James Boyk Home
Articles by James Boyk on Audio & Music
Recording Credits
James Boyk's Audio Consulting Services

James Boyk: Reflections on music & audio
1 The most important thing to know about audio...
2 The definition of accuracy in audio equipment...
3 How do we measure the accuracy of audio equipment?
4 When will digital audio be good enough?
5 "Perceptual crossover"
6 Can a good hifi-system enhance my appreciation of my classical CDs?
7 Beethoven's ear
8 Piano student's preparedness
9 The piano as a tool

The most important thing to know about audio...
is that Music Is Beautiful. By "beautiful" I don't mean "pretty-pretty"; I mean simply that music draws you in, rather than puts you off; that you want to experience more of it. This is the most obvious thing about live acoustic music: people gather 'round when it's played; and when the players stop, they're asked to continue.

This beauty is very complex, and certainly multi-dimensional, matter. But if a system of recording & playing back does not capture some of this quality, then it is missing the most important element it can have.

Some audio people in effect claim ugliness as a proof of accuracy in recording & playback systems, and put down a love of beauty as "a preference for euphonic colorations." It makes me wonder whether they've heard much live music; and if they've heard it, whether they've understood it.


The definition of accuracy in audio equipment . . .
is Musical Accuracy As Determined By Listening (because listening is the purpose of the equipment). Therefore the obvious fundamental original authentic test is a listening test. (And not just any random kind of listening test, either; but that's another matter.)

Listening tests are perfectly do-able, though expensive & time-consuming. Of course engineers want easy, & easily repeatable, measurements to use in designing or modifying a product. But to justify using any such test, one must first validate that test with reference to listening. What does it mean to validate a proposed test? A four-step process:

Step 1: Carry out a valid listening test to rank a large group of components of some particular kind: amps, speakers, whatever.

Step 2: Measure the same components using the proposed bench test.

Step 3: Demonstrate that the proposed measurement does in fact correlate with the ranking from step 1.

Then comes Step 4: Do this again & again for more components of the same kind. After enough tests of enough components, people will start to believe that your proposed test really is valid, and we'll all have a party. I'll bring the Veuve Clicquot!

But so far, no one has validated in this way any measurement that applies generally to audio components.

It is often argued that various test-bench measurements are "naturally" or "obviously" important. This is naive. Not enough is known about the ear-brain system to know what parts of the music signal are important to preserve. Preserve them all perfectly, and you're home free; but we can't preserve them all perfectly; that's why this is a branch of engineering, not theology! Engineering's glory is its ability to deal with trade-offs. Engineers know that every time you make something better, you make something else worse.

A good engineer can make any aspect of performance good, as least up to a point. Say you're a speaker designer. Choose whatever seems important to you. Is it flat frequency response, extended bandwidth, constant radiation pattern, good impulse response, non-excitation of room resonances? These all seem "obviously" important. But when you make any of them better, something is being made worse.

And what if the aspect of performance being made worse is more important to the ear than the thing you're making better? If you don't know what you're trading off, or you don't know its perceptual importance, you're not in control of your design. (And if what's being made worse is only the cost, you're exceptionally lucky.)

The Ear Is The Definition of what audio is about.
The Ear Is The Definition of what audio is about.
The Ear Is The Definition of what audio is about.
Wrestle with this fact as you will, jib against it, hate it, wish that it were different—it's still a fact.

So please don't tell me about measurements. They're fascinating. I deal with them all the time. When gathered in the right way (much tougher than anyone realizes who hasn't tried it), used with caution, and tested against hearing, they can be useful—a bit. But one of the best crossover networks I've heard was designed entirely by listening.

Whenever someone claiming to take a scientific or an engineering point of view claims that measurements tell the story, just ask him to point to the test that validated the measurements against listening. Until there are measurements scientifically validated as described above, listening tests will be the only way to judge. It's unfortunate, it's frustrating, it's expensive, but after all it's glorious. How much more pleasurable it is to listen than to measure!


How do we measure the accuracy of audio equipment?
I don't know of any manufacturer of consumer audio gear—including the highest of high-end components---who does the kind of testing required to determine sonic accuracy. It is impossible to overstate the importance of this simple fact! In the pro audio world, the situation is no better, though I've heard claims that there are one or two manufacturers who actually listen to their equipment in a reasonable way. More typical however is the story I heard from a designer of a digital recording console, who told me that even though he begged for listening tests, the maker wouldn't allow them. So this half-million-dollar unit had mike preamps and A/D converters never tested against live sound!

I'm going to risk making an analogy to a field about which I know nothing, namely video. If the analogy is full of errors, ignore it.

Act I
If I want to judge two TVs for the accuracy of their pictures (not just for which is my preference on material I watch), I would take them into a setting where I could look at some objects with my own eyes and then look at the reproduction of those objects on each TV. I would judge accuracy, not by comparing the pictures on the two TVs with each other, but by comparing each with the original appearance of the objects in front of the camera.

If I want to judge two methods of video recording for accuracy (not just for which is my preference), I would, as above, try both methods in a setting where I could compare the results of each to the original. Using the very best equipment for every component, I would look at the peach (say) in front of the camera while taping it with both machines. Then I would play back one videotape & look at the peach on the monitor. