Bow Testing - for Paul N. & co.

June 20, 2019, 12:20 PM · Hi guys,

From this thread: Nickel mounted vs Silver mounted bow - difference in tone?

I made a video:

Here's some information about this vide and what variables I could control, and what I couldn't:

First, the recorder is a ZOOM Q2 HD Handy Video Recorder. It's not super fancy, but it doesn't suck. I have it set to the highest audio recording quality, and placed it on a bar stool about 25 feet away from my position.

Second, the bows. They are (not in the order that I played them):

A. a CF bow from Shar. The mark has worn off the frog, but I believe it was called "Presto." I have had it rehaired, so these are not the original hairs it came with ;) but even so they are older than the other hairs in this group. The ribbon is still in pretty good shape, however.

B. a fine old bow stamped "KITTEL." It is unprovenanced, but possibly the real thing. The stamp is genuine, so if not a real Kittel, it could be a Knopf or Bausch or something along those lines.

C. an early 20th century bow by Albert Leeson stamped W. E. Hill & Sons.

D. a contemporary bow by Pierre-Yves Fuchs.

I did not feel like getting out the rubbing alcohol, but I cleaned off all the hair with my microfibre rosin-cleaning cloth until it felt slippery and the old rosin dust stopped coming out. Then I applied 4 strokes of Bernardel to each bow.

To the best of my ability I set each bow to the same tension, meaning that it felt like it took the same effort to push the belly of the stick all the way down to the hairs.

For the long notes I tried to keep the speed even, confine the bow to contact point #3, and let the natural weight of the bow create friction.

For the polishing stroke in the Kayser etude, so that you could hear how each bow responds to receiving energy from the hand, I bore down just enough on the stick that I could feel the springiness of each bow engage and push back against me.

If you feel like it, leave your guesses in the comments! :)

Replies (140)

June 20, 2019, 1:10 PM · What's the question exactly? Be brief.
June 20, 2019, 2:17 PM · 1. Do they sound different
2. Can you tell which one is which
Edited: June 20, 2019, 2:37 PM · Nate, first, thank you so much for taking the time, you're very nice. You deserve "best user of the month. Really, much appreciated! :)

Steve, I guess the question is: guess the bows.

It's fine that we can get the chance to guess, but a proper experiment does not let the audience know if you're playing one bow or another.

That means, in your video you could perfectly have used the very same bow in the first 4 notes. That's when an "expert" can mess it up massively saying "the bow 1 was making the sound brighter, it was so obvious", and then it's when you tell them "it was the same bow". I'm certain in this experiment, properly made, we would find this situation so many times. But again, thank you for the time spent and for at least creating a kind of "blind test". I will listen to it properly with headphones later and tell if I can hear differences. I definitely should hear them in the pieces or melodies, I guess, because you surely control each bow differently.

Also, guessing bows is just very vague, and I'll explain myself: one should explain its reasons why he's exactly choosing one bow or another. If not, you could be randomly guessing. That's why it's so important to not let know the audience if you're using one bow or the other. By process of elimination, you have way more chances to correctly guess, if you're given 4 options. Instead, if you don't know anything about the bow and also you have to explain why you choose a specific bow, that's when one that can actually differentiate the bows can easily explain the reasons and guess correctly, while someone guessing randomly and heavily influenced by the Placebo effect will simply fail so bad, expressing contradictions constantly.

Anyways, as I said, I will give you my opinion, whether if it's that I can't hear any difference or I do, but I'm not home now and I must use headphones.

Edited: June 20, 2019, 2:59 PM · Hi Paul,

Not that you're wrong, exactly, but there is a certain amount of good faith involved in a project like this. What I mean is, I might have lied about what bows I'm playing on, because I want to feel like I'm important and have famous bows, or even just to make the test better. Maybe I don't really have any good bows, and those are just some crappy fiberglass bows I got at a yard sale. Or maybe they're 4 identical Coda bows from Shar. You can only trust my word (or not!) that they're the bows I said.

Similarly, I trust you to listen honestly, and maybe jump around in the video to listen to the bows in a different order, so that, for example, you hear 1 and 4 next to each other, and so on. Plus, since it's a recording, you can download the video, extract the audio, and load it up in something like Audacity that will let you visually inspect the signal, and do spectral analysis and other complex things. I'm not saying you should necessarily do that extra work, but that is an option if you want to visually confirm that the sounds produced by the four bows had different effects on the microphone's membrane.

Anyway, have fun. :)

P.S. Something you might find instructive is to click on different timestamps in the video so that you switch from hearing one bow to the next bow in the middle of the stroke. So, say, click 0:22, then right away click 0:33, then 0:46, then 1:00.

Edited: June 20, 2019, 3:35 PM · Yeah, sure, I will thoroughly check for differences.

Well, of course you could have lied to us and made us believe you were using 4 different bows while all of them were a cheap chinese $40 bows from a factory. That's a "nasty" way to caught a golden ear audiophile, people I've met sometimes and never get tired of proving wrong.

Nevertheless, I think it's OK to properly explain the experiment to a mature audience: "You will listen to 16 bow strokes performed in the 4 different strings by a violinist, played as consistent as possible; we have 2 different bows, CF and wood. Just for the sake of God, each bow will be played at least once, but nothing else you will know. Only that the bows are using the same hair, tension and rosin, and its amounts will be equal as well. What will change is basically the "stick", between each sound. We could trick you and play the exact same bow 16 times, but we want to be good guests. Guess the bows"

BWT, it's not "that great" to accurately analyze the sound wave. Why? Because the same bow with the same stroke will never ever produce the very exact same wave, so starting from there, you can imagine how chaotic analyzing would be, and also you get as much precision as your mic lets you. A mic could record a different wave even if it's being exposed to the exact same wave over and over again. It's what happens in the real world, magnets have errors, there's friction... so, don't go crazy, hahahaha.

Edited: June 20, 2019, 3:35 PM · Very interesting, thanks! My appreciation of the bows:

no.1: neutral
no.2: a weaker, softer sound
no.3: almost metallic sound
no.4: between 2 and 3, good volume, probably the best bow.

But I don't know enough about bows to then match this to the actual bows you list. But they clearly sound different from each other, although this may be psychological as Paul correctly warns us.

June 20, 2019, 4:54 PM · Paul N wrote"
"Well, of course you could have lied to us and made us believe you were using 4 different bows while all of them were a cheap chinese $40 bows from a factory."

Sure, any number of things are not outside the realm of possibility.

Edited: June 21, 2019, 5:38 AM · God, I love the violin, in general. I listened to you and just wanted to say that, hahaha. I liked your playing as well, fantastic.

So, I've listened to this with a pair of good headphones, once, and here's the quick review.

The first open notes sounded pretty much exactly the same to me, I hardly, hardly believe I can cut the 4 fragments to create 4 samples, queue them in a long repeated list randomly sorted and correctly place the sample with its bow, where it corresponds, recognizing the samples of the same bow, and that for each one of the samples. Long story short, I don't think I would be able to tell them apart, by any chances or means.

Second, in the piece, there I noted some very tiny little nuances, but I sense I'm getting huge expectations influences. I say this because the third one sounded a little nasal to me, I wanted to check again before writing that down and I now can't really say it's nasal anymore.

I once again believe that I would have a very, very hard time figuring out, if I listen to the four samples of the piece in a continuous list, randomly queued, which sample is the same. In other words, I'd listen to the sample of the bow 1 and I'd say it's the bow 3 sample, but later I'd recognize the same sample as the bow 2, etc...

There's a cheat though, I could easily recognize patterns in your playing, because you did not play at all like a robot, that means, the same notes were played faster with one bow, you vibrated differently each time, your tempo was not a Swiss watch... Patterns, basically.
So, I could focus on the patterns of each sample to, at least, be able to recognize and correctly select always the sample of a bow. But, then I'd be comparing differences in the way you played and the tiny difference between how you performed each time, not in bows. That's why it is so important to suppress any kind differences due to the violinist playing differently, which is inevitable even when doing a simple scale run. That's why an open string with a consistent sound is the best possible way to test if the violin sound changes, because you are not including the violinist ability to control of the bow, hence you control more or less the variables you want to be constant and not disturbed.
In a simple, long stroke, of an open string, you can't do pretty much anything: you can not do vibrato, so we can't cheat focusing on the differences of vibration in each sample to recognize when it's sample 1, 2, 3 or 4, you should be able to control the attack of each bow, etc... Basically, a long bow stroke in an open string kind of guarantees that everything stays the same, except the bow. That's why I insisted so much in this.

And, I love how you play, hahaha, really. I might do the A-B thing to test if I can really at least recognize which samples are from the same bow.

June 21, 2019, 12:57 AM · I think number 3 is the carbon fibre bow, brighter than the others less warm
June 21, 2019, 2:10 AM · A snappy question deserves a snappy answer, but unless Nate gives us a selection of boxes to tick he won't get one!

1. I think 3 sounded different (help - I'm with Lyndon!) but it was marginal and I might change my mind on further exposure. It's important to prevent the experimental subject from second-guessing.

2. Unanswerable because this requires a preconception (based on experience or prejudice) of how a CF bow for example is supposed to sound. And of course the danger is that if we think this CF bow stands out from the others in one particular way, we may assume that's true of all CF bows

But above all, from a listener's perspective I'd say the differences simply aren't important. The importance to the player on the other hand is entirely a matter for them to decide

June 21, 2019, 5:20 AM · The 3 seems to have a drier sound, but it seems to project well, in high register too.
I like the 4 better than the 2, the 2 better than 1.
June 21, 2019, 9:18 AM · Hi Paul,

As a matter of fact, I agree with you, at least up to a point. The four open string samples clearly sound different (i.e., they are obviously not the same bow played 4 times in a row), but if I wait a few days I will probably forget what order I played them in. When that happens, I won't be able to tell you which one is which just from listening, and these are my *very own bows!* (Don't worry, I wrote the answers down on a piece of paper so that if the thread goes on for a while I'll still be able to give the answers. :)

I'm afraid you are also correct, I wasn't able to be quite as machinelike as I could wish. I also spotted a little vibrato creeping in to the etude here and there, darnit! I don't normally use a continuous vibrato, but I've been practicing it because they like to hear it at auditions. Now it's rearing it's ugly head when I haven't asked for it. Grr.


For bow trials, this sort of default generic "what does the bow do when I leave it alone" type of playing is the first step. Long tones on open strings, trying to interfere as little as possible. The reason for this step is to eliminate bows that have problems - basically, ones that do things when I haven't told them to. Some bows will randomly jump, or let go of the string, or skid around, due to flaws in the stick. At this stage there will not be a big difference in *sound quality* from bow to bow. If I play a bunch of open D strings on the $2,000 "normal bow" from the TwoSet video you mentioned, you will not feel like making an OMG face when I switch to the $200,000 D. Peccatte.

Later today I'll record another test that you may like better, but for now I have to go run some errands!

@Steve Jones,

I'm not sure if you mean "snappy" as in quickly, or as in irritated, but I didn't mean to be rude. I agree with your comment in full, especially point 2.

June 21, 2019, 9:51 AM · Nate B
Thanks for posting, I only had time to listen to the open strings, very interesting differences, not big but still noticeable.
Edited: June 21, 2019, 11:04 AM · Steve Jones point 2 is partially right, not entirely right. "...this requires a preconception of how a CF bow is supposed to sound"

That's not entirely true. Sure, in order to say "this is a CF bow", you of course need to know how a CF bow sounds, if it is true that it sounds or make the violin sound in a certain way.

What's wrong about point 2 is that you do not need to know, and I will go further, you should NOT know how a CF or wood bow behaves or "sounds" to tell them apart, because then you have a preconception, an expectation, and that means Placebo will hit in, granted.

In example, you don't need to know how a violin sounds and how a piano sounds in order to tell them apart. You can obviously hear a difference in the sound (say you play an A note in both), but you can't say which one is a piano and which one is a violin. That's OK, we are not testing that, we are testing if you can hear a difference in sound, and you can, in the case of a violin vs a piano. Hence, you can prove you can tell a piano and a violin apart, even if you have never ever heard them and can't connect the sound you are listening to one of them.

That's why a blind test is necessary, because even with all kind of prejudgments, looks and price tags in your mind, they will be useless and your words will be put to test: if you are actually hearing a difference between bows, you will easily recognize a pattern in the change of sound. If not, you will fail the test.

It's very important that you are testing the changes in sound. Why? Because as I said, if you only play once bow 1 and bow 2, you can recognize patterns of a singular performance, and with that reference, at least, be able to tell when you are hearing to bow 1 and bow 2. You would be cheating because you would be using the performance as a reference and not the sound, but the test wouldn't be able to caught that.
That's why in my experiment design, I didn't say "play bow 1, play bow 2, record them and tell them apart in a blind A-B test". It wouldn't work.
You could be lying to us and to yourself, because you will be making decisions based on the patterns of the performance (the first bow started a bit more aggressive, it reached the maximum amplitude of the open string earlier...) instead of the patterns in the change of sound, which is what we are testing.
The patterns in the performances are inevitable since we are not machines with parameters, that's why in my experiment I tell you to play 16 times, because that kills any chance for you to even recognize if it was bow A or bow B basing your judgments in the patterns of the performance, because there would be 16 different patterns, not a single reference you can use to compare. Only changes in sound, if it is that there are any and if it is that you can hear them. That's why it's so important that the hair and the rest of variables stay the same, so we are completely isolating the effects of the bow and testing its effects, and not the effects of a singular performance or the ability of a violinist with a specific bow to play a more steady open A due to the perfect balance and condition of hair vs a very bad unbalanced stick with hair jumps and slides due to bad "plastic" hair.

In example, if you are comparing bow A (60g) against bow B (10Kg), since the violinist won't be able to create a constant sound between the bows due to the impossible maneuver of the heavy one, you will surely know when bow A or B are being played. But, I hope, luckily, no bow will be able to overcome a player's ability and make him or her unable to play a constant sound.

Edited: June 21, 2019, 10:40 AM · Nate - by "snappy" I meant your question was quick and to the point, just as I requested! (actually I was being slightly rude about Paul's prolixity..)
Edited: June 21, 2019, 11:05 AM · I'm sorry for my Prolixity, it's simply that I need to explain myself as good as possible so I don't get misunderstood and someone cherry picks my words. Explaining as good as possible, due to the audience behavior, most of the times means you have to explain every single thing to the maximum reasonable details, if not, your ideas will get out of context constantly.

I've been accused of saying that the bow is not important, that sums all up. Hence, I shall bring the bricks. Sorry, I don't like it neither, it's better a cute, short message, and more appealing.

June 21, 2019, 11:09 AM · @ Steve Jones,

All good! :)

@Paul N.

Clear communication is always hard, and especially hard about specialized technical subjects. For my part, I appreciate the effort, and I'm not too bothered by the length of your posts. I think that the distinction you're trying to draw is that there is a difference between:

A. noticing that two sounds are different

B. inferring from from a sound what the cause of that sound was.

For example, when you play open A and open E strings for a new student, they can always tell when you switch from the first string to the second string, but they can't specifically identify the E string when they hear it, or, sometimes, tell which of the two strings is higher in pitch.

Is that fair?

What I'm going to do now is take two of my bows and set up the coffee table so that the screen of my computer is between the bows and my camera. Then I will roll a 10-sided die. I will play the first bow that many times, putting it back on the table after each time. Then I will play the second bow enough times that I will have played 10 times total. Your job is to figure out at what point the switch happens.

My die is numbered 0 - 9; if I roll a 0 I will roll it a second time. If I get an odd number I will play bow 1 all ten times, and if I get an even number I will play bow 2 all ten times, so there is a (somewhat small) chance that you will be listening for a change that never happens!

Edited: June 21, 2019, 12:11 PM · Yes, of course, that's another fair example. I was simply trying to explain that you don't need to know how the A note and E note sound in order to tell them apart. As I said, I must go further, you under any circumstance must not know what to expect from a CF or a wood bow, because that's destroying your judgment, you are setting yourself up to listen to those things you've been told must happen, you're establishing a base, a limit. In a correct blind test it won't work, no matter what you know, but it gives better, more clear results. Why? Because the reasons given in my experiment, if you don't know how a CF bow should sound, will be based entirely on your hearing, entirely on the facts, the reasons will be truthful about what you're hearing, and no prejudgments will take place. It's clear that if Lyndon says he's hearing a CF bow, the reasons given will be "bright" sound most of the time, because he supposedly knows information about what he has to say before doing the experiment. It shouldn't be a problem for honest people, but your mind will scream to you "you think it's CF so you most of the time must put what you know, not what you hear". He will still fail very bad the test if he does not hear any difference, but basically the reasons given will be made up mostly, influenced by external factors to the experiment.

The perfect experiment is to not tell which bow you are using, play it 16 times or so, swapping randomly, OR NOT! (you could play with the same one the 16 times), and then let "experts" explain what they are hearing after each stroke. Let them say which bow was used and explain why, as specific as possible. If you get constant correct results from most of the people, then you can say they are clearly noticing something in sound that make them easily say it's bow A or bow B. Then the fun part it to read the reasons: if there are two experts that correctly answer the first 5 strokes, but then you read the reasons and observe complete opposite reasons, well, something's smelly.
Could an expert say the sound of the 3rd stroke clearly reveals a bow change because it was brighter and more open sound than the 2nd stroke, while other expert say it was a different bow because it clearly showed a more dead sound, with less treble than the 2nd stroke?
That's why I insist in explaining why, although I shouldn't insist at all, if you are choosing one bow or another it's because you are hearing something.

Police and insurance companies do this all the time when figuring out if a bunch of witnesses are making up a story or not, specially when it involves receiving money from insurance companies or when it involves a crime. Witnesses will explain, isolated from others, the story, and answer questions. God, the questions, most of them are very, very good and revealing, with a lot of thought behind them.
If they all claim to have seen what happened, testimonies must tell the same story, with some tiny little unknowns. It's when you start hearing a lot of "I don't know" when something smells bad, or when directly the testimonies are not compatible at all. Sure, all witnesses can talk right after the accident and believe a canonical story, but it's really hard to make up a perfect story of the events, and the questions will be designed to try to reveal the loose points of the story.

I don't know why you make it easier with a dice and other things. Your experiment is kind of blind but still you are handling unnecessary information, hahaha. The probability to randomly pass your test is 10%, as I understand you have to guess just 1 number from 0 to 9, way, way, way higher than my experiment, and as "hard" to do as mine. The probability of my experiment to randomly pass it with all correct answers, is, if my stats knowledge is not forgotten, less than 0.002%.

June 21, 2019, 11:54 AM · Here we are:

June 21, 2019, 12:40 PM · From the first video, my preference was: 2, 4, 1, 3, with 2 and 4 being pretty close.

I'll be really interested in hearing the results revealed - especially if it turns out that bow #3 is NOT the CF bow as we all seem to be assuming...

Edited: June 21, 2019, 1:02 PM · Listening to it with my PC speakers (less accuracy than with my headphones, but whatever). I absolutely can't say I heard any kind of difference in sound, only in your performances (3rd time you rushed like hell, hahaha), let alone notice a difference from previous strokes and observe that difference remains in the rest of the strokes. Heavy A-B test required for me to try to "guess", or should I say make up, a number, and even then, I would be very, very distracted by each different performance. After a lot of A-B testing, I'd get used and recognize some mistakes to differentiate between stroke 1 and stroke 3, in example, being the 3rd faster.

Your video title suggested me that you should compose a bow concerto or something, hahahaha.

Nate's Bow Concerto No. 2 in B? major, Op. 85, pour solo bowino et cordes.

June 21, 2019, 1:00 PM · Sorry Paul, but you have demonstrated that your opinion has no credibility whatsoever.
June 21, 2019, 1:06 PM · @ M. D.

:D I will keep the answer secret for a little while longer in case some other people want to contribute guesses. I am also enjoying this a lot.

@ Paul N.

Hahaha, yes, you're right, I noticed myself as I got more used to the procedure I started moving through it more quickly. That's why towards the end the strokes start getting long again. Before I reveal the answer, I hope some others will submit their guesses for video no. 2. If a lot of other people can tell the right answer, then it may be that your headphones aren't very good, or your listening skills need to improve, or something along those lines.

Edited: June 21, 2019, 2:29 PM · Where, and how, Lyndon?
And please, instead of simply "insulting" without giving any reasons, and creating noise, do your guesses. In the first one you simply said the 3rd bow was CF, yet you didn't guess the others. Now you enter simply to call me things and don't even participate in the test.
Could it be you're dying inside because you can't tell?

And these tests are "so easy" compared to a properly made one, with changes in the bow, or not, and no reference in the performance mistakes.

Nate, yes, and of course, I didn't thought about you, I mean, the violinist. Of course, that's a very good point, the violinist can also change unconsciously his/her behavior during the test, not directly affecting the output, but changing something that should not change, like the tempo. It's very clear, if you repeat an action like this, many times, that your mind gets "tired" of the action and start to speed things up. A metronome with in-ear headphones or something should help you.

About my hearing, of course, I accept and I'm totally fine with the conclusion that there are some people that can clearly pass my test and prove they hear the difference between bows. That would mean me, and many others, are not capable of doing so.

Please, don't reveal the results any time soon, many people had the time to answer to me many times how wrong I am, and ignorant, not to know that bows make a difference in sound, my God. They should come here and proudly explain how there's a difference, nailing your 2 tests, but specially mine if you do it.

Even though your tests are not really blind because you're giving away information, and they are kind of bad from a stats point of view.
For example, the last one test should be passed at least by a 10% of the people that do the test, even if we are all randomly guessing. If we are all lying and can't hear any difference (difference that remains during the rest of the bow strokes and you should notice that as well), even then, a big 10% of people will pass, and that's a lot, a whole lot.

If you are trying to prove something wrong, you need to design a test that is right (as mine, there no cheats or obstacles for the guessers) and that proves almost 100% wrong, or right. If you want to prove telepathy is fake, and you say "guess where's the dice between these 10 cubes", and end up with just a 90% of accuracy, that means the experiment is really weak. A good 99.9% solidly proves something right or wrong.

Your first experiment, is not a good example neither, because we only have to put 4 shocks in their respective box. That means, almost 5% of the people randomly guessing your bows will correctly answer all of them.

A 10% and a 5%, those are very high numbers. That's not all, those percentages are the number of people passing everything 100% correctly. The number of people correctly guessing randomly just 2 bows in your first test is higher than 10%, so "I only failed 2 bows, 50% correct, not that bad" is an expected answer that can lead to a very wrong conclusion.

June 21, 2019, 1:19 PM · How could I possibly guess the three pernambuco bows when I know next to nothing about them??? They all sound decent, there is that good enough for you.
Edited: June 21, 2019, 3:55 PM · Yeah, that's a bad designed experiment, it really doesn't lead to any solid conclusion, just like the TwoSet experiment.

But I'm upset, Lyndon, they "all sound decent"?
Is that all you have to say about all the changes you noticed in the violin sound?
That looks like something my 7 years old sister with zero contact with even the sound of classical music could say.


The second one should be easy as pie for you!
You only need to tell us when you noticed the sound clearly changed due to the bow. Please, don't stick just to the number, explain to us how the sound changes, let us know what you hear. Notice your answer also leads to the conclusion that you heard that change in sound permanent in the rest of the strokes, if any were left. You even have a big 10% of probability to guess correctly. Not that you need such a gigantic number that invalids the experiment if we want a solid conclusion, but you should nail it.

Nate, now that you're all in, like in Poker, you should do the test I explained. Choose the CF bow and the best wood bow you have, and create a 16 number list, throwing a dice 16 times. If the number is 1, 2 or 3, you play with CF, if it's 4, 5 or 6, you play with wood. Follow that list, and for the sake of the video, blur it, or focus on the floor or roof. For the sake if it, 4 strokes in each string, to test all strings, although it shouldn't have to be necessary at all. Repeating this experiment 2 or 3 times (different from each other) and scoring extreme values like 15-16/16 or 0-1/16 in all of them would certainly prove you can totally hear something changed, and that CF bow "sounds" different.

Lyndon and all the people that "insulted" me should nail it.

June 21, 2019, 2:38 PM · Hi Paul,

I think it is possible that you don't fully understand the mathematics of how the data collected from tests such as this are to be analyzed. It's OK that some people might get the right answer by guessing - we depend on that fact to make predictions and draw conclusions.

It is wrong to think about it this way: "if people can't tell the difference then they shouldn't be able to get the right answer by guessing."

The right way to think is: "if everyone is guessing then we should see the same number of correct answers as we would if we used a fair random number generator."

It will not take very many people who are not guessing to skew that number up, even in a test with a small sample such as ours. If, for example, just 4 people get the right answer, that should convince you that at least some of them are not guessing, unless it turns out that 50 people came to take the test!

June 21, 2019, 2:56 PM · LOL! OK Paul, I give!

Edited: June 21, 2019, 3:41 PM · Ever since I began lurking here, I've been noticing Lyndon Taylor frequently making personal attacks and snide remarks whenever someone disagrees with him. Just ignore his toxic comments. Wish there was a ban function like reddit...
Edited: June 21, 2019, 4:36 PM · IMPORTANT!!!
Let's clear things up:

Premise: violinists can hear a difference when swapping bows. Not telling which bow is which, but simply hearing a difference, hence, knowing when it's the same bow playing and when it's the other. Notice how that's even easier that recognizing if it's a CF bow.

Objective of the experiment: prove the premise RIGHT or WRONG.

Answers after each stroke: same or different. The first answer can be both, it's your reference.

Results of the experiment: scoring extreme results such as 1/16 or 15/16 would mean you are noticing a difference. Scoring between 3/16 to 13/16 proves you can't tell the difference, because the starting point is 8/16.

Guessing: estimate or conclude something without sufficient information.

Since we DO have all the information available for the experiment, which is sound, the correct word is not guessing, but doubt, uncertainty, confusion, not having idea.

Antonyms of guessing? Truth, fact, knowledge, reality.

So, yeah, someone that "guesses" one bow, counts as "I can't tell the difference, hence I'm proving the premise is wrong in my case". People "guessing", or I should say making things up, have the same veracity as people using a dice to determine which bow select.

But then, how we differentiate guessers that have no clue but do it correctly, from violinists that are actually hearing the difference and can explain why?

The experiment and your score. A dice will obtain near 50% of the answers correct, while a violinist proving the premise right should have near 100% of the answers correct, or near 0% of the answer correct (same and different swapped).

My experiment is kind of like telling heads or tails, if you want to prove you can see the future, you better obtain 100% of the answers correct, which is almost impossible for 16 events if you randomly guess, because guessing 50% correct in 10 events is quite easy.

End of the story: get 0-1/16 or 15-16/16 to prove you can hear a difference. If you get 80% right means absolutely nothing, because, I know this sounds shocking but it's the truth, since there are only 2 bows, a clueless person, a dice, should get exactly 50% of the answers right, after repeating the experiment many times. We are trying to prove a if we hear a difference, not if we are able to recognize how a CF sounds. Looks similar but it's not at all the same.

Edited: June 21, 2019, 4:28 PM · Read my previous message as it is very important:

If we are just testing if we can tell the difference in sound when we swap bows, then the answer of the events is not CF or Wood, but Different or Same.
That means that it's equally hard to score 15 right or 15 wrong, because you would be right in both cases about knowing when it's the same bow and when it's the other. Scoring anything between 3-14 would mean absolutely nothing, you can't tell the difference, because statistically, our starting point would be 8/16, not 0/16. The more you move away from 8, the more idea you have.

IF we are testing if you can tell which one is CF, then of course the answer must be CF or Wood and only scoring 15 would be valid to determine you know the difference.

I hope it clears things up. You know, proper work requires paper, pencil, thinking and time. I put some of those in things I'm interested in, but I needed more here, sorry for the confusion.

Edited: June 21, 2019, 5:54 PM · Hi Paul,

Sure, I agree with your clarification; we want to know not if the listener can identify specifically which bows are used, but merely whether or not they can consistently detect the change when it happens. Really, any of the (now 3!) videos I made can be used to successfully demonstrate this, given enough participants. But since we're now talking about tests rather than bows...

All you really need to do is randomly generate list, say 10, of these tuples:

(CF, CF)
(CF, W)
(W, W)
(W, CF)

(Just for example, it could also be CHEAP BOW vs. EXPENSIVE BOW or some similar token system, maybe just "BOW A" and "BOW B")

After each pair the listener answers "same," "different," or "can't tell." If the answer is "same" for (W, W) or (CF, CF) the listener gets a point. If the listener answers "different" for (W, CF) or (CF, W) the listener gets a point. The listener gets half a point for "I don't know," and no points for being wrong.

This not only allows for tracking different skill levels, it also allows for the listeners' varying level of confidence in their own abilities.

(Why yes, I am a big fan of Daniel Kahneman, why do you ask? ;)

Basically, this is just the same as teaching a new student to hear high vs. low pitch. You play two notes on the keyboard and ask the question: "Was the second note higher or lower?" At first they can't tell or they get it wrong a lot; as they practice they get better and better until they never miss. Eventually it's automatic, just like a language.

Edited: June 21, 2019, 7:51 PM · Exactly, we are testing something easier, that is if we can tell if it's bow A or bow B. I'll only let do the hard test to people like Lyndon that can clearly tell how a CF bow sounds, at least compared to pernambuco. He can say CF or wood instead of A or B.

Yeah, I guess your tuplets would work as well?
You know, statistics is one hell of a subject, full of apparently logic deductions but so, so, so easy to make a mistake. So I was thinking what would be the difference between your test and mine.

I don't understand your points awards though. Why not being able to tell the difference gives you points, nothing less than the 50% of a correct answer?
What's the difference between being wrong and having no clue of the answer?

If you don't know the answer you should be honest and retire, because the experiment is exactly trying to prove that you can't tell the difference, that you don't know the answer. You would be guessing based on nothing.

According to your point system, a person that "doesn't know" 5 answers, hits 3 and fails 2, would get a 5.5/10. Pass? For such a clueless user, 5.5 is a huge punctuation. It should be more like 0.5/10.

We are looking for consistency, not luck or "I kind of tell them apart but not really". A Pass! should be someone that clearly hits 99.99% in a experiment like this, with only 2 options to choose from.

Another story would be if we were using 4 different bows, but since I want to make it very simple and to the point, nothing more to the point than CF vs wood, we are only limited to 2 options, hence the number of hits must be very, very close to 100%.

Just think about the heads and tails example. Imagine I say I can see the future, thing that it's completely false. I'm right now gonna "GUESS" 10 head or tails based on nothing, and compare that with a random generated seq from an online webpage:
My guess: H H T H T H H T H T
Web page: H H T T T H H T H H

Look, I hit 8, 8/10! And I swear this was pure randomness, both sequences. See why I insist you really must nail every event? And retire when you don't know?
What would an expert think if after doing the test, obtains 8/10 right answers, or 14/16?
They would mistakenly induce they really know the answer and that can tell perfectly the difference, with only 2 bad guesses, nothing important.

And I also insist in the explanation of the WHY. If one expert says different because the first sounded dull, dead and dark, and another expert says different because the first one sounded more open, wide, harsh and bright, even if they are both correct, they are clueless because the reasons tell opposite facts, so they both can't be right, so they guessed based on nothing. Both failed.

If you're testing if violinists can tell apart a 440Hz A and a 443Hz, one says different because the first was slightly sharp, and other says different because the second was slightly sharp, what's the conclusion?
They have no idea and can't tell them apart. This is a bad example because one is actually right. But since we're talking about sound, that's way more obscure, the reasons really matter. Why? Because if we compare the reasons of the only 2 out the 50 users that hit 10/10, and the reasons only tell the same story in 2 of them, and in 8 of them there are contradictions, that proves one got lucky and lied, or both got lucky and lied.

Edited: June 21, 2019, 10:28 PM · Hi Paul, those are good questions. Here are some things:

First, even an experienced listener will experience hearing fatigue when performing tasks that require this level of concentration, and every person taking the test will experience cognitive depletion. For this reason the test should not be too long - 7 to 10 "questions" max. Your test, with 16, is too long. In fact, just for fun, I took your test on video 3 that I recorded earlier. About halfway through, sure enough, I experienced a concentration lapse, and wasn't sure whether I was on no 9 or no 10. As a result, when I checked my score, I found that I aced the first 8 in a row, then switched the next two (i.e., I detected the change, but I had the CF bow as the real bow and the real bow as the CF bow) and then went on to make two errors, fair and square (which I put down to having already practiced for 5 hours and thought about this *waaay too much* today).

Second, when actually implementing tests of this sort, there is a psychological element involved. Committed test-takers will experience more mental strain if they are put in a position where they feel they cannot answer honestly. If they really can't tell, they should not be forced to choose "same" or "different." Providing a "can't tell" option means that the results will more accurately reflect the subjects' true abilities, thus increasing our confidence in the reliability of the "same" / "different" answers and minimizing the occurrence of random guessing. People like getting points and will be more likely to take the safe half-point, thus making the "same" and "different" answers more likely to reflect their true convictions. In other words, it takes mental effort to select an answer other than the default one, and people are risk averse - they don't want to get the answer wrong. Awarding a half-point for "can't tell" establishes "can't tell" as the default answer. People will be more likely to only risk getting a zero for a wrong answer when they feel fairly certain that the answer they are giving is the correct one.

Third, to explain the scoring system more fully, 5.5 is not a "huge punctuation." ;) People will rarely score less than 5 in this system. A score of 5 is the same as randomly generated answers. 5.5 means you did just barely better than random guessing. A score of 10 indicates both competence at the task and confidence in your ability. 9.5 and below indicates the presence of either doubt or error.

Fourth, your coin-toss point about random outliers is well taken, but there are some ways to mitigate it. Most easily, run the test multiple times, with suitable rest periods between, using the same subjects, but a different set of tuples each time. Further, we should trust the good-faith of the test subjects to answer "can't tell" and not just guess. If we're worried about the honesty of the test-subjects, the solution is to provide a financial incentive. Anyone who gets all 10 right gets a $100 cash prize. Anyone who gets fewer than 10 right gets $1 per "can't tell" answer, and nothing for the other answers, right or wrong. Those amounts might need a little tweaking, but this is the method by which such tests are actually conducted by real life psychologists. Giving them points should be enough, but if points don't work, give 'em money. :)

Finally, the "explanation WHY" that you're seeking is actually worthless for this test. The part of our brain that decides what the answer is does so long before the part of our brain that creates verbal explanations about it can supply one. It really doesn't matter if the stories the experts tell about their decisions match or not. The only thing that matters to the test is whether or not the decisions consistently correlate with the tuples. This is also true of the 440 vs. 443 pitch example you gave. The data of your hypothetical test indicate the opposite of the point you were trying to make - both your (fictional) violinists can tell that 440 and 443 are not the same pitch. In fact, in real life, it is much easier to tell that two pitches are different than it is to identify which of the two is higher, as anyone who has had to teach a beginning student can tell you. :)

I encourage you to stop thinking about this as a binary (pass/fail) test of a single person's ability. We're not trying to determine whether or not Superman can fly. Think of it more like an adaptive pitch test:

There will always be a range of results, dependent on several factors, including experience, hearing sensitivity, and test difficulty. (By difficulty, I mean that some bows are more similar than others, so every pair of bows will provide the test with a different degree of difficulty.)

Edited: June 21, 2019, 10:37 PM · Hahaha, this is endless, but funny, hahahaha.

"Performing tasks that require this level of concentration"
Notice how that sentence looks so innocent but already tells a lot about the experiment, according to you. You are assuming differentiating bows is a very arduous task, you are almost making a statement that determines the verdict of the experiment before doing it. Oh, we are not trying to trick anybody, quite the opposite, it's not just any bows, but one CF bow vs a wood bow. Lyndon and like 10 more users from the other post mocked me because I was talking about "a lost battle", something so obvious it was even insulting for them to be questioned. I was told it was tested every day by them, give up, I was asked about my violin, bow, bow shopping experiences... because that's relevant, right?
I answered all the questions without hesitation. So, no, their clear point was that in fact bows changed so drastically the sound and it was dumb to even question that.

Second, about the number of events. No, sorry, 7 events or 10 events are very, very short, specially if there are only 2 possible answers. The more events, the more accurate and the more difficult for lucky users to pass it. About 16 events is a nice number, and you are expected to hit 15/16 at least. It's the only way since I want to make it as easy as possible for the "experts".

Hahaha, you misunderstood my experiment though. You were supposed to proceed this way: make a 16 list that determine the bow, and then play x4 G, x4 D, x4 A and x4 E. Simple, constant open string bow strokes. That would mean that you could do all the G strokes with the same bow if the 4 first bows in the list are the same one. I think you did a full run of open strings, hahahaha. But doesn't matter, although the more complex the exercise, the less consistent your performances will be. You could be interfering positively or negatively, I don't know, but surely, a constant bow stoke wouldn't accept excuses, and that's what I was aiming for.

"If they really can't tell, they should not be forced to choose "same" or "different" "
Exactly, if they can't tell, the premise is proved wrong. You are literally questioning the exact premise we want to prove right or wrong. If they can't tell, they can't tell. Premise proved. You can't be trusted for the rest of the results, specially in this very easy test. Because, if we start to get passes on this, how do we explain it when they are right or wrong?
Don't you see why reasoning each answer is crucial to conclude if the listener is basing the answer in what is the sound saying, instead of a divine intervention?
It's very easy: they can't explain why it really happens. OK, doesn't matter, I can't explain neither why magnets attract each other, but I can prove to you that they do. But then you're trying to say that the reason is not important neither, and should not be compared afterwards.
So, you are trying to say that we should believe something they can't explain, which is fine, but they can't neither explain why they choose one answer or another. Really?
That's the same as saying "coherence in the reasons doesn't matter". So, how are they telling one from the other, divine intervention again?

What do we do if we start watching the testers give reasons, for the same correct answer, completely opposed? That means nothing to you?
Of course it means something: inconsistency. It's precisely this inconsistency the one that lowers the score to almost 0/16 if it's clear you can't really tell them apart, instead of the statistic 8/16 a dice gets. A dice should get exactly 0/16, by definition. Why? How? Precisely because it has no reason to chose one bow or another, it's the reasons what differentiates you from a clueless dice. You give no reason or explanation to tell us what made you choose one option or another, then you're acting like a clueless dice, literally.
Unless, you say that each violinist interprets brightness, metallic... on its own. Then it's all valid and for me this violin sounds like a strad.

"it is much easier to tell that two pitches are different than it is to identify which of the two is higher"
Yes, definitely, when the pitches are so close. So, what does that tells you?
It tells you that you can clearly differentiate the sounds, not the pitches, you can't say if one is higher or lower, meaning you've reach your limits of distinguishing pitches. The premise was "how many Hz can you tell apart?", the experiment was testing you ability to analyze pitches, not tiny little differences in sounds. When you can't tell which one is higher or lower, game over, you're not basing your opinion in the Hz's anymore, end of the experiment. That's what invalidates a right, because the reasons were completely random and not based on the abilities we are testing.

Edited: June 22, 2019, 12:17 AM · Hey Paul,

Remember, I'm on your side, more or less. You are not dumb to question. I don't reject your position outright, I'm saying it needs some tweaking. In the realm of of generic playing on open strings there are perceptible differences from bow to bow, but not drastic. Certainly not such that it's similar to the difference between, say, tuba and clarinet, or such that, given two decent bows, an expert will always be able to tell the difference without error. I'm 100% opposed to the idea that somehow only pernambuco is eligible to give "life" to the violin sound. In fact, in my collection of around 10 bows I have a snakewood bow and an IPE wood bow in addition to the CF one.

When it comes to the number of events, you are simply incorrect. If you study cognitive psychology more, you will learn more about it. In this context, 16 is a too large a number for most human brains.

About the experiment, yes, I misunderstood you. I did wonder about that. It seemed to me that it would be better to do the entire test on just one string, maybe repeating the whole test once for each string.

You said: "What do we do if we start watching the testers give reasons, for the same correct answer, completely opposed? That means nothing to you?"

Yes, that is correct. Any reasons that might be supplied by the test subjects are unrelated to their ability to select the correct answer or not. That is really the whole point of testing: to escape from natural language descriptions of things. We don't care if the subjects understand why, and in fact, if we are scientists, we don't want to know what their explanation is. We only care about whether or not they can tell the difference; not how they explain why they can tell the difference.

P. S. I was going to say something about this:

"If they really can't tell, they should not be forced to choose "same" or "different" "
Exactly, if they can't tell, the premise is proved wrong. You are literally questioning the exact premise we want to prove right or wrong. If they can't tell, they can't tell.

In my previous post I made a comment about testing to see whether or not Superman can fly. This is exactly what I meant by that. You seem to conceive of this problem as though we are trying to locate a superpower that is always on, and that if you ever discover one mistake you have proved that the superpower does not exist. Real life is just not like this. It is a simple fact that human beings encounter moments of doubt - even experts - and it is perfectly legitimate for them to do so. In order for a test to be at all useful it must account for the relevant experiences that the subjects are reasonably likely to encounter. That is why in the adaptive pitch test I linked to earlier a wrong answer simply makes the test slightly easier and goes on to the next question. It does not end the test immediately with the answer "SORRY, YOU ARE DEAF!"

Edited: June 22, 2019, 5:29 AM · In order to do this experiment to obtain valid results, one would need to create a machine that draws each bow across the strings at the precise point, speed, and pressure of the other bows. The device would need to be able to vary these parameters, and the digital recordings would need to be identically made.

You would also need to control for hair - number and quality of the hairs, hair tension, quantity and quality of rosin, humidity, and who performed the re-hair.

Finally, the identically-recorded results would need to be analyzed by an algorithm that can compare multiple digital acoustic fingerprints to determine if there is a statistical difference between the sounds of the bows.

Human beings are incapable of performing this experiment or evaluating the results. The mere fact that human ears hear different and limited frequency ranges makes it impossible. What might be "different" to one person might be physically imperceptible to another.

So this is another "how many angels fit on the head of a pin" debate.

What is not debatable is the fact that many people believe that different bows sound differently on different violins. This is my own personal experience.

Edited: June 22, 2019, 7:36 AM · Good morning!

I think it's about time to reveal the answers to the first video. I will delay a little bit to give anyone who wants to the opportunity to chime in. Last chance to spot the carbon fibre bow! :)

For your convenience, since this thread has waxed long, it's this video, from the top post:

June 22, 2019, 7:50 AM · Hi George,

Of course, such machines have been made, and similar experiments to the one you describe have been conducted. But I think you may have missed the point. We're not talking about a comparison of the physical properties of bows, we're talking about the limits of human perception. You can't investigate human perception without humans. :) When you say "What might be "different" to one person might be physically imperceptible to another" you are quite right, but that's the main feature, not a bug!

Try this analogy:

I can easily hear difference in pitch between 3Hz and 1.5Hz, and with a little effort down to about just under 1Hz.

Some people can reliably hear difference in pitch at .75Hz and below, but to me a lot of those pitches are indistinguishable - it sounds like the same note twice.

Suppose I don't believe anyone who says they can distinguish a difference of, say, .2Hz, and accuse them of claiming to have a super power. How could they prove to me that they really can hear a difference?

Edited: June 22, 2019, 8:22 AM · Yeah, sorry if I was, too aggressive? I don't know if that's the word. I'll try to express the ideas in a more simple and relaxed way, I think.

