The first was a different gauge and different guitar on a different site.) I started off that test doing recordings, but quit after the first three or four because of difficulty keeping them consistent, and the hassle of indexing them for people to hear online.

I haven't yet ruled it out, but didn't want to promise it.
 

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
Random observation - as soon as I saw the photo of the "bent" string on the jig, I saw a butterfly and a grasshopper in the shadows. I know, I know, it's not a Rorschach test, it's a string test!
Haha! I see them!

The grasshopper is a shadow. The butterfly is a stain. This piece of wood was originally just a random scrap of ply that I used to put under the front of the snowthrower when bringing it in wet, so it wouldn't stain the garage floor. It was in my scraps box and it was more or less the right size, so I gave it a couple of light passes through the planer to clean the worst of it off the surface and reused it for this. No need to cut it deep enough to remove all the stains.
 

Nuuska

Enlightened Member
Joined
Jan 18, 2016
Messages
7,706
Reaction score
6,077
Location
Finland
Guild Total
9
Sounds great - except for one thing.

After testing - give the strings back to you wife with a test number written on paper to go back into the package.

This way you can not tell which strings are left when you eventually get there. First after all sets have been evalueted - she can show you the packages.

I know this might be nitpicking - but after all tha tedious & longlasting effort - you do not want to know that the last set to be evaluated is "brand X" - knowing it might influence your judgement.
 

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
Sounds great - except for one thing.

After testing - give the strings back to you wife with a test number written on paper to go back into the package.

This way you can not tell which strings are left when you eventually get there. First after all sets have been evalueted - she can show you the packages.

I know this might be nitpicking - but after all tha tedious & longlasting effort - you do not want to know that the last set to be evaluated is "brand X" - knowing it might influence your judgement.
If I do that, then I can't post anything useful to readers of this thread until the entire test is complete. It doesn't help you, as a reader, to know that I've tested sets A, B, C, D, and E so far, and D was bad but B was great. Without telling you what brands they were, there'd be no point in documenting progress in this thread; I'd basically just have to go away for a year and come back with a completed spreadsheet.

It is a flaw in the process, I grant you, but to keep it interesting for the readers I'm just going to try to work around it. If I don't keep reviewing the original list and comparing it to what's been tested so far, I stand at least an even chance of not remembering a year from now what all has been tested and what is left to go as I come down to the final set.

Or, here's an idea, I can just test the final two the way you suggest. So even if I know Brand X and Brand Y are the final two, I won't know which is which until both have been tested. This delays reporting of the penultimate brand's results only. What do you think?
 

Stuball48

Senior Member
Gold Supporting
Joined
Oct 22, 2017
Messages
4,785
Reaction score
2,582
Location
Dickson, TN
Looks like a test for those who really play. I like it.
As a retired teacher, I never had much confidence in those with a PHd leading a teacher discussion who had never been a classroom teacher.
Bet you I will try your "winner."
 

mavuser

Enlightened Member
Joined
Sep 15, 2010
Messages
8,213
Reaction score
2,743
Location
New York
this is super cool and I love your jig! for selfish reasons, I would like to see u add Martin Vintage/Monels and some of the newer Ernie Ball acosutic strings Earthwoods I think they are called. ("silk?"). and do the whole test in 11 gauge sets 😬 and then 9 gauge 12 strings sets 😬🤷‍♂️🤘
ur wife is keeper btw
 

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
this is super cool and I love your jig! for selfish reasons, I would like to see u add Martin Vintage/Monels and some of the newer Ernie Ball acosutic strings Earthwoods I think they are called. ("silk?"). and do the whole test in 11 gauge sets 😬 and then 9 gauge 12 strings sets 😬🤷‍♂️🤘
ur wife is keeper btw
Going to pass on the Monels; this test is just for phosphor bronze, so no 80/20, Monels, nickel bronze, or aluminum bronze. But Earthwoods, yeah, I'll add those. I just ordered a set, they'll be here Jan 3. Editing the list in the first post to reflect the addition.

The reason I didn't put Earthwoods in from the start is because they're visually identifiable even out of the package. They have gold-colored core wire, but no color coding on the ball ends. But if I can be semi-objective about the Rotosounds, I guess I can be about the Earthwoods too.

I don't have any guitars that take those other gauges. But if you do, please swipe my idea and run some tests. :)
 

krysh

Guildarist in the mod squad
Joined
Apr 28, 2007
Messages
4,427
Reaction score
907
Location
near hamburg*germany
Guild Total
6
I‘d love to hear 1min recordings of the fresh sets with the same song. And dont come with John Cage.
 

dwasifar

Senior Member
Joined
Apr 19, 2021
Messages
1,080
Reaction score
935
Guild Total
1
No Elixer or Monals ?
Nope. This test is only for phosphor bronze. No 80/20, nickel bronze, Monel, or aluminum bronze either. No Elixir because the test is intended for uncoated strings. I only included one coated set, from D'Addario, and I only did that because otherwise the EJ16 set would have been immediately identified.

I have to draw the line somewhere, otherwise this would turn into a Test Of Everything and take four years. 🙃
 

Wellington

Member
Joined
May 27, 2021
Messages
128
Reaction score
110
Guild Total
1
Nope. This test is only for phosphor bronze. No 80/20, nickel bronze, Monel, or aluminum bronze either. No Elixir because the test is intended for uncoated strings. I only included one coated set, from D'Addario, and I only did that because otherwise the EJ16 set would have been immediately identified.

I have to draw the line somewhere, otherwise this would turn into a Test Of Everything and take four years. 🙃
I believe Veritas are coated too, well the core is anyway.
 

Rayk

Enlightened Member
Joined
Apr 19, 2015
Messages
5,786
Reaction score
1,184
Nope. This test is only for phosphor bronze. No 80/20, nickel bronze, Monel, or aluminum bronze either. No Elixir because the test is intended for uncoated strings. I only included one coated set, from D'Addario, and I only did that because otherwise the EJ16 set would have been immediately identified.

I have to draw the line somewhere, otherwise this would turn into a Test Of Everything and take four years. 🙃
Ok then I have the winner in mind lol though it’s shelf life might not make the 2 week limit . Lol 😂
 
Top