Keith Hill responds to Prasad's post

Instrument maker Keith Hill responded to Elatia's query related to Prasad's recent post on violin sounds. It failed to appear in the comments section (TypePad!). I am publishing Hill's opinion as an independent post.

Elatia, 

This is the third time I have read about this interesting experiment, from three different sources.

The first thing I would say is that I have yet to encounter a great antique violin that has not been tampered with by repairmen over the last two centuries. Each instance of tampering degrades the quality of the sound. What we hear in the great old instruments is probably 75 - 80% of what they should sound like had their sound not been too degraded. So it is not so amazing that violinists may select a new violin. 

I have noticed since the publication of my discovery (back in the 1980s) of Area Tuning in the old violins that a few of the best makers today are using that discovery. And when they have used it, they have figured out how to make it work...some perhaps even more successfully than I have. Still there was a great deal known by the 17th and 18th century violin makers about the art -- and it is an art (to contradict Mr. Hunt) -- of enhancing sound that we have yet to discover. 

What makes the violin so challenging to puzzle out is its incredible complexity. To date, I have uncovered 22 distinct tuning systems in a violin and more than 113 acoustical adjustments to get a violin to exhibit the 34 criteria I have learned of from the great old violins. And that doesn't even include anything relating to the preparation and varnishing.

But my work has nothing to do with the modern science of acoustics, as practiced by many modern makers, because it is ear-oriented not eye-oriented. When I hear the sounds of the great antique violins what I hear is sounds that stir my soul because they sound like human voices. In art, the aspect of paradox is essential -- otherwise our senses are quickly sated. When I hear the work of modern makers, what I usually hear is loud violins that sound like loud violins (no paradox), having little sculptability (flexibility) and sounding tepid from the point of view of carrying power (intensity) 

“Loud” doesn't necessarily have carrying power. I have made intense sounding clavichords that can fill a large concert hall--and a sound that is difficult to mold and sculpt isn't worth very much except to a mere technician. Intensity of sound and flexibility of sound are cardinal traits of an acoustically enhanced sound and must be tested for at a significant distance. At a distance of 100-300 feet the sound of most modern violins falls off and gets lost when accompanied by an orchestra. Whereas in this same test, the sound of the great antique violins blossoms, almost doubling in volume the further you are from the violin, and their exquisite timbre cuts through the sound of an entire orchestra to be heard distinctly over all the other competing sounds. These qualities are what the greatest musicians value most, according to my understanding. And these are what the violins made in Italy from 1600 to 1790 (give or take a few years) excel in. 

Ease of playing was another trait the violinists in this experiment mentioned. However, it is my view that violins that are not acoustically enhanced for intensity and flexibility are easy to play because they are not subject to the distortion resistance effect that makes the sound difficult to start without whistling, an effect that Stradivari's violins are known for. The real trick is making a violin that is both flexible and intense to be, as well, easy to play -- so that the music flows out of the instrument, instantly following the will of the player. This assumes that the player is a true artist and has a will backed by real musical understanding.

 
Keith Hill - Instrument Maker

Comments

[Dean C. Rowan]

It is a treat to read Mr. Hill's reply. Rereading Prasad's post, I again question the target of the debunking. Is it "magic-fiddle stuff" or merely pretentiousness? I opt for the latter. The rest of Prasad's concluding paragraph, on the other hand, is heartily supported by Mr. Hill's comments, although I wonder what Mr. Hill feels about celebrity. He is, in a way, already well known. I've known of him for years courtesy of LP and CD liner notes, and I'm no musician or aficionado of violins and harpsichords. He's not as well known as Stradivarius, it's true, but the Italian is perhaps less well known than some present-day tabloid cover feature of the week.

[Elatia Harris]

Dean, do you mean "The Kardashian"? That's a really awful sounding celebrity instrument that marries badly with just about anything. I may have skipped over an item in Prasad's post, but here's what I am wondering.

One, how do you control for distortion caused by the metal get-up that prevents the player's seeing what he's playing? I am of that vintage of music lovers who remember when shaggy woolen coats and furs were thought to interfere with audiophile pleasures, and had to be bundled out of the room where the turntable was.

Two, how can a violinist ever tell what his instrument sounds like to a human listener, rather than a mic, placed 100 to 200 feet away? Would the player not need the real thing -- an ultra-discerning listener -- to help in that blind-testing of the instrument?

Also, Keith Hill's web sites are extraordinarily rich. Here's a link to the long article about area tuning that he references.

http://www.instrumentmaking.keithhillharpsichords.com/areatuning.0.html

And, to read an interview Keith Hill was kind enough to give me, for 3 Quarks Daily --

http://3quarksdaily.blogs.com/3quarksdaily/2007/08/building-sound.html

[Dean C. Rowan]

I'll read your interview in a bit, Elatia. Yes, the welder's hat and perfume are obvious vulnerabilities in a so-called double-blind test, but they're only extreme examples of a wider problem with the procedure, which often creates an artificial context that undermines the purpose of the test. I, too, half-heartedly subscribe (if only theoretically these days) to weird notions of audiophile mysticism. I recall reading in one of the audiophile magazines an account of improved hi-fi sound resulting from orienting the faucet in the kitchen sink a particular way. I did not test the hypothesis.

Yes, I was thinking exactly of The Kardashian. I've not heard it played, but my understanding is that its bottom end is a bit emphatic.

[Norman Costa]

@ Keith:

Thanks so much for your reply to the 'violin experiment.' As a scientist I can say that the experiment was extremely limited. I met you in St. Paul about a year ago and still think about the little bit I learned about area tuning. It was fascinating. Of course, I thought about you when I read Prasad's post and wondered how you might react to the story. So, my thanks to Elatia for soliciting your reply.

[prasad]

Keith, thanks for the extremely thoughtful reply! I did not expect there would be a response from an eminent violin maker, much less one who's contributed to the present state of understanding. A few scattered thoughts;

- re violin tampering, this sounds quite important. (Of course, it doesn't really cut against the idea that violinists *should* pick the new ones, just for different reasons) I am curious however, have you handled old violins that have been tinkered with more or less than is typical? I'd be interesting to know to what extent such differences are comparatively apparent. Of course, it could also be that "bad" violins get picked upon for tampering more often! That'd be a pretty hefty confound…

- Re carrying power/timbre as experienced by audiences/in a large concert hall/orchestra, I'll make the obvious points. There have been plenty of such (single-blind) tests too! Quick link dump from a google search here and here. A recording of an old program (haven't listened yet) where eminent judges Pinchas Zukerman/Isaac Stern fail to pick out their own (!!) very old from new instruments. Admittedly this last was in a studio setting, although I've never yet in any other context (besides testing) heard someone argue that if you had a top violinist in an acoustically good - but small - room you were missing out the essential beauty of their playing! Indeed, at least this article claims that the old violins were in fact typically made with *smaller* settings in mind, that a lot of the tampering above is to make them sound bigger in large venues, of a sort which weren't in wide use when they were first made.

It's not my impression that concert audiences *can* distinguish old/new in these terms. Frankly, I'm not sure given how the typical member is mostly there for the status seeking, that they'd distinguish violins from cellos on average, so it's not obviously a better test than this. But even educated audiences are not really able to hear differences of the sort you mention in the post.

- Re ease of playing, one of the participants in this welding glass test comments extensively, including on this issue, here.

- I'd also pose the following meta-argument - if there were any study showing that the old violins sound much better, in any setting, to anyone, it would have been endlessly touted, and it hasn't so there isn't...

[Dean C. Rowan]

You said it, Prasad, this is whack-a-mole. Look, this is turning into a good example of the well-intentioned goal of objectivity missing the point. Nobody is asking whether old violins really sound better than new ones. That would be an absurd question. The multiplying variations on testing cropping up in these comments get closer to the truth: there are a range of factors that can affect a performer's or auditor's appreciation of a violin's sound and playability. Climate is one not yet broached here, but I'd bet it plays an important part in the quality of an instrument. My inclination not to extrapolate excessively from such tests is based somewhat on experience with hi-fi gear. Musical instruments, like audio equipment, are like friends. They have vicissitudes of personality that require time to get to know. Short term tests that demand an immediate expression of preferences have the strengths and weaknesses of speed dating.

[prasad]

Perhaps so, Dean. I sure hope violinists on the market appreciate the importance of spending a few weeks with each of the candidates before picking a violin! It'd be a pity for such decisions to be made using speed dating type methods that make a mockery of objectivity. But perhaps that misses the point too :)

To completely shift gears, climate? Ask and it shall be given…

[Ruchira]

I don't think that the point is being missed, at least not by much. While an artistic (or romantic) experience is terribly subjective and we cannot exactly quantify what (or whom) someone will or will not like and why, in this case, all other things being equal, the test was pretty precisely controlled for just one thing - the age and maker of each instrument. As Prasad points out that in other tests, the "wow" factor of a venerated label did similarly influence the perception of the listeners / players. This is nothing remarkable - it happens with various other commodities. Big names influence our value judgement. Whether that means anything in the wider world of music appreciation is not really the issue. For whatever it is worth, it sheds light on the workings of the human mind. We all have our own spheres of dazzlement. If it is not the violin or the maestro, it may well be a Kardashian or a Steve Jobs.

[Dean C. Rowan]

Yes, Ruchira, but my point is that the test doesn't properly address the quality of the instrument over the long haul, and so the question of which is the better (or let's say, the preferred) instrument is left unresolved. Yet one perhaps implied message of the debunking is that Strads aren't really better than newer instruments, because if they were, a greater number of violinists in the test would have more frequently chosen them.

My original puzzlement had to do precisely with the unsurprising finding of a "wow" factor influencing snooty proclamations about the value of older instruments or pricier wine. That there is such a factor doesn't determine which violins, new or old, a player or auditor might prefer given adequate time to get to know them. It proves that there are extra-musical influences to our experience of music. To me, that's generally a good thing.

The macro view of climate's effect on violins is not what I had in mind, Prasad, but wow, that's an interesting take! I was thinking of issues of local climate at the time the listening and playing take place. Materials expand and contract according to the weather and conditions in a venue, for instance.

[Ruchira]

A comment that Keith Hill was unable to post for reasons we don't know. (How to go about figuring out the fickleness of TypePad?)

Prasad, Elatia, Dean, and Ruchira,

I recently had the opportunity to acoustically inspect the Stradivari from 1690 once owned by Leopold Auer. As I was inspecting the resonances on this particular instrument I could instantly tell that both endblocks had been changed. The owner of the instrument confirmed that indeed both endblocks had been replaced at some point in the recent past on that instrument.

Further, one of the first Stradivari violins I ever had the privilege to acoustically study, owned at the time by Gerri Lucktenberg, now deceased, some 35 years ago, and the one on which I discovered the principle of Area Tuning, was subsequently put into the hands of an extremely well known repairman/restorer because the belly under the bridge was sinking. That repairman saw fit to glue in a patch on the inside of the belly directly under where the bridge stands and did so in order to "press" out the visible sinking effect caused by 2 centuries of down draft on the belly. By press out, I mean that the patch was installed in such a manner to force the belly back into a perfectly arched shape. The result of this all too common bit of cosmetic restoration was that the sound changed dramatically for the worse.

This kind of tampering is precisely the type I was referring to in my initial comments. Every time a violin maker or repairman pops apart an Italian violin from the 18th century or earlier, the sound most certainly must be affected, and I say this from all that I know about acoustics. This is because the more sophisticated the acoustics are the less material needs to be removed to make a huge change in the sound. That statement is equally true were the words "more" and "less" be reversed.

Another comment regarding audiences. It is my experience that the more untutored listeners are, the more purely they experience the truth about sound and music. Musicians tend to have too many vested interests in what they want out of their musical instruments. Whereas untutored listeners are receiving the sound as it was meant to be received, that is, how it touches them in their innermost being.

As for the experiment, my main objections were that the initial encounter was experienced with a "filter" on. Wearing a welding mask, if that is what they actually used, while playing a violin would be like testing out the quality of a pair of running shoes with both ankles shackled. The experiment basically tried to demonstrate that violinists can't be trusted to be objective and impartial when they are judging the quality of their tools used for their livelihood. I personally have not found this to be true especially about the best players. The wow factor is more to be found in the neophyte. The snooty factor, one I have personally encountered, is more to be found in those who are not especially confident of their ability to judge truly.

Finally, musical instruments that are highly sophisticated acoustically are altogether less susceptible to change in sound with changes in climate conditions and, I might add, altitude. Instruments I made early in my career were extremely fussy about changes in temperature and humidity. Those I make now sound and play about the same in any kind of humidity or altitude. I mention these two conditions because one of my acoustical technology students, John Weigand (Prof. of Clarinet at West Virginia University), who asked me to teach him how to acoustically enhance the sound of his clarinet reeds, mentioned repeatedly how his reeds were really fickle depending on the humidity, temperature and altitude. He reported that a reed that worked perfectly at sea level at a location that was neither dry nor cold would misbehave high in the mountains. My response to him was that the more purely wrought the quality of the sound of a reed was made to be the less fickle it would be and the longer the reed would last. I said this because it was my experience that when a reed, or harpsichord, or violin, or French Horn has been purely tuned most of the performance problems that plague players disappear. And the tuning of an instrument doesn't disappear unless someone has made a point of ignorantly changing (tampering) the instrument materially thus altering the acoustics. Eventually John reported that the reeds he was making sounded and played wonderfully no matter the conditions and that they lasted sometimes as much as 6 times longer than those he used to make.

[Elatia Harris]

Yeow! That's pretty sickening about sticking a patch inside a Strad to restore the appearance of a perfect curve. That reinforces what you say, Keith, about makers and restorers catering to the eye. I want to know about "purity" in tuning -- the kind that overrides climate vagaries and altitude variation. I have an idea you mean something very specific.

[Dean C. Rowan]

Mr. Hill's follow-up comment is at least as instructive as the first. It accords with some of my experience with stereo gear. For example, slight adjustments in placement of decent to extremely good loudspeakers can effect a substantial change in the sound. So-so speakers are not so amenable to such tweaks.

The remarks about climate are perhaps counter-intuitive. I was assuming otherwise, but I was coming from an assumption that sophisticated musical instruments are delicate, fragile. But why should that be so? A better instrument should be more securely constructed, and therefore less vulnerable to climate alterations. Makes sense to me. (On the other hand, I know that my own clarinet fitted with the best reed and played by me at any altitude sounds wretched.) Like Elatia, I'd like to know more about purity in tuning. I'll guess it has something to do with the Area Tuning process Mr. Hill has adopted for violins and, perhaps, harpsichords. By the way, I listened yesterday to the entirety of a 3-disk recital of William Byrd's "My Lady Nevell's Booke," a Naxos recording that used an instrument built by Mr. Hill. Lovely stuff.

I consider myself an untutored listener, in the sense that I have little formal musical training and can barely distinguish portamento from pimento. (Both improve with gin, however.) I'm familiar with a lot of music, but mostly perplexed by its production. I marvel at the fact that, say, Korngold wrote his sinfonietta when he was 15, or Rossini his sonatas for wind quartets when he was 12. I think I tend to miss out on "the sound as it was meant to be received."

[Elatia Harris]

Dean, Keith Hill's wife, Marianne Ploger, teaches a course at Vanderbilt University on listening to music at the speed of music. Both Keith and Marianne have lectured extensively on "receiving" sound in this way -- as well as producing it.

[Ruchira]

I would like to thank Mr. Hill for taking the time out of his busy schedule to educate us and our readers about musical instruments, sound and their vagaries. And many thanks to Elatia for facilitating the exchange.

[Norman Costa]

@ Keith,

Like Ruchira, I am so appreciative for the wonderful education on instruments and sound and enjoying music. Like Elatia, I winced when you described the patch stuck to the inside of the Strad violin. What the hell do I know about violins? It just seemed wrong on its face.

@ Dean,

How were you able to determine that a piece of music was performed on one of Keith's instruments? Now I'd like to find out where I can listen to his, and, perhaps, Cremonese instruments and others.

[Keith]

Elatia, Prasad, Dean, Norman

Since you asked about purity of tuning, tuning wood is vastly more complex than tuning strings. You can read what I wrote on Area tuning by visiting my website: http://www.instrumentmaking.keithhillharpsichords.com/areatuning.0.html Basically purity, from a sound point of view is all about ratios and proportions. A unison (one sound) is a 1:1 ratio. So anything that is purely tuned to a specific pitch or frequency produces a 1:1 ratio. If the tuning is not pure, the ratio would look like 1:0.9999999 or 1.111111111:1 In other words, 1:1 means 1:1 and nothing else. If two sounds are out by even one millionth of a hertz, a wave interference will be detectable...albeit one would have to wait a long time for the interference to manifest but it would eventually. Tuning purely means total elimination of wave interference between two pitches or frequencies. In tuning strings you just turn a peg of some sort to change the tension; in tuning wood you scrape a bit of the wood away. The problem with wood is that once you have removed it there is not way to replace it except by gluing a patch and starting over. I hope that is clear.

Elatia, you said that Marianne Ploger teaches listening to music at the speed of music. Actually, she teaches musicians to know what they are hearing at the speed of music. This is an ability (to know what they are hearing at the speed of music) that the greatest musicians of the past were able to do. Today we treat anyone able to do this as being extraordinarily talented because we know that Mozart could do this. Marianne has figured out how to teach anyone who is dedicated to learning this skill (to know what they are hearing at the speed of music) to do the same. In other words, Marianne has rediscovered how Leopold Mozart likely taught little Wolfgang to know what he was hearing so that he could write down an entire piece of music after only one hearing. Granted this skill requires intense concentration and memorization but if someone wants to learn how to do that, Marianne can teach that to them. Its not much different than doing dictation exercises in grade school, only a lot more sophisticated listening is required.

[Elatia Harris]

Thanks, Keith -- a slight but meaningful correction.

[Dean C. Rowan]

Norm: See here for the discography of recordings using Hill's instruments. I learned about the Naxos Byrd set when I stumbled across it searching for titles of recordings on Wildboar, a label devoted to early music and especially to high quality recordings of solo harpsichord recitals. The founder of that label, the late Joseph Spencer, used to run Berkeley's Musical Offering, a pretty cool classical music store near the Cal campus. I first learned of Spencer ages ago when he hosted a radio program, Chapel, Court, and Countryside, featuring early music. Subsequently, I would mail order Wildboar recordings from my home in Los Angeles. Now that I'm in Berkeley, I visit Musical Offering with some frequency.

I would love to hear from Keith which recordings of his instruments most impress him in terms of musicality, but also of sonic fidelity. Who did the best job of engineering a recording of one of his instruments? Superbly well recorded acoustic instruments are hard to come by.

[lee]

not bad! thank you for your kindly sharng!

[Dean C. Rowan]

Missed this coverage in the NYT. I like the bit in the main article about the psychological effect on a player of handling an old instrument. This, too, would be fun to test.