Judging a violin by tapping

Edited: August 18, 2017, 1:25 PM · Every now and then, lower-end instruments are being offered without strings/bridge installed. Then, one cannot judge the most important thing: the sound quality.

Would it be meaningful to test the sound of a violin just by tapping and listening to the sound, or maybe use software to obtain a frequency spectrum of the resonances? Probably, you'd need to install a bridge, held in place by at least one string (dampened by a rubber band), which can then be tapped on from the east/west-facing edges. If it isn't an original bridge, the feet wouldn't be cut to fit the belly surface. How bad would that be?

Supposing that it is possible to measure and judge the resonance spectrum, are there important acoustic properties of the violin that you would miss and that will only show when you actually play a fully set-up instrument?

Edit: never mind tapping a bare violin; just consider violins that are set up with bridge, soundpost, and strings.

I'm always worried about the subjectivity of my perception. I know from things like wine tasting, loudspeaker comparison, and even buying a new violin bow and trying out a friend's violin that I find it very hard to judge in absolute terms; I will mainly note differences (without good/bad label) when I switch and I get confused after a few tries.

Example case: I tried measuring tap spectra on my DIY travel violin (different thread). I was convinced that the bass response had increased after a move of the sound post. But the tap spectra were virtually the same.

Replies (25)

Edited: August 18, 2017, 12:56 AM · Adding any bridge or strings without a soundpost is a bad idea and also will affect the sound. If they're being sold without a bridge or strings, the soundpost (if there is one??) will have likely fallen loose at some point and need to be reset before you can do much of anything with it.

RE: Bridges, if you're testing these violins, and assuming by some grace of nature their soundposts haven't fallen down, they make bridges with adjustable feet that you can use. And then if you're going to go through that trouble, you might as well just put strings on it and try it out.

Just my .02c

Edit: I also feel like this would be an incredibly inaccurate way to test violins for quality. The important factors in buying a violin are how it sounds and how it plays. I'm not an expert on the topic, so maybe there are some numbers that can predict how it sounds, but you won't know how it actually plays or actually sounds until you give it a try. It might be subjective, but most people can pick a favorite out of a line up. It doesn't matter if it's objectively the best, it matters that it's the best for you.

Also, if it is a store selling these to end users I would be very uncomfortable. If it is a private transaction, I would be very wary of damage that might not be apparent while not under tension.

August 18, 2017, 1:20 AM · I think that you will hardly see any concrete differences in the spectra of different violins with your method.

IMO what people like in the sound of a violin is completely subjective. It is probably related on the relative amplitudes of different frequencies rather than on if they are there or not. I think these would be really hard to quantify.

This being said, the setup (bridge and strings) is a very very important part of the sound IMHO. In my understanding, a shitty chinese VSO can sound better than a masterpiece if it has a proper setup (and the other doesn't).

August 18, 2017, 3:45 AM · Both the above replies have made many excellent points, IMO.
August 18, 2017, 5:55 AM · It is impossible evaluating violins by tapping. Instruments with similar "tappings" will sound very different.
With experience, you can judge if the plates are over thinned by tapping a bit heavily (I learned that with Zukerman) but don't do that with old instruments.
August 18, 2017, 6:21 AM · I think tapping is most useful when choosing a ripe watermelon :)
August 18, 2017, 8:20 AM · everytime i walk into a music store and see violins without bridges and strings, and i asked them about them, they always tell me the same answer "thats how they are stored" i always feel bad for the people that will buy them knowing the sound post probably moved and has bad strings that havent been used yet
Edited: August 18, 2017, 8:33 AM · I bet the store clerk did not know or care about the difference between a guitar and a violin in terms of maintenance requirement. I certainly was not aware there is such a thing called the sound post hidden inside the violin when I first bought the instrument. I also wondered why the bridge is not fixed (or glued) to the body.
Edited: August 18, 2017, 8:33 AM · Actually you can tell a clunker by tapping on it quite obviously, a violin that has good tap sounds usually sounds good when strung up, but not always, certainly better odds than random, tapping can be useful for violins that are not strung up, but its something of a skill to acquire from years of experience IMHO, not easy to teach someone.
August 18, 2017, 11:01 AM · I agree, also there are often peaks or holes in the spectrum that can tell a bit.
You wont be able to tell if the violin will sound good without having a soundpost and pressure on the top, but for many you can tell they wont sound good!
August 18, 2017, 11:16 AM · What you can tell is if the wood rings and is resonant, that's a good sign, if the tap tones sound like cardboard, dead as a door nail, there really no hope.
August 18, 2017, 12:00 PM · "Sung Han
August 18, 2017, 6:21 AM · I think tapping is most useful when choosing a ripe watermelon :)"

I have a friend who authored a study on "thumping" of watermelons by folks to determine ripeness as validated or invalidated by MRI.

Doesn't work for watermelons either.

Edited: August 18, 2017, 12:35 PM · "Sung Han
August 18, 2017, 6:21 AM · I think tapping is most useful when choosing a ripe watermelon :)"

I have a friend who authored a study on "thumping" of watermelons by folks to determine ripeness as validated or invalidated by MRI.

Apparently it doesn't work for watermelons either.

Edit: I am aware of some well respected members of the trade who incorporate the tapping of a pencil-the eraser end-on the plates in different places to aid in adjusting an instrument. If it works for them and those who take their instruments to them, more power to them.

August 18, 2017, 1:23 PM · OK, those who mentioned that a bridge-less violin is likely to be soundpost-less as well: good point. I hadn't thought about that. So, I'll now just consider the usefulness of tapping the bridge of a violin that is set up.

@Michael, "It might be subjective, but most people can pick a favorite out of a line up. It doesn't matter if it's objectively the best, it matters that it's the best for you." -- for some people this works in the sense that they will be happy with their decision. For me, "subjectively the best" would mean that the choice is the same in a blind comparison, because it is well known that subjective perception is very sensitive to one's expectations. Do you know how people in the 1990s believed that green ink on the edge of a cd would improve the sound, and that many audiophiles today "know" that the music will sound better if they spend 1000s of dollars on loudspeaker cables or digital interconnect cables?

Blind comparisons are rather difficult to organize with violins. I have read many times on this forum that one should try out a violin for a week or two before deciding to keep it; how would you ever make the right decision for a choice between several violins without driving the shop owner crazy?

Because proper subjective testing is so difficult, I'm looking for an objective method as the next best thing.

@Luis, "It is impossible evaluating violins by tapping. Instruments with similar "tappings" will sound very different." -- Is there a physical explanation why this is so? What makes the sound of a violin is how the sideways motion of strings at the bridge drives the resonances of the violin body. Surely the sound from tapping the bridge must have something to do with the sound from vibrating a string at the bridge?

It seems that there are violin makers that do such measurements (Schleske showing data of his own violins and a Stradivarius), with calibrated hammers ("Impulshammerpendel", averaged over multiple angles): "The measurement process is based on impulse excitation. A computer is then used to compute the 'transfer function'."

August 18, 2017, 1:43 PM · obviously there's no better test than playing the instrument properly set up, no scientific testing is ever going to be a more accurate indicator of the tone than actually playing the violin, I was just speaking for the common case for me as a luthier of having the option to buy a violin that is not set up properly to play, at least for me tapping on violins can tell me which violins to avoid, and if the tap tones are good, more often than not, the tone is good as well.
August 18, 2017, 1:47 PM · duane,

Could you give me the reference for the MRI study you mentioned regarding watermelon? That would be an interesting article. Thanks.

Edited: August 18, 2017, 2:07 PM · Han,

Because tastes vary, room acoustics vary, set up, strings, bows, and skill vary, there is no objective test for the best violin. Taste is subjective, therefore best violin is subjective. It can only be the best based on one persons criteria. Maybe I like violins that sound dead in the upper register (extreme example, yes). A double blind study has reinforced this idea.

One of the reasons a simple in shop trial isn't really enough and you need to take it home is because of acoustics. You need to hear the instrument in different settings. My instruments sound very different in my music room, compared to a store front, compared to an open field, compared to a church.

August 18, 2017, 2:49 PM · Sung,

It is a Japanese study. I'll be there next week and I will ask him if it has been translated or will be.

Edited: August 20, 2017, 3:38 AM · Han N, I can understand your desire for an objective test, and fascination with a measurement approach. Evaluating violin tone with spectral analysis has looked so promising, that researchers have been aggressively chasing it for at least a couple of decades. However, it still has not managed to replace subjective impressions as the final arbiter, even for these researchers. (Martin Schleske is the only researcher I know of who may think it has.)

One of the problems, in my opinion, is that FFTs don't display everything that the human sound processing system "hears", with its complex system of signal acquisition (the ear), and brain processing.

With your background as a physicist, perhaps you can come up with an analytical system which is more useful, than what other physicists have come up with so far?

If that's something you'd like to do, then as starting point, you might want to read all the violin research that George Bissinger has published.

August 20, 2017, 8:34 AM · One of the reasons a simple in shop trial isn't really enough and you need to take it home is because of acoustics. You need to hear the instrument in different settings. My instruments sound very different in my music room, compared to a store front, compared to an open field, compared to a church.

Yes, I've taken and instrument that sounded fantastic in the dealer's small room only to find 30 minutes later in a recording venue that it sounds really awful. Uneven strings and a much worse sound.