1. Sure, there are differences way more obvious, like telling apart a tuba sound vs a violin sound, but the premise is "we can tell bows apart". By the way they have spoken to me, it was so obvious, so it's not at all a valid answer or excuse to say "but it's very, very hard". No, it is easy, simple and obvious, because each bow make the violin sound one way or another, and in CF vs wood, it should be a screaming difference. They have not told me "well, Paul, it affects sound but it's really, really difficult to hear the differences, they are so subtle or invisible for most". They have said "I have 3 bows and I can tell so easily, the difference is so obvious, give up, one makes the violin sound harsh, bright, the other make it darker and dull..."
Besides all of this, I'm making it as easy as possible: only 2 options, and the most obvious ones, which is CF or wood. It really can't get any more simple and easier than that, literally.

2. Number of events: let's do an example. The sequence given by a dice tells you:
So, you will play an open G string with CF W CF CF, with a pause between each other, leaving the bow in a table and picking up the next one, then the D string all CF, and so on.
Is that exhausting?
Sorry, no. If you claim to recognize so clearly each bow, that's as simple as the test can get. I even change strings, 4 different pitches, what else to wake up your senses?
A person claiming to have perfect pitch, which to us can sound false and made up, will pass a 16 note test so, so easily. May be, may be, in a 1000 number of events, that person fails 1 or 2 notes, or not, mostly not.

3. "Any reasons that might be supplied by the test subjects are unrelated to their ability to select the correct answer or not. That is really the whole point of testing: to escape from natural language descriptions of things"
I really don't understand your reasoning there. It's so, so confusing how you believe that reasoning your choices does not matter. Reasoning your choices is what makes you capable of doing this test. It's whats differentiates you from a dice, that's a fact. If you have no reason to explain why the sound changes, and you neither can explain why you choose one bow or another, sorry, I'm 100% forced to believe you're making up your choices like a random rolling dice, and that your hits and misses during the experiment are simply a matter of accident.
There's a very good quote that says "I can only hear the differences if you explain them to me". AKA Placebo. If I present to you 2 identical samples and tell you notice how in the last one, played with a $500000 bow, the sound is more open and bright, you will surely "hear" that.

It's only when we check that the reasoning is compatible between people that passes the test when we can conclude that, effectively, these people are basing their judgment in what we are testing, which is sound (that's why I request them to describe the differences in sound), and not anything else like patterns in the performance, which we reduce them as much as we can, but they are still there, in example, you could unconsciously be playing, with perceptible more airiness, a light bow.

4. "You seem to conceive of this problem as though we are trying to locate a superpower that is always on, and that if you ever discover one mistake you have proved that the superpower does not exist"

In this test, with only 2 bows?
Yes, one mistake is a huge deal.

If you feel more comfortable, we can throw all 4 bows in and tell them to select in which one of the strokes, if any, the CF was used. I can guarantee you there would be plenty of excuses they can shield with the moment you put 4 bows. So, to make it as easy and simple, 2 bows, hence, you better answer correctly almost 100%, given how easy it is to get one answer right randomly, and how easy it is to randomly get 50% of the punctuation right.

5. George "In order to do this experiment to obtain valid results, one would need to create a machine that draws each bow across the strings at the precise point, speed, and pressure of the other bows. The device would need to be able to vary these parameters, and the digital recordings would need to be identically made. You would also need to control for hair - number and quality of the hairs, hair tension, quantity and quality of rosin, humidity, and who performed the re-hair"

Yes, and no. If all these people that told me I had no clue about what I was talking about, have to rely on all that machinery and control parameters to be able to tell them apart, then that's quite the verdict of the experiment. Why?
Because it would mean that the slightest change in speed, pressure, angle or hair grip would confuse them and couldn't even tell a CF bow from a wood bow.

In other words: "Yeah, I can totally differentiate the sound created by a black bow and a brown bow, but only if I go to a lab with $1000000 worth machinery that plays them exactly the same way, under NASA lab accuracy"

Yeah, then one would ask how they were able to tell them apart when a violinist was playing.

Nate, nobody is participating. I suspect many of the users that mocked me and called me ignorant have seen and read this thread, but they don't dare to test their beliefs. Look at Lyndon, he didn't answer the test 2, he limited to say 1 number between 1-4 from your first test, 25%. It's clear he fears to be completely wrong, specially after mocking me. Imagine he can't tell the difference, how chaotic. So, a data revealing now, with only 3 or 4 participants, is useless.
I am also taking for granted that they are not analyzing the sound with a PC or anything. In example, if we are testing if people here can tell apart +0.1Hz, it would be so easy to use a pitch analyzer and get all correct answers.

Edited: June 22, 2019, 8:35 AM · Hi Paul,

[Edit again! I copied a different thing than what I was actually replying to!]
1. They have said "I have 3 bows and I can tell so easily, the difference is so obvious, give up, one makes the violin sound harsh, bright, the other make it darker and dull..."

Right, but *I* have not told you that. I have told you that sometimes the difference is big (like between a heavy viola bow and a light French bow) and sometimes the difference is small, and that some people are better at detecting such differences than others, due to naturally good hearing, skill, and the like.

2. Right, I understand your test now. I'm not saying that the test is exhausting for me to create. It's not exhausting to play open strings 16 times with a few different bows! ;) I'm saying that the design of the test will taint the data by causing the test subjects to make unwarranted errors due to cognitive depletion. This is not a simple topic, but here is a quote to illustrate the principle:

"A disturbing demonstration of depletion effects in judgment was recently reported in the Proceedings of the National Academy of Sciences. The unwitting participants in the study were eight parole judges in Israel. They spend entire days reviewing applications for parole. The cases are presented in random order, and the judges spend little time on each one, an average of 6 minutes. (The default decision is denial of parole; only 35% of requests are approved. The exact time of each decision is recorded, and the times of the judges’ three food breaks–morning break, lunch, and afternoon break–during the day are recorded as well.) The authors of the study plotted the proportion of approved requests against the time since the last food break. The proportion spikes after each meal, when about 65% of requests are granted. During the two hours or so until the judges’ next feeding, the approval rate drops steadily, to about zero just before the meal. As you might expect, this is an unwelcome result and the authors carefully checked many alternative explanations." - from "Thinking Fast and Slow" by Daniel Kahneman

3. " It's so, so confusing how you believe that reasoning your choices does not matter. "

Yes, I understand that; it's confusing for a lot of people the first time they learn about this. We humans are very committed to the idea that we are rational beings; that our reasons precede our decisions. The (somewhat simplified) truth is that our reasons are a story we tell ourselves about our decisions afterwards so that we can feel like the world makes sense. Don't misunderstand, I'm not saying that the reasons are unimportant. How we describe things is critical for sharing information, getting along with each other, and and catching errors made by our unconscious decision-making machinery so that we don't act on them. However, for *this test,* not only are reasons unimportant, they are harmful to the data. We're trying to escape from the stories people tell and find out what really happens.

I'm not sure I can explain it to you convincingly within the confines of this forum discussion. You might just have to trust me, or agree to disagree. I would recommend that you read the book I quoted from earlier, "Thinking Fast and Slow" by Daniel Kahneman. I think you would really enjoy it, and it would help you be a better thinker about topics such as the one that we are discussing, which you are clearly very interested in.

Edit, to reply to your edit :)

I haven't been posting here very long, but I've been a lurker for many years. To be honest, I was surprised that Lyndon even ventured to try and spot the CF bow in video #1. Lyndon is not really my target audience. The youtube video has over 90 views; I suspect there are a number of people who are not actively posting who are still interested in the investigation and will want to know the results. I just wanted to make sure to give, for example, people who only read on the weekends, a chance to chime in. I'm going to go make some breakfast and then I will REVEAL THE BOWS DUN DUN DUNNN

Edited: June 22, 2019, 8:48 AM · Paul N.

There is a difference between saying something is true and proving it is true.

The very same bow can sound vastly "different" depending on how it was re-haired.

Like I said, this is another "how many angels fit on the head of a pin" debate.

Nate B.

We are also talking about whether differences are real or imagined. :-)

Lyndon T.

Your blanket dismissal of CF bows is absurd. Many CF bows can sound great and perform superbly.

Edited: June 22, 2019, 9:23 AM · OK, so a very simple question then. I'm open to be wrong, that's for sure, but here it is.

Well, first, let's clear this: when we are told the sound is more open, brighter, darker, sweeter... although they are not scientific terms, we do understand them, right? We assent when our teacher explain why this part or that part must be sweeter, louder, brighter...
Even for emotions, that are so subjective, most people agree when a piece is sad, happy, fun...

So, the question is: you are testing if a bunch of violinists can tell a +0.1Hz difference. Out of 100, 86 retired due to honestly being unable to tell if one pitch is higher, lower, or the same. They knew they were just guessing without any real conviction about their choices. Then 12 failed the test so bad, and only 2 got all the questions correctly. One of them hit each answer and correctly told which pitches were higher, if it is that they were, because some were the same. The other one, got 80% of the different pitches wrong, that is, he said the lower pitch was the higher one.

Tell me, would you determine that this second violinist is able to recognize +0.1Hz differences? That is, recognizing the height of a note, don't mix that with recognizing that the sound was different.

Which pitch is higher?
-The second one (choosing an option)
OK, you are listening that second is higher, right?
-No, I don't know (unable to reason why)
You don't know?
-No, but they are different, that's for sure
OK, so what made you conclude that?
-I don't know

For me, and I believe for anybody, that means that this violinist is unable to distinguish pitch, but can tell differences in sound. He's not using pitch as a reference, which is what we want to test. It would be really weird, because imagine a violinist like this passes all the time the test, just like the one that passes the test because he or she knows correctly the answer.
It would be a really weird case, probably an error in our experiment, because we are just changing the pitch of a sound, the Hz, nothing else, and yet one can pass the test without using this variable that changes as a reference, because he all the time mixes highs and lows, so it's clear he is not using that as a reference.

So, see? From my point of view, the reason would be fundamental to even pop up errors in our experiment. If you wasn't counting the reasons, then this violinist that somehow passes the test without using the apparent only variable that's changing as a reference, would be determined to be able to use changes in pitch. As I say, it's a real weird situation because it doesn't make sense, you totally must base your reason in the only variable that's changing. If you pass test and test without using this variable, then the experiment is changing something more than the pitch.

If you are not reasoning what makes you select one option or another, you are a dice, and dices get 0 points because precisely they can't reason.

June 22, 2019, 8:55 AM · Paul,

Statistics are useful for large sample sizes (MANY people guessing the bows) when the detailed physics of the causes and effects are not known.

Your objections are based on conclusions drawn from one or two individual guesses. That is not how statistics are properly used.

The default position is that the bow does not matter in producing tone. That is called the Null Hypothesis.

Have a bunch of people observe the test and see if they can discern a difference among bows. If, as an entire group, they cannot distinguish tone difference among bows on average with greater than random chance probability (called the confidence level), then we have no statistical reason to reject the Null Hypothesis.

An individual successfully distinguishing the bows among all bow trials may be because they actually can, or they got lucky. We have no way of being sure without designing additional tests.

Like the infamous Paris experiment, there may be experts that can tell the difference between well regarded new and old violins, and specifically identify which are old Italian and which are not.

But statistically, the chance of that happening was no better than flipping a coin.

We can argue if the videos represent valid scientific randomized tests, but it still is fun to participate if they are not.

My 2 cents for video 1: 1 and 3 sounded very similar, as did 2 and 4. I thought 2 and 4 created a brighter sound, that is, more high frequency content.

Edited: June 22, 2019, 9:30 AM · Hi Paul,

"Well, first, let's clear this: when we are told the sound is more open, brighter, darker, sweeter... although they are not scientific terms, we do understand them, right? We assent when our teacher explain why this part or that part must be sweeter, louder, brighter..."

The truth is, I really do not. I think the language we use to describe sound is next to useless and presents a real impediment to any kind of objectivity when, for example, shopping for violins. Over the years people have said things to me like "oh your violin has such a great dark chocolaty tone, I bet it would really like STRING BRAND X." Or, "your violin has such a nice bright sizzle." Or "your engine has a really great growl on the G string." Or "your G string sounds kind of fuzzy and covered like your violin has a cold." Now of course, violins sound different on different days, but I play a pretty solid modern instrument with modern strings. I really have no idea what to make of linguistic descriptions of sound. When I try violins at a shop and the luthier says "do you prefer a light or a dark sound" I don't really know what to say. And when he says "here, try these ones, they have a great dark tone!" I never know what to expect, and when I play the violins they seem to me to be no way similar to each other. And yet we all talk about sound using words like "dark," "light," "closed," "open," "broad," "sweet," concentrated," etc., and there are websites like Peter Zaret's devoted to language of describing sound, as though we are talking about some kind of universal collective perception. This is why, a long time ago in a thread far far away, I said to James T. that I would love to see more scientific descriptions of our instruments and how they work.

"Tell me, would you determine that this second violinist is able to recognize +0.1Hz differences?"

In short, yes. The question "are the two different pitches different" is a separate, and much easier, question from "is the second pitch higher or lower?" Further, at least for me, "second pitch is lower" is much easier to identify than "second pitch is higher," at very narrow differences, say .5Hz or less.

Edit, once again, to reply to your edit:

The problem you are having is that you are only considering two possibilities: either your brain is a random generator, or your brain is a list processor. But what we are really talking about is a sensor: a particular stimulus is received, and an alarm goes off in your brain. If the alarm goes off, it means you can hear that the two pitches are not the same. It doesn't matter if you can explain that one or the other is higher.

June 22, 2019, 9:23 AM · The answers! Spoiler alert, Carbon Fibre was bow #2. :)

Bow 1 was early 20th century Hill by Albert Leeson
Bow 2 was Shar Presto CF
Bow 3 was Contemporary bow by Pierre-Yves Fuchs
Bow 4 was fine old bow stamped KITTEL

Edited: June 22, 2019, 9:45 AM · Thank you Nate!
OH! I knew it, I was suspecting something about the dread, more bright sound of number 2 bow. Yeah, now I'm listening to it again, and I can clearly hear how number 2 sounds overly harsh and dead. It's kind of annoying actually.

OK, I'll stop hahahaha

Now seriously, thank you for making the test and having this conversation. You seem to be a nice guy. I'm curious about your playing corner. It looks so comfortable, sweet home and appealing, with all those great windows and natural light and flowers. I'm used to play in cabins without windows, small and with artificial light, so, yeah :(

Is that like you own single story home, typical duplex with it's own garden and area?

June 22, 2019, 9:36 AM · Fantastic.

Is the "fine old bow stamped KITTEL" actually a Kittel bow?

June 22, 2019, 9:42 AM · LOL.

OK, so we have got to a point where I think maybe we can have a more interesting discussion. It took a little longer than I thought it would, but hey...

The truth is that what we have been dissecting in such great detail is not all that interesting, from a musical perspective. Once we have used generic long bow playing to reject the obvious duds, there's no reason to keep playing like that; we're rarely called on to play in that way in real life!

For almost 200 years bow-makers have been re-creating, or slightly improving via minor tweaks, the mathematical model that Tourte created. We really should not be surprised if after all that time the most generic sound made by even a relatively low-quality modern version is not drastically different from the real thing. It might be more interesting to compare the difference between, say, a snakewood baroque sonata bow and a modern CF bow, or between an ancient ironwood bow (there's a gorgeous ironwood D. Peccatte at Carriage House right now, if someone wants to lend me $30k or so... ;) and a modern pernambuco one.

I think the right question right now is, why do players have preferences about bows at all? And, by extension, can we talk about what makes a bow good or bad using quantifiable (more/less) language? This is really what is interesting to me about bows and bow trials. I think the answer is that we can. Before I go on I want to ask you if you know what "transients" are.

June 22, 2019, 9:47 AM · @Paul Deck,

There's a chance that it is, but it has no provenance. It has a real KITTEL stamp, and it plays fabulously, but there were also some great bows made by Knopf for the KITTEL atelier. Paul Childs said it was "spurious" about 20 years ago, but that was before all the Kittel research and the Kittel book. The head looks Knopfy to me. I can upload some pictures of it, if you'd like.

June 22, 2019, 9:49 AM · @Paul N.,

Thanks, I've enjoyed this too. About my playing corner, the niceness is all due to my girlfriend and her orchids. It is our own house (well, we're renting) but we just moved into it a few months ago, so the rest of it is much less appealing, with cardboard boxes and things stacked all over the place.

Edited: June 22, 2019, 10:12 AM · Well, not a lot of people participated, only partially doing the test 1.

As I suspected, I believe we can say that what really matters when selecting a bow is how good we can control it, basically, not at all if it's from 1800 or 2010, if it's wood or CF. The differences in "sound" will be due to our great control with one bow or another, and also one should take real care when comparing bows: different hair condition is crucial. It would be really sad to reject one bow because it has old wasted hair and you feel you need to press down more in order to get the sound you want out of your violin.

If I ever go bow shopping for real, with $2000 in my pocket?
Wear blind glasses so materials and looks don't influence me, and play. Then select the most comfortable ones for my personal way of playing, and finally take off the glasses and see price tags, materials, looks... We are not robots, we also like to own gorgeous bows. I'd say for the sake of tradition I prefer wood bows, but oh man, some CF bows just look amazing, full of details and surprises.

Sad part of the story?
If between the bows we select after the process, on of them is $349 and the other 2 are $1399 and $1600, chances are we would reject without hesitation the cheap one, we would immediately look down at it, even we felt previously we could play with it in the same level of comfort and quality than the other 2.

I guess "cheap" bows in violin stores use a not so great hair, even the re-hairing could be done not that great in most of them, having mistakes, uneven re-hairing... while the most expensive ones would have the best hair possible and surely fresh and new.

I guess I would have to request that the store rehairs all the bows I'm gonna try. Here I come! Yeah, it's guaranteed the violin shop dealer would kill me, but I warned you about it several times.

Edited: June 22, 2019, 11:13 AM · Paul N.

For sure, on more than one occasion I've had a dealer rehair a bow I was trying because a bad hair-job was interfering with my ability to judge it. As for the sad part of your story, I do like beautiful things, and a couple of my bows are real lookers ;) but at least one time I did the opposite. When I got my "KITTEL" stamped bow it was up against a "real" Paul Jombar bow that was probably by Victor Fetique and almost twice as expensive. I liked the cheaper "KITTEL" better, and I bought it, and I've never been sorry I did. Over the years it's always been in my top two favorite bows from every batch I compare. It doesn't look like much but it sure can howl. :)

June 22, 2019, 1:55 PM · Wow. In retrospect I'm glad I hedged with saying that I liked 4 almost as much as 2, but there's no running from the fact that I chose the CF bow as my favorite.

Yikes. What does that say about my ear??

Edited: June 22, 2019, 3:51 PM · M. D. don't worry, even an expert as Lyndon couldn't tell a wood bow from a CF bow. Your ears are perfect, you should trust them under the right circumstances, like these. Well, not just Lyndon, but nobody. And I'm 100% sure he does the same experiment again and you would choose different favorites the second time.
It's OK, bows don't make the violin play better or worse, but you, the performer. Also the hair, the hair must be equal in amount and condition to test fairly 2 bows, or any number of bows. If not, the fact that you notice one bow slippery and uncontrollable, could be perfectly just the hair, specially when comparing it to other bows that have a perfect fresh rehaired hair, you would notice a difference in an instant, but due to the hair, not the bow, in that case.
June 22, 2019, 3:44 PM · Quite happy that the bow that I found the weakest turns out the CF bow :-)
June 22, 2019, 3:53 PM · I still believe number 3 was the CF bow!!
Edited: June 22, 2019, 4:41 PM · Don't worry Lyndon, it could be, Nate could be tricking us. Nate, did you say the correct answers, or was that message another test?
Hahahahaha, now this is getting out of hand.
Oh, no, hahahaha, I'm dying. Now I really think that Nate doesn't have anything to prove the answers he's giving and it's all a matter of faith.
June 22, 2019, 4:32 PM · muahahahahahahaha.....
June 23, 2019, 5:13 PM · The original thread was removed. Thank goodness...
Edited: June 23, 2019, 6:11 PM · You're welcome, Nuuska. My question was answered with the very first reply. Most of what came after could have powered a wind turbine farm.
June 23, 2019, 6:08 PM · Yeah, I just noticed. Feel free to do the 2 remaining tests though. Good luck!
June 23, 2019, 8:50 PM · If I want to conduct tests devised by an unqualified beginner, I'll devise them myself.
Edited: June 23, 2019, 10:28 PM · Well, thanks for trying!

It's wonderful how, except Lyndon, that bravely dared to participate and fail, all the experts in the room won't dare to guess and tell the so obvious differences a bow makes in sound. I expected that though, the moment I dared to question such a sacrilegious thing I knew I was gonna get a lot of heat, and I knew that those people wouldn't dare to question their blind beliefs, but rather insult and mock me. So, I'm good, it's understandable. Have a nice day!

June 23, 2019, 11:27 PM · There are definitely violins that pick up a nice bright pep from a well-matched CF bow. It's not true that all CF sound is bad. In some cases it can loan extra projection due to its tendency to pull out the high frequencies.
June 24, 2019, 12:54 AM · OK, like Lyndon I thought No3 might be the CF bow, but that was based on what others have said in the past rather than my own experience so actually (like everyone else..?) I was right all along! Whatever sonic differences might exist due to the material of the bow are insignificant in comparison with the many other factors that make a scientifically conducted trial practically impossible.
June 24, 2019, 3:47 AM · I really enjoyed the conversation between Paul N. and Nate B! I understand Paul's excuse for the so-called "prolixity", but have found that most of the time a higher word count just gives more individual sentences to latch onto, so there's really no winning against cherry-picking. Luckily, Nate was a very fair partner in the discussion.

Basically the main point of contention was the Nate was treating the task of differentiating between bows as a reasonably difficult one so his proposed methodology was tailored to that. He was trying to test plainly how good of a sensor an expert human ear is when it comes to detecting a difference between bows. The question of what the sensor would give as a post-hoc rationalization was irrelevant for that purpose. Though I would argue that the clarity and certainty of the post-hoc rationalization, as well as the speed at which it arrives, might also be indicative of something.

On the other hand, Paul was coming at the issue from a more specific point of view, where he was taking into account the attitudes of the people that had claimed to him the differences were obviously obvious. He was trying to devise and experiment that would minimize the amount of excuses available to subjects if their sensors proved to be less accurate than they had hoped. A slightly less pure approach, but perhaps more appropriate for the situation.

Actually, with clear enough differences I think Nate's argument about post-hoc rationalizations is less salient. For example, if subjects were tasked with differentiating between a dessert and a main course their explanations would be quite consistent with objective reality outside of their brains. They would explain that the dessert is a dessert because it's a cake. They would explain that the main course is a main course because it is a stake. In that case I think the immediate subconscious decision and the post-hoc rationalization would be quite well aligned.


The responses of some of the other members were less satisfying. So much snark and snooty rhetoric. I mean, what is it about the test that shows it is devised by an 'unqualified beginner'? How would it be improved? Or is it such a thing that it should not be tested too much at all? Why the need to delete the previous thread? Why should it be so irritating that it exists, and why come to its reincarnation only to profess your gladness that its previous incarnation is gone?

I mean, I understand the hypothetical situation where people who are professionals in a field would be annoyed by some ill-conceived objections by a casual participant, but I think sometimes the convenient role of this annoyed professional is unduly donned because of its believability and used merely as a crutch.

I must say, though I'm not proud of it, I got some satisfaction from one self-assured answer in the thread turning out to be wrong.

Edited: June 24, 2019, 7:34 AM · Thanks, Urban.

Yes, Nate was and is totally open minded, very kind because he did the tests (that's something already, huge step) and doesn't need to call me things and look down at me to disagree or agree. He reads what I think, with respect, and if he doesn't agree, he explains why.

For example, we disagree on the importance of the reasons. I forgot to mention just one very important thing about this. Assuming that the only variable that really changes between the events is the bow, only people that can notice changes in that variable could pass the test. For me, the importance of the reasons is that it really is a guarantee of the experiment, that is, it checks if the isolated variable was really isolated.

Hair is so important, it's literally the finger of a guitarist, it's like having thousands of fingers in a raw. So, a difference in hair between bows is like a guitarist playing with more nail or more flesh, it of course will change the sound. So you must have very similar hair to correctly do this experiment, or else you will hear differences due to the hair, not the bow.

I'm not devising anything, I only thought about a proper experiment. People here that disagrees with me don't need me as a supervisor to check if they are right or wrong, their guesses and the correct answers will tell them. I've been called worse things, don't worry.

After all the requirements told to me, I can assure you now that in order to develop the "bow detector", I need a $5000 violin (old italian if possible), $3500 bows, become a semi-professional (a beginners can't tell) and have been bow shopping. Yet, a non-violinist, non-musician mother can totally tell the differences. Yeah, this is what I'm dealing with.

June 24, 2019, 9:29 AM · What's obvious under the ear to a player might not be obvious to an audience member or someone listening to a recording.

A few years back here on, I asked about the issue of bow switches in Holst's "The Planets". The col legno in the "Mars" movement is pretty much a call for "use your CF bow", but there's a concertmaster's solo later where I really did not like the brightness of the timbre produced. Nathan Cole told me, just use that CF bow, nobody in the audience will notice the difference. He was right, and the recording of the performance sounds fine.

June 24, 2019, 10:06 AM · Let's keep it simple, it's enough with 2 bows, let alone introducing an orchestra. The premise of the others was clear: a bow makes the violin sound brighter, darker, sweeter, harsh, loud...
I told them that it could only be because 1, the hair is different in the bows, 2, the way they handle and control one is different form the other, or 3, they are being tricked by Placebo effect. Nothing to do with the bow itself. Then you read terms like bow sound, I'm telling you I've been to a bow maker conference and the most important thing about CF vs Wood (all hated CF) was the sound of the bow. I was mocked, so I decided to design a test, and here it is, and no one is taking it. Well, Nate did the experiment, I didn't. I could but I'm sure I would be immediately rejected, even from the first moment my main objective was to know if they can. I believe they can't at all, but hey, I designed an experiment that can prove me wrong, so I'm all open.

Now if you want to change the premise by "we can hear the difference, but only if we are the ones playing". OK, then Nate should put the microphone right over the violin and test again. But notice a mother said she could hear the difference 2 rooms away, and this I think it was Andrew that said he had tested the bows while playing them like a cello. So, your premise swap is contradictory with what they've said.

June 24, 2019, 11:14 AM · Oh we're still going!

+1 Lydia. I used to make a big deal about switching to a cheap bow for the "stroke with the wood" section of in Mahler 1. Nowadays I just use a cheap bow the whole time or stroke lightly with a good one.

Paul, I'm pretty happy with how this discussion wound down, and I'm not sure that there's anything to be gained by expanding it at this point, unless you want to talk about why we still have preferences about bows even though there is not a drastic difference in the basic sound produced from one bow to the next.

However, I will say that we shift focus from "the experience of listening from a distance" to "the experience of playing," that there is 0% chance that I would ever make a mistake about these bows. I mean, for example, if I had been blindfolded during the "all in like poker" video and a friend was deciding whether to hand me the CF bow or the wood bow, I would know which it was likely as soon as I touched it, and certainly immediately once it was on the string. There is just too big a difference between their comparative weights and stiffnesses for me to confuse them, even when fatigued.

However, that would not be true if the two bows were very similar, say, a couple of nicely machined Coda Diamonds, where the quality control is high. That would be much more difficult and possibly impossible task for me.

Edited: June 24, 2019, 5:28 PM · Paul N. wrote:
"I'm telling you I've been to a bow maker conference"

Oh golly-dang, which conference was that? Please specify, since I have attended maybe 50 such conferences over that last ten years, and don't recall having run into you.

June 24, 2019, 12:41 PM · The bow sound that makes a difference IS the sound that the player hears because that is the sound by which the music is guided to the ears of listeners.

By now I've probably played hundreds of bows - although some for only the number of seconds it took to go through my bow testing routine. I always tested 2 bows at a time and then kept one to test against the next and so on and on and on - for up to 66 in one session. I've done this with violin and cello bows - not so much with viola bows. I will admit I have never tested a bow worth more than $10,000 (although my own Voirin might be worth more had it not been damaged and repaired).

About 18 years ago I prepared an EXCEL spreadsheet "bow calculator" of almost 30 bows I had in hand for testing. What I learned from that exercise that the parameters that Paul N wants to keep as constants in his tests do not reflect the way musicians use bows. The "bow calculator" was available on line for about 12 years while I still had a website but I resurrected it again earlier this year so I could link to it here on so here it is again:

I will cite just two of the things that will not work in bow testing:

1. No bow should be haired with a standard amount of hair. Every bow should be haired with the optimum amount of hair for that stick's physical characteristics.

2. No player uses two bows in exactly the same way. Every bow is used in the best way to bring the desired performance out of that bow.

One of my violinist friends used a CODA Classic bow and later a Rolland SPICCATO bow on his Enrico Rocca violin (basically choosing to use a $600 or $1,600 CF bow on a $100,000 fiddle) in preference to his Lamy bow. He related to me one experience he had had in a string quartet when he loaned his Coda bow to the other violinist who did not like it - but then when the other violinist tried the CODA on the Rocca he could see why my friend chose it.
(So much for blind testing!)