Another point, are we talking about tapping the back/belly or sound-post? I've not heard of anything useful from SP tapping, but maybe I've missed out on this.

August 20, 2017, 9:01 AM · Doubtless there are violins around where only a cursory inspection is needed to determine that the tapping is best done with a 2 lb hammer to put the poor thing out of its misery ;)
August 20, 2017, 9:25 AM · Peter, the problem with in-store testing is that it may not be just a relatively empty room, perhaps like a recording studio, but a space in which the walls are adorned with perhaps dozens of violins and violas each of which is a resonator for the instrument under test. So what the tester is hearing is not just the sound of the test instrument but an overlay from all the surrounding resonances.

I once accompanied to a dealer an orchestral colleague who was looking for a better cello. The shop's cello room had about a dozen cellos in it, all of which replied vociferously to anything my friend played on the instrument she was trying out. It was ridiculous, and impossible to make a reasonable assessment. She eventually decided on one and took it home to give it an extended workout. I think she decided to buy it.
Edited: August 20, 2017, 11:47 AM · Trevor - the room in question was very small and long and it had highly reflective surfaces, walls, windows etc. The end result was that it made the instrument sound great because of all the reflections and at the same time muddied things up so the strings all sounded fairly equal. (Well known dealer in central London).

In the recording hall (it was a famous venue for orchestras and chamber music which does not exist now) the instrument in the large space sounded dreadful. (Kingsway Hall).(This was way back in 1981).

August 20, 2017, 11:57 AM · @Michael "there is no objective test for the best violin. Taste is subjective, therefore best violin is subjective." - my worry is that taste is not only subjective, but also irreproducible. One might do the same comparison the next day and decide that the best one is a different one from yesterday, because of mood differences, because they were listened to in a different order, or because someone else mentioned that one of them is really beautiful/expensive.

@Peter "Another point, are we talking about tapping the back/belly or sound-post?" - neither; I mean to tap the bridge (tapping the vertical narrow edge), which is as close as you can get to where the strings couple into the instrument. See pictures by Schleske.

@David, actually, I think signal analysis is an excellent approach to analyze a sound. If you're willing to accept that you can judge a violin played by a good violinist from a good recording, then all data is in the digital signal. The main issue with FFT-based spectra as a particular way of representing data is that you don't see the phase relations between the frequency components. I'd guess that that piece of information is not very important in the perception of sustained tones, though.

However, as a physicist, I do see some issues with tap responses that have nothing to do with how the brain processes sounds:

* The point where one might easily tap the bridge is not the place where the strings interface with the bridge; therefore, the effect of the bridge on the sound is different for the tap versus the strings.

* Part of the sound color comes from the feedback of energy from the bridge back into the strings; with a tap test, the hammer is no longer in contact with the bridge by the time the bridge recoils from the impact.

* Finally, a violin will not radiate sound uniformly into all directions. The sound color will therefore depend on where the listener is, and of course, the room reflections. Schleske takes the average response over 36 angles. That's a particular choice, which may or may not be representative of what a typical listener experiences.

Re "perhaps you can come up with an analytical system which is more useful" I think I'll first need to become a halfway decent violinist. :-)

August 20, 2017, 12:35 PM · As a maker (and retired aerospace engineer), I have tried to use various means of judging instrument quality at all stages of construction, with the idea that I could make modifications at that point to improve things. It's pretty hopeless. At best, with a soundpost in place, you should be able to identify something truly horrid vs. something decent... with enough experience. Perhaps with a massive research program, you could get finer gradations of bad and good... but why? In the end, people often disagree on what's bad and good anyway, even with all the parts in place and able to be played.
August 20, 2017, 2:04 PM · At one point, a Strad was considered to be the sound to emulate. But with further experience and research, it was discovered that one Strad can sound radically different from another, and that in a lineup, a Strad isn't always the most preferred instrument by either the player or the listener.

So Han N, in addition to all the other challenges you and I and Don Noon have rightly mentioned, what would you use as the "reference standard" for your objective sound evaluation system? That's turned out to be a major sticking point for the main violin sound researchers, and something they're still trying to sort out.

How would you suggest getting past that sticking point?

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

Facebook Twitter YouTube Instagram Email

Violinist.com is made possible by...

Shar Music

Pirastro Strings

Coda Bow

Violin Lounge Academy

Yamaha V3 Series Violin

Dimitri Musafia, Master Maker of Violin and Viola Cases

Metzler Violin Shop

Gliga Violins

Corilon Violins

Leatherwood Bespoke Rosin

Anderson Musical Instrument Insurance

Bobelock Cases

Fiddlerman.com

Fiddlershop

Nazareth Gevorkian Violins

Los Angeles Violin Shop

Pro-Am Strings

Wangbow Violin Bow Workshop

Subscribe