June 24, 2019, 1:03 PM · Fascinating sheet, Andrew. Also, insane. Did you really count the individual hairs? Yikes!
June 24, 2019, 1:03 PM · On my previous violin, the tonal differences between bows was less stark, to the point where sound almost didn't factor into my choice for a bow for it. Then I tried an Arcus that Bernd sent me, and it was very harsh on my violin, but very good on a friend's dark-sounding instrument.

I just put a set of Rondos on my current violin, and they are bright to the point where they're a little too harsh under the ear. Interestingly, this has the effect of significantly diminishing the tonal difference between my CF backup bow and my usual bow. Usually the CF brings out the high frequencies. These strings are bright enough that the high frequencies are sounding regardless. The difference between the two bows is not inaudible, but it's not a big deal. In the hand, of course, the bows are radically different.

June 24, 2019, 2:11 PM · As long as we are moving toward specifics— sometimes the best bow today is not the one that was best yesterday. A change of humidity can affect not only the bow (esp hair length) but adjustment of the violin. And results surprise. I have one bow that is a little on the soft and dark side that I would normally use to make a fiddle more forgiving in cold weather. Its best moments as far as sound go during the dead of summer, when it often outperforms brighter-sounding sticks—especially on a Guarneri-model Croen, which is probably the darkest of my instruments. Even when I am traveling with only two bows, I will try each to see which is doing better that day.
June 24, 2019, 2:26 PM · Nate B, Yes I did count individual hairs and related the number of hairs to the stiffness of the stick and what seemed to me the best playing tension to achieve optimum strain in an individual (average) hair. It's all in the spreadsheet.

This was my project up to the morning of 9/11/2001. By 9am PDT that morning I had decided to stop the project and devote my time to using bows only to make music - which was too bad because my wood-working son had just made me a portable jig for taking to where the bows live to make more measurements. Jay Ifshin had given me the OK to make the measurements on bows in his shop - but I never did.

June 24, 2019, 2:52 PM · Andrew, not much to say to that except "wow!"
June 24, 2019, 3:16 PM · Conference, not conferences. It was one conference set by the "apparently" most famous bow maker locally, although then I've asked my three teachers over the years and none of them told me they had their bow rehaired by this bow maker. I didn't ask them if they knew him, simply asked to which one they went, and none was this man. The conference was about the history of the bows, how bows are made, and whatever the bow maker wanted to talk about, and of course the difference between CF and wood bows.

"The bow sound that makes a difference IS the sound that the player hears because that is the sound by which the music is guided to the ears of listeners"
OK? and I've asked you several times to hear those obvious differences with these tests and tell me where are the bow changes?

About the bow hair amount, I accept a few hairs more or less, that's fine. I guess, though I'm not sure, that it would not make a noticeable difference. Of course, hair is very, very important, it's our right hand fingers, like the fingers for a guitarist, so, they affect quite a lot the sound. Since I believe most bows use a very similar amount of hair, my point was that you can't compare bows that have big hair amount differences. Again, then you would not be comparing bows, but amount of hair + bows, hence you unbalance the equation.

"No player uses two bows in exactly the same way. Every bow is used in the best way to bring the desired performance out of that bow"
Agree 100%, but to assure the there's such thing as bow sound, and that bows change the sound of the violin , you must keep things equal in the comparison. That's my test. My point is precisely your very same statement right here, that any difference in sound are due to the way you play, and not the bow, its material or if it's gold or bronze. That's why in my experiment you are supposed not to hear any difference in the way of playing, because I've removed it, or tried to, hence steady open strings, fairly easy. Any influence of the bow or its materials will pop up.

June 24, 2019, 4:16 PM · "A few hairs more or less." But you see, Paul, the bows I tested when working their best had a range of hair numbers from 133 to about 250 for (the decent) bows with stiffness ranging from 135 to 333 N/m (Newtons/meter). My point is if you hair all the bows the same amount you will not have optimized the performance of some, if any. The spreadsheet has the equation I derived for optimum number of hairs given a bow's stiffness and some other parameters (also in the spreadsheet) - it was a fair match to the numbers of hairs counted in the bows I tested.

Now, if you were able to tighten each bow so that its hair was stressed (i.e., stretched) the same amount then you would achieve one optimization but not the one for straightening each bow its optimum amount.

It is like comparing the talents of the world's best basketball player with those of the best soccer player. I know rationally it seems as though one should equalize all things to make the bow tests but in a way each bow plays a different game. For one example, I have found that only about 5% of the cello bows I tested played spiccato and sautille as though they had a built in motor such that if you did anything close to the right thing with your right hand the motor would "turn on" and away you go - but only 5% --- and it seemed almost impossible to find a bow like that for sale - at least not in the bows I tested in shops at less than $10,000. I know if I owned a bow like that they would have to tear it from my dead hand! I suspect that is why they are hard to find.

I can hear the differences between the tonal qualities my bows are capable of. I've owned some of them for over 65 years - and I can tell you that through different rosins and many rehairings these relationships have remained the same on the same violin and when I have acquired new instruments to play with them and new bows to compare to them the relationships have remained the same --- even with new rosin brands the relative relationships of the bows to each other remain the same.

Edited: June 24, 2019, 6:48 PM · Paul N wrote:
"Conference, not conferences. It was one conference set by the "apparently" most famous bow maker locally, although then I've asked my three teachers over the years and none of them told me they had their bow rehaired by this bow maker. I didn't ask them if they knew him, simply asked to which one they went, and none was this man."

Thanks. I have regular contact with violin and bow makers from all over the world.

"I've asked you several times to hear those obvious differences with these tests and tell me where are the bow changes?"

Me, or others? I haven't been a part of this thread for quite a long time. I frankly wasn't interested in your newbie tests, since I have already been involved in such tests for about 50 years.

Yes, different bows can produce different sounds. No, I will not undertake the burden of convincing someone who is resistant to learning. I am not your Momma.

June 24, 2019, 7:18 PM · Ooh, snap.


Edited: June 24, 2019, 8:44 PM · So are people here now claiming that the differences are not obvious in the recordings, but are for the player (which is obviously also important)?
David Burgess, are you saying that you could have differentiated the bows from the recording, but chose not to? Or are you saying that you could differentiated them as a player?

I ask fully expecting nothing but unnecessary derision.

Edited: June 24, 2019, 9:07 PM · So, let me get this right: you say bows have optimum values (I'm not gonna argue that), which are certain number of hairs and certain tension, of course with errors, but doesn't matter.

Well, first, I thought the tension of a bow really depends on the player, his way of playing and the piece. You're saying though that there's a fixed perfect value?

Anyway, even if you're 100% right, how in heavens can you conclude that bows affect the sound of the violin, if the hair, which is what actually makes the sound happen, big impact, is completely different, both in tension and amount?

A lot of questions come up with your declarations. If in fact bows have optimum values of tension and amount of hair, then it should be even more obvious the difference, since Nate had the hair pretty equal. Why you still haven't guessed the 2 left experiments? Any complaints about them? We are comparing CF to Wood, it really can't get any more easy.

David, experiments don't convince, experiments prove. You pass, you prove you can tell the difference. Convincing is what clueless people try to do when they have no reasons to explain or prove their beliefs. Then they'll try to convince you. I'm free to reject the proven truth, but that's my problem, that I'm dumb.

By the way, the fact that some of you have been playing for 70 years doesn't mean anything, related to this. It means you know how to play the violin, you know how to read music, improvise, practice, teach may be, compose may be... but it does not give you any kind of extra points when talking about these topics.

June 24, 2019, 9:11 PM · If you listen to the four bows in the recording, can't you tell a difference between them, even if you might be hard-pressed to say which you like better?

I think that there are violins and bow combinations that have bigger tonal differences than Nate's demonstration.

I also think that for the noobs or the less-ear-sensitive people that can't hear the differences, well, you don't need to worry about it. When your ear improves to the point where you can, you may begin to care and can re-evaluate then.

I can hear rosin differences under the ear that another player ten feet away can't perceive. It doesn't mean that the differences aren't there, but that they are too small to be worth thinking about when choosing a rosin.

I spent a lot of money upgrading bows in order to get a tonal differences sufficient to make a projection difference on my current violin. Maybe I'll do some audio sound samples to demonstrate the delta.

June 24, 2019, 9:19 PM · And speaking of rosin - it is actually the ROSIN that makes the sound happen. Try plying with a newly rehaired bow that has never been rosined and you will hear a faint whisper, at best.
June 24, 2019, 9:40 PM · Paul N.

When you watch a bow in slow motion being pulled across a string (see video) it is easy to see that some of the energy of the vibrating string is transferred to the bow hair which is then transferred to the bow stick. Because of this phenomena, the stick of the bow is going to dampen the movement of the hair caused by the vibrating string. It is therefore quite obvious that the stick is going to have an effect on the sound relative to its dampening qualities.

Now you can argue all you want that the bow does not make a difference to the sound, but since the string is vibrating the hair and some of that energy is being transferred to the bow (particularly at the tip), then of course the bow is going to have some effect.

If you study bows, you will notice that many of the finest bows have sticks that are quite narrow in back of the head, the heads are refined, and the bow hair mortice is farther back. If you experiment with enough bows, you will generally find that strong bows with this kind of configuration sound better than bows with thicker less refined heads. I suspect that it has to do with how the stick interacts with the vibrating ribbon of hair, and that thicker and weaker ("whippy") sticks dampen the sound more.

But, regardless of whatever hypothesis you choose, clearly the stick is interacting with the vibrating string through the hair.

June 24, 2019, 10:30 PM · +1 Lydia again!

"I think that there are violins and bow combinations that have bigger tonal differences than Nate's demonstration."

I think so too. In fact, I think I could probably demonstrate that if I brought in a couple of other bows. One (maybe the main) reason is that, even though they spanned a wide variation in cost, all of the bows I played were decent ones. I could try figuring out which two of my bows sound the most different and try making another video, I guess.

Not totally related, but about rosin; my personal philosophy is that these choices are most critical. Violin playing is something that we primarily do by ourselves, for ourselves. Even if we perform a lot, the amount of time that other people spend listening to us is dwarfed by the amount of time we spend playing and practicing in private. Any variable that we can set that makes the experience of playing more pleasant for ourselves is of supreme importance. There are a few rosins that feel really great to play on. I think that the fact that the audience can't tell what rosin I'm using doesn't make rosin choice unimportant; the fact that I can tell the difference makes it very important. :)

June 24, 2019, 11:18 PM · The tension on a bow and the hair quality make a difference to the feel of the bow in the hand, but I haven't noticed this have a major difference on timbre. It may also make a difference if you're "bottoming out" a stick. When my bow isn't perfectly rehaired, it feels like the hair is overly stretchy, and the weight I'm putting on the string is flattening the stick against the hair, which results in a rasp in the sound.
June 25, 2019, 7:05 AM · "If you listen to the four bows in the recording, can't you tell a difference between them, even if you might be hard-pressed to say which you like better?"

The experiment is not correct: I know, after each event, the bow changes, hence I'm expecting a change, so, it's not really a blind test.
A correct blind test is when you don't know anything about what you're going to hear, and you are forced to use your hearing as your only tool, and can't hold onto something else (beliefs, knowing after each event the bow changes, etc...)
So, a correct test is that one when you don't know anything, it could be all 10 events are using the same bow... any combination you imagine. If you correctly guess those 10, then it's quite certain you're noticing the changes, if it is that there are any.
About your question, no, specially in the open strings, I can't hear any kind of differences at all. They all sound different in terms of performance, because Nate is not a robot. I've explained several times, but if you start comparing complex passages, it's guaranteed you're not gonna focus on the sound, but on the way of playing, the patterns of each sample and the mistakes Nate makes.
If you are so sure, guess number 2 experiment. Although, you have 10% of chance to hit randomly, but only 1/10 of you should hit the correct answer, so it's OK.

I'm tired of replying to each question, I'm building a wall with all the bricks. Please, those so sure about it, guess the experiments. Until then, all I can read is a person that believes something that's not minded to prove or test.

Edited: June 25, 2019, 7:53 AM · Have you ever walked on the moon? If not, how can you believe that anyone else has?

You should devise a test the astronauts can take to prove it to you. If they don't, it must be bunk. ;-)

Many decades ago, Norman Pickering had a machine which moved each bow the same way. You are very late to the party! LOL

By the way, you don't need to reply to this.

Edited: June 25, 2019, 10:44 AM · Paul N.

I posted a video that shows in slow motion how the vibrating string under the bow also vibrates the bow hair and explained how those vibrations are dampened by the bow stick. So there is an observable and reproducible physical explanation of why different bows can produce different tones and sound qualities. A bow that is less dampening (for whatever reasons) is going to produce a different tone than a bow that is more dampening.

So you now know that the bow stick does effect the sound based on the empirical observations of experts, the centuries-long refinement of bow design by archetiers, and visual physical proof of the interaction of the bow stick with the vibrating string.

Your question is now moot.

June 25, 2019, 9:21 AM · I believe Paul's main point was that the differences are at least not as glaringly obvious as the reaction of some people to his original thread would imply. If things are very obviously different you don't really need all the rigorous isolation of factors. Bow differences, at least for the audience of a recording, seem to not be that obvious. It also does not follow, George, that if a difference is physically detectable that it is meaningfully detectable by the human ear.

I haven't tried many bows myself - especially not expensive ones. I'm agnostic to the issue. I would be very inclined to believe violinists (in before a snarky remark about the value of my opinion), but I'm still not clear on what the consensus even is because of so much of this snappy attitude. Again, should the differences between the bows on the recording be obvious? Lyndon appeared to believe that at least the carbon fiber bow should stand out.

Edited: June 25, 2019, 10:11 AM · Urban, I don't think you'll get snarky replies, unless you post multiple long-winded diatribes, repeating much the same thing over and over and over.

Yes, the tonal differences between some bows can be heard on recordings. The bigger the differences between the bows, and the shorter the time between the different bows, the easier it is. Same thing with fiddles. Extremely detailed memory of sound is very short-lived. Yes, this has been tested, allowing various amounts of time to elapse, and seeing how much the accuracy deteriorates.

Edited: June 25, 2019, 2:22 PM · I won't repeat myself anymore, and I would need to do it because your replies and questions really show you either don't read my messages or don't understand them. Urban, you can substitute me because you are the only one that seems to read my mind when replying. I'm not agnostic, I'm 99% sure bows make absolutely no difference in sound, but since I don't fully trust myself I prefer to see if experiments prove me wrong. So far, total disaster, not even able to tell apart a CF bow from a wood bow, not even Maestro Lyndon. The 2 remaining tests are being totally ignored, and that's screaming what I think, so...

I give up. I'm listening to Moon River, and that's how it's gonna be, happily :)

June 25, 2019, 2:14 PM · I have a question for David -- have any contemporary violin-makers and bow-makers deliberately made pairs, where a violin and bow are specifically made to complement one another?

Was this ever done in the past? (Fred Oster once suggested that I try Vuillaume-shop bows from approximately the same year that my violin was made, and that turned out to have some excellent tonal matches even if none of the bows were ideal playing-wise.)

Edited: June 25, 2019, 3:27 PM · Lydia, not that I am aware of. Can't say it hasn't happened. But most violinmakers and most bowmakers seem to be satisfied with the usual routine of trying a number of bows with a particular violin, and trying a number of violins with one or two favorite bows.

Paul N, someone not agreeing with your messages needn't mean that they don't understand them. Wouldn't defaulting to the belief that someone who doesn't agree with you has not read what you have said, or isn't smart enough to understand what you have written, be highly arrogant? ;-)

Right now, there is a stringed instrument acoustics workshop taking place near Cleveland, with about 40 acoustics engineers, physicists, and fiddle tradespeople in attendance. There will be many and various sorts of tests. If you are truly interested in learning, and not just arguing for the sake of conflict, why are you not there?

June 25, 2019, 5:33 PM · «After all the requirements told to me, I can assure you now that in order to develop the "bow detector", I need a $5000 violin (old italian if possible), $3500 bows, become a semi-professional (a beginners can't tell) and have been bow shopping. Yet, a non-violinist, non-musician mother can totally tell the differences. Yeah, this is what I'm dealing with.»

In case you were referring to me, I would like to set a few things straight.

1: I am not a mother, I am a stepfather.
2: I am not a violinist, but that does not mean I am a non- musician. While music is not my main occupation, I have been an opera singer for over 30 years, classical guitar player for even longer, and also play flute, recorder, tuba, and viola da gamba - and I have perfect pitch.
3: Remember I said that one violin (quite a bit more than $5000 and French, not Italian) gave a much clearer response to different bows than the other (a bit less than $5000, new-ish, and Italian). One was the first instrument completed by a student in Cremona, the other one of the last instruments made by a master restorer and rebuilder of classic instruments. I think the main difference between the two is not the age of the instrument, but the result of a 60 year career of handling and working on some of the best instruments in the world at that time. The bows would have other differences - elasticity, Young’s modulus, stiffness, distribution of mass, and a lot of other small differences that together add up to very different tonality.

June 25, 2019, 7:48 PM · One question: if 3 extra hairs with 10% more age can influence sound enough to invalidate a test, why not 3 extra grams of pernambuco spread at random over the stick?
June 25, 2019, 8:11 PM · OMG, this is so tiresome.

Paul you are so right. The fact that someone is a famous VIOLIN maker, but not a BOW maker, the fact that one plays forever, 60 or 70 years, or any of the other factors mentioned (e.g. trying 200 bows or whatever) etc.means nothing.

This type of 'discussion'is so old. The religious believers are so passionate, but it's still mythology until proven otherwise. If there are differences, the differences should be universally demonstratable (and easily done so since the differences are more or less stated to be so obvious), not subjectively felt. No one is demonstrating anything, just stating what they personally feel.

And by the way, whenever on these threads someone tries to be more objective, and doesn't detect what has been pre-decided they should notice, others say it's just because the hearing isn't developed enough, The problem is that truly objective tests which are exceedingly rare, just don't prove what you all claim individually. Get it together and prove your case if it is so obvious.

Edited: June 26, 2019, 12:00 AM · The premise is that there is no difference in sound production from different bows. Whether I can identify CF bow from a wood bow is not really relevant. That a difference can be observed with a degree of consi5stency is all that matters. Basic statistical rigours is of course a must, and proper statistical analysis with a set degree of confidence (aka probability) will discern whether the null hypothesis (as stated above) can be accepted or rejected. The size of the sample necessary to draw statistically valid conclusion will be dictated by the test used (I vaguely remember 32 as the min sample size for T test, but I could be totally wrong and I am sure others will quickly point out which is the proper test and sample size). The test should be to play two bows (or the same bow twice randomly), and the listener (not through a recording, but live) must identify each time if the bows are similar or different. The right/wrong answers are then tabulated, and the statistical test applied to the data to determine if the number or right and wrong answers statistically different than that of an equal number of purely random determinations (flipping a coin for e.g.) i.e. two sets of data: one obtained from observation and one obtained from simply flipping a coin for each play of bow combination). The result of the data analysis of the two data sets will determine if they are significantly different from each other or not. If not, then and only then may you conclude that the null hypothesis is true, i.e. the bow's sound cannot be deemed to be significantly different from the listener's observations. That said, the result is only valid for the given instrument/bows/player/listener combo, so we'll be back arguing if one player can bring out the subtle differences, while another can't. Also some instruments possibly lack the degree of dynamic range and complexity necessary to react to the subtle nuances between one bow and another. You could eliminate the listener variable by repeating the test with 32+ listeners to determine if the results between listeners are significantly different or not with more sophisticated data analysis (e.g. variance analysis). This way you would be able to conclude if any given listener can reliably differentiate between 2 given bows, and whether several listeners can reliably achieve the same result and that for a given degree of confidence. Any one willing to take that on ought to write and publish an article!
June 26, 2019, 12:23 AM · There are already scientific articles out there that look at the sonic differences between bows, performed using a machine. Try this one for starters: LINK

Personally, I'm not especially interested in arguing with every noob who wants to prove that they are somehow smarter or more "objective" than experienced players. This is the kind of thing that has regrettably driven away a lot of the knowledgeable posters that seemed much more prevalent in's earlier days, at the archives bear witness to.

June 26, 2019, 5:07 AM · Lydia, it gets tiresome, doesn't it! The alternative would be to let misinformation go unchallenged, which carries its own set of negative consequences.

Thanks for posting the link to that study. Maybe that will give some of the doubters a sense of how "not-up-to-speed" they are with the research being done. Wouldn't it be nice if they had the initiative to do a little homework on their own, before pontificating?

Edited: June 26, 2019, 7:34 AM · Miss-information spreading? Going unchallenged?
Now if that's not ironic hypocrisy, I don't know what that is...

The paper is fine, kind of interesting as a whole but boring sometimes and not very clear with the conclusions, although many concepts or ideas there are not understandable unless you're an expert of the subject (aka engineer or physicist that really studied those specific subjects).

Anyways, it literally says at the end: "No clear experimental evidence for an influence on the string or bridge motion from the bow has been obtained as yet, only indications of interactions would be possible under certain conditions (I guess they refer to matching resonances?). In any case, those effects, IF captured, would be small. In summary, a lot going on in the bow, but very little, IF ANY, seen in the string or bridge"

So, capturing those differences would be, according to the paper, quite challenging using sound equipment, if it is that there are differences, let alone a human being able to detect them.

I don't understand how, after all that work, they didn't do a simple experiment testing if trained violinists with good developed hearing could tell the differences in a fully blind test. I sense because they knew 99% of violinists think bows makes the violin sound different (tonal differences as in the paper), so it would be "rude" for these experts to call violinists to prove them wrong in a strongly personal belief.

Just like scientists wouldn't call the leaders of religions after the discovery of a very revealing fact (Christianity, Islam, Buddha...), ask them questions and finally show the evidence and prove them wrong. Scientists would simply publish the discovery, and then probably most leaders and religions would tweak over time their "unmovable principles" or "facts described in the holy book" to adapt and be compatible with scientific facts, as it's been always occurring.

Edited: June 26, 2019, 7:39 AM · In my experience, people who are actually accomplished don't usually feel the need to use rude terms like "noob" to refer to people who are less experienced. This is name-calling in place of argument.

Hmm...maybe those early posters who left this forum in droves did so because of the condescension they saw here, not to mention the lack of substantive discussion.

Edited: June 26, 2019, 8:09 AM · Paul N, didn't you write this four days ago?

"OH! I knew it, I was suspecting something about the dread, more bright sound of number 2 bow. Yeah, now I'm listening to it again, and I can clearly hear how number 2 sounds overly harsh and dead. It's kind of annoying actually."

June 26, 2019, 8:08 AM · Please, tell me you understood it was irony at its most elevated exponential. I was simply imitating an hypothetical "expert" that couldn't tell at all the difference in the blind test (failing it), but once they are told which bow is being used, suddenly all the differences pop up and are so obvious. I really hope you got the joke without this explanation.
June 26, 2019, 8:11 AM · Paul N., if you haven’t seen this already, you might find it interesting to read these thoughts from some highly experienced and highly regarded bowmakers:

One quote from the article (by Charles Espey):

“What makes bows sound so different?”
“Bows of course do not make sound themselves but they accentuate different ranges of an instrument’s potential sound spectrum as well as simply mobilizing more or less sound from the instrument. There are a number of elements contributing to the bow’s sound generating potential. First and foremost is the wood; its density, its grain structure, presence of perturbations in the grain, its stiffness and the quantity of extractives such as pigments and waxes in the wood. Secondly there is the bow’s structure; its camber, its graduations or diameter from tip to butt, the height of frog and head, weight of frog and winding. Of course the qualities of the wood are reflected in the bow’s structure, strong wood will permit finer stick diameters for example. In addition the quality and quantity of hair has an effect. All these qualities of wood and structure are combined in an infinitely variable way in a given bow.”

I know you’re dug in, and you’ve become fixated on your “test,” but I'm afraid I don’t like it much, at least the one in this thread. In my experience, sometimes, between two specific bows, there may only be a small sonic difference – and these can be further diminished by the sound quality of a YouTube video. Also, a trial of 4 bows owned by an individual (that is, bows that have already -- presumably -- been through a selection process by the owner in which bows more sonically compatible with the specific violin and the player’s taste were chosen) would not necessarily reveal big differences. In other words, the test sample is small and not at all random. Others have mentioned additional issues.

My experiences (merely as a parent of a player): 20-odd years, numerous bows, most ranging from around $5k to $50k.

Sounds elicited in bow trials: most trial bows huddled in a similar more-or-less “okay” range (I’m not saying they sounded exactly the same) – with sometimes a few outliers on either side. Positive outliers would be worthy of consideration.

The positive and negative outliers, when/if they appeared in a given trial, have not been shy about making themselves known.

These determinations were not linked to maker/price, both of which were (intentionally) not known at the time of playing.

I have witnessed that these sound differences persist across players. Player 2 using the same violin and bow would certainly not sound like Player 1, but the fundamental differences in sound of different bows have been discernible regardless of the player – and player observations/preferences were aligned in this regard. The players did not know anything about each other’s preferences in advance of playing.

I’m only addressing issues of sound here and have ignored the hugely critical issue of playing characteristics.

Edited: June 26, 2019, 8:31 AM · Paul N, please go out and try some bows yourself, (which I think you said you have not done yet), rather than insisting that others need to "prove" something to you. Some us are very busy. You can also set up some tests for yourself if you like, and I would suggest that as the best way of convincing yourself.

Yes, it's very difficult to set up a single-variable test with either violins or bows, and that's one of the reasons violin acoustics is so fascinating to people like Boeing Aircraft vibration engineers, one of whom has been a regular attendee at the stringed instrument acoustics workshop near Cleveland taking place right now, which I mentioned earlier.

Again, if you are really more interested in learning than being combative, why are you not there?

Edited: June 26, 2019, 8:39 AM · I'm not "dug in" my test, I'm dug in facts and evidence. I designed that experiment real quick, and I'm sure it can be improved by an expert statistician, specially those that work with sense experiments.

"Player 2 using the same violin and bow would certainly not sound like Player 1, but the fundamental differences in sound of different bows have been discernible regardless of the player"

Yes, put bow A with old worn-out hair and bow B with fresh hair. Both violinists would say the same thing, even if they play very different styles or even genres. Or unbalance one of them. Just 2 examples.

God, do the tests or design a better one and prove you can tell the differences. Else, you're just dodging and dodging the bullet with beliefs and words instead of facts.

The time excuse doesn't work, you have plenty of time to reply, read and cherry pick, but you don't have exactly 2 minutes to listen and prove your beliefs. No, I have not gone ever bow shopping, but I told you I've tried several bows, my teacher's, other student's, various violins. Yeah, I feel like I'm trying to say to a muslim to prove me Allah exists, or something his religion says, and he answers "we don't have anything to prove to you, you need to do the Ramadan, pray 5 times each day, attend Mosque and then you will notice Allah exists".

I'm really done, and the paper right there that I didn't share, but you, concludes what I think. End of story, unless proven wrong.

June 26, 2019, 8:45 AM · "God, do the tests or design a better one and prove you can tell the differences. Else, you're just dodging and dodging the bullet with beliefs and words instead of facts."

Hmm. I kept my tone civil, Paul N, and we haven't communicated before -- so maybe dial the frustration down a tad -- I'm just talking here, offering my experiences and comments from some people who really do know what they're talking about. I'm sorry, I don't need to "prove" anything to you, and and I'm not dodging anything. For me, my decades of experience in this regard are facts -- and the only ones I need. You remind me a bit of the blind men and the elephant here. Only I'm not sure you're even touching the elephant. Like David said, and others, it might help to stop pounding the table on your test and just try some bows yourself.

Edited: June 26, 2019, 9:29 AM · Sorry Sean, not my intention to come up to you as rude, I liked you message and your tone, you're totally nice. I'm just exhausted to repeat myself to death, and argue about broad things that won't conduct anywhere, I feel it's simply a way to avoid answering the only main reason why this thread exists.

There's a fallacy though, that you even recognize quite openly: you prefer decades of experience over facts. Your experience does not give you facts or conclusions, it's simply that, experience. Your experience is an indicator that kind of guarantees that you know how to play the violin, you control bow techniques and strokes, intonation, you know how to read music, how to play louder, softer... but doesn't give you any kind of extra points about how the violin produces sound, how bows are made, and pretty much anything related to the violin that you've not tested, or should I say, tested properly.

The elephant and blind men parable doesn't remind myself at all, but precisely what depicts the paper: a violinist declaring that a violin is harsh. God forbids someone questioning the words of a 60 years old career violinist. The conclusion is wrong, the violin is simply not very loud, and the violinist, to balance that, puts more pressure on the bow, hence making it sound harsh. Not the bow, not the violin, but the violinist performance, is what made the violin sound harsh.

Anyways, see? We keep talking and talking about this nonsense, while the only purpose of this thread is ignored, covered under thousands of words and nonsense. What we can tell for sure is that none of the participants could tell apart CF from wood, and one claiming that they are so different failed as well.

I'm totally cool with you and, with everybody actually. Have a nice day!

Edited: June 26, 2019, 9:44 AM · Paul W. wrote:
"The time excuse doesn't work, you have plenty of time to reply,"

LOL, compare our word counts, or number of posts in this thread! Your choice.
Yes, I am very busy. Not only do I have a huge backlog of commissions, but also have teaching gigs in the US and Australia coming up soon. How about you?
That's largely why I thanked Lydia for providing a link to a research paper, even though it doesn't come anywhere close to representing the entire knowledge base.

June 26, 2019, 9:32 AM · @Sean Gillia:

I think your criticism of my 4-bow version of the test is pretty fair - it shouldn't be a surprise that my bows aren't drastically different, since I have preferences about bows and, generally, don't buy bows I dislike. :) However, as a point of interest:

"Also, a trial of 4 bows owned by an individual (that is, bows that have already -- presumably -- been through a selection process by the owner in which bows more sonically compatible with the specific violin and the player’s taste were chosen) would not necessarily reveal big differences. In other words, the test sample is small and not at all random. Others have mentioned additional issues."

As it happens, only one of the four bows were auditioned on the violin I was playing in the video; the other 3 predate it (one of them goes all the way back to high-school!) One of them I got in 2005, the other in 2008. The violin was made in 2010.

June 26, 2019, 9:33 AM · I meant you had plenty of time to test instead of reply with words, over and over, it was not at all a personal attack, and I hope you didn't take it as such.

My word count is because 80% of the messages are directed to me, I like to explain myself as good as I can and reply to everyone, as a matter of respect. Also, I type really fast and I many times replied while commuting. I'll stop now, knowing this is not gonna change anytime soon. Thanks for participating you all, though.

June 26, 2019, 9:41 AM · Paul N wrote, "a violinist declaring that a violin is harsh. God forbids someone questioning the words of a 60 years old career violinist. The conclusion is wrong, the violin is simply not very loud, and the violinist, to balance that, puts more pressure on the bow, hence making it sound harsh. Not the bow, not the violin, but the violinist performance, is what made the violin sound harsh."

This is where your cavorting in your noob ignorance is glaringly obvious.

If a player digs into an instrument, and the violin responds with harshness, it's the violin's fault. Yes, the player initiated the action that asked the violin for something that the violin couldn't give, but it's the violin's fault for not giving it. A good instrument doesn't "bottom out" by cracking or sounding harsh under a lot of weight (not "pressing").

June 26, 2019, 10:03 AM · I reiterate, I think this kind of name-calling -- "cavorting in your noob ignorance" -- is inappropriate and unnecessary. Make an argument. "I know better than you" is not an argument. Even "I know better than you because I've been playing X years longer" is not an argument.

Furthermore, I think "it's the violin's fault" is a cop-out. We can always have a better instrument or better bow. I guess that provides a convenient excuse for not playing well. I believe it's the musician's job to make the equipment work for them. And the good ones do - I'd certainly rather pay to hear Hilary Hahn play my cheap beginner's violin than pay a beginner to play her Vuillaume.

Edited: June 26, 2019, 10:35 AM · Jesus, I swear this is my last response, see David?
All posts are directed to me, and I feel I have to defend my point or explain in case someone "attacking me" misunderstood.

It's not me who says that, it's your paper, for God's sake. Please, read before you start calling me names. I quote your paper, not mine:
"... the players have an acting role and try to compensate things to control the system (I guess it means to control what they want to hear). This phenomenon was observed by Weinreich when a violinist was testing instruments. The player remarked that one instrument was scratchy, they tested that violin and came to the conclusion that the output was low, so the violinist had to compensate unconsciously to compensate this bad property of the violin, and forced it more than the others"
You just called ignorant noob to Gabriel Weinreich, Professor Emeritus at the University of Michigan in Ann Arbor and preeminent figure in the world of musical acoustics, because I believe that's the Weinreich they are referring to. Congratulations.

So, was the violin scratchy?
No, it was not
Was the player playing the violin much harder than others?
Yes, he was
Was the violin, as a consequence, sounding scratchy?
Yes, it was
Was it because the violin "is" scratchy?
No, answered in the first question
Did it sound scratchy because the violin sounded lower and required a much harder stroke?
Yes, that was the reason
Was it a good violin?
Apparently no, although that is personal, one could be looking for a low output violin for chamber, baroque, solo recital with an oboe... besides, it could be the violin, or a bad set up, or a bad bridge, bad sound post position or size... one should check those things before making some definitive statements about a violin.

Have a nice day everybody! I appreciate your participation, but I won't comment or defend anything anymore, some of you come at me a little harsh and that's not productive. The experiment is there, you're are free to do it, ignore it or improve it. Not more personal attacks since it's not nice, and we've proven that we can discuss things without going crazy. There's not much to discuss about this very specific topic, you've been droning around for ever without getting to the point of the thread, and that is endless. Good bye and we will see us in other threads.

Edited: June 26, 2019, 10:38 AM · "The elephant and blind men parable doesn't remind (me of) myself at all."

Of course not. That's typically how things go for people who are severely lacking in either knowledge or experience.

Lydia has been a valuable contributor here for many years. You have a long way to go before you catch up to her.

Edited: June 26, 2019, 10:47 AM · Paul N and Emily B

Blaming the violin is definitely not a cop-out, not for anyone whose livelihood depends on the sound they produce. Can a good player adjust for nearly anything? Yes they can but in the end it's not worth the effort. And the audience will notice you played on a cardboard box by the sound. You don't need the worlds finest tools, just something that's fit for concert playing and fits your natural technique of playing the violin.

Edited: June 26, 2019, 11:15 AM · @J I, I did not say that it's not worth it for a professional to have good tools. Obviously good instruments make it easier to play well. What I did say is that it's a cop-out to blame equipment for a bad sound. It can make it harder to produce a good sound, but ultimately a good player can produce a good sound even on an inferior instrument.

@David Burgess: Valuable contributor? Maybe. I don't think that's an excuse for rudeness and ad hominem attacks. What exactly are your values?

June 26, 2019, 10:59 AM · Paul N wrote:
"Have a nice day everybody! I appreciate your participation, but I won't comment or defend anything anymore, some of you come at me a little harsh and that's not productive."
Yes, I get that you have highly relied on lashing out. Hoping that you will eventually arrive at a better place. :-)
June 26, 2019, 11:12 AM · Emily, I understood what you said, but to me a bad violin still reveal itself no matter who is playing it! It is especially noticeable in the higher positions. Now if we are talking about cheaper trade violin with good playing qualities but not quite the richness, then I agree. Hilary will absolutely sound good or even great on a cheap instrument but in the end it will not compare to her best.
Edited: June 27, 2019, 5:30 AM · Some really good violinists can certainly make a really bad violin (or bow) work and sound well. A lot depends on their distribution of available resources, and how much they are willing to devote to making a bad violin sound good, versus using them for other things.
Edited: June 26, 2019, 11:34 AM · I don't understand the few credulous voices here. Paul N has some kind of thesis, which is not well defined, and he has "designed" an experiment which will prove something, and instead of just following the scientific method, running his experiment, analyzing the data, publishing the methodology and results and bringing it to the scientific community and letting people be the judge, he sort of haphazardly critiques people with more experience for even bothering to humor his nonsense.

If his problem is that people have ingrained opinions on something, which is not meaningfully defined, and that they are stuck in their received "wisdom", then given how sure he is of his "thesis", he can go out and run his experiment and stick it to them. Otherwise, I will take my experience, and the much greater experience of other people here that have tested a wide variety of bows, over our own version of some engineer crank on the internet that has his own grand unified theory of physics that is going to show the whole world, as soon as everyone agrees to a bunch of arbitrary terms specified by said crank.

Somehow, Terrence Howard's new take on new math is more honest to me than this nonsense

It's a shame that Paul has declined to reply any more ;-)

June 26, 2019, 11:43 AM · David B is describing a classic Dunning-Kruger effect: LINK

The paper description you quote is very different than the one you implied: The player remarked that one instrument was scratchy, they tested that violin and came to the conclusion that the output was low, so the violinist had to compensate unconsciously to compensate this bad property of the violin, and forced it more than the others"

A low-output violin that you end up forcing is scratchy. You can choose not to force it to get more volume, in which case you will then have to live with the fact that it's too soft. Players will describe this in various ways -- usually what they'll say is that the violin has no depth. More experienced testers will tend to use more precise language.

Good players can compensate for bad violins, but that doesn't mean that the violin isn't bad in the specific ways identified.

June 26, 2019, 11:49 AM · Paul N, bow-sound skeptics (which include me) never survive these discussions without wearing a few eggs and tomatoes. Remember that the "industry" has a pressing material need to ensure that we keep buying their wares, and that includes bow- and violin-makers not to mention shops and dealers. Professional players and teachers, likewise, have a material interest in convincing the lay public that their knowledge and experience exceeds ours in a way that conveys monetary value. I believe it does have value -- but that's not the issue here. The issue is one of bias.

Someone who claimed that Paul N was "fixated on his test" also quoted the following:

"Bows of course do not make sound themselves but they accentuate different ranges of an instrument’s potential sound spectrum as well as simply mobilizing more or less sound from the instrument."

I think I know what "accentuate" means (perhaps increasing the resonance of specific frequencies?), but phrases like "mobilizing sound from the instrument" are so vague that they are useless. That's usually what these kinds of discussions revolve around from the "pros know" side of the argument: Vagaries that are non-testable. You find the same thing when people discuss pre-amps for violin pickups or any type of hi-fi equipment: Weird language designed to obfuscate any issues that might be subject to legitimate controlled study. Much of that is intentional. I remember when people swore up and down that they could hear the difference between cheap lamp-cord and murderously expensive "monster" speaker cables; skeptics were called "noobs" and worse. Until Stereo Review financed a careful test and found -- oops -- there's no difference.

I'm known to be skeptical (NOT antithetical) with respect to the "bow sound" concept. Invariably, the question comes up as to how the bow would "accentuate different ranges" because, from the standpoint of mechanical engineers (I work with these people), pretty much the only way that would happen is by resonant vibrations in the stick. These have been measured and they're very small. "Elements contributing to the bow’s sound generating potential [such as] wood density, grain structure, perturbations in the grain, and 'extractives'" may indeed change the vibrational response of the wood -- the trouble is that the magnitude of vibrational coupling between the instrument and the bow stick is extremely small. On the other hand, my guess is that we are already dealing with a situation where the human brain (perhaps yours but not mine) has learned to discern these very small differences.

Having said all that, we must also remember that even if the bow stick makes a meaningful contribution to the violin's overall sound, it will be very difficult to separate that effect from the bow stick's playability. The more playable the stick, the more readily the violinist should be able to draw a pleasing tone. That's very important in tests where the violinist might only play each bow for a few minutes.

Finally, the kinds of "controlled studies" that people think they're doing when they try a basket-full of bows at a shop are flawed because the hair, rosin, etc., are usually not held even close to constant.

Edited: June 26, 2019, 12:47 PM · Paul Deck wrote:

"Paul N, bow-sound skeptics (which include me) never survive these discussions without wearing a few eggs and tomatoes. Remember that the "industry" has a pressing material need to ensure that we keep buying their wares, and that includes bow- and violin-makers not to mention shops and dealers."

Yes, all that can be true. As someone who may be "sold out for life" (assuming that I am either living or productive to the age of 75), I had hoped to present another perspective, not based on greed alone.

Not everyone in my business is a scammer, although I can understand why it might appear that way at times.

June 26, 2019, 12:38 PM · I never said anyone was a scammer. It's just that kind of comment that represents a needless escalation of what might otherwise be a reasoned, civil discussion.
Edited: June 26, 2019, 1:04 PM · "Scammer" was my own description, and not anything I was attempting to attribute to you.

I've had some major run-ins with the lower-integrity people in the trade. Maybe you have, or maybe you haven't.

If you want to label this as "escalation", or lack of civility, I think that is weak, and will put your blame for escalation or lack of civility back on you.

June 26, 2019, 10:23 PM · I'm voting for Christian Lesniak for president.
June 27, 2019, 4:02 AM · I don't find the test as useless as some of us do.
Same session, same violin, same player, plus the "shock" of a quick swap.

But hair from the same batch (same horse?!) with the same quantity of the same rosin? At the same tension? (Or not).

June 27, 2019, 8:43 AM · Hi Adrian,

Two of the bows were rehaired by the same person within a month or two of each other, so its *possible* the hairs were from the same horse...

They did all have the same quantity of the same rosin and, to the best abilities of my spidey senses, the same tension.

Of course, if you wanted to be extremely "scientific" you could get your archetier to stand by and transfer the hank from one stick to the next as you play, so you use the EXACT SAME HAIR each time! ;) :D :D :D

June 27, 2019, 8:57 AM · Assuming you mean the US president, then that would be a vast improvement.
And I have to ask, since this thread is all over the shop anyway, Christian, why do you have a rainbow lorikeet on your shoulder?
June 27, 2019, 9:15 AM · I think he might be a pirate??
June 27, 2019, 11:22 AM · The costume was bestowed upon me by a very perceptive coworker, who I believe was trying to express some higher self that he perceived in me. I AM currently rocking the mustache, so I just need to get my eye poked out and get the bird, and I think enlightenment is probably around the corner. So, Rosemary, Lyndon's guess is as good as mine, but it seems like a pretty snazzy familiar to have around.
June 27, 2019, 11:28 AM · Ahoy, Matey!!
June 27, 2019, 12:04 PM · Yarrr!
June 27, 2019, 12:11 PM · I think you mean "Ahoy, Natey!"


June 28, 2019, 12:41 AM · I got the pirate bit, but it’s an Australian bird, so not normally associated with pirates and wondered if there was some significance there
I did meet a sailor once who went everywhere with a parrot on his shoulder
It was scary sitting next to him. While you had a largish parrot stretching to get onto your shoulder, and contemplating a juicy earlobe, there was a stern German accented voice “ Do not antagonise the bird”

This discussion has been archived and is no longer accepting responses.

Facebook Twitter YouTube Instagram Email is made possible by...

Shar Music
Shar Music

Yamaha Violin Finder
Yamaha Violin Finder

Pirastro Strings
Pirastro Strings

Dimitri Musafia, Master Maker of Violin and Viola Cases
Dimitri Musafia, Master Maker of Violin and Viola Cases


Vsound: Electric Violin Pedal

Bay Fine Strings Violin Shop

Bobelock Cases


Los Angeles Violin Shop

Nazareth Gevorkian Violins

Potter Violins

Pro-Am Strings

Violin Lab

Wangbow Violin Bow Workshop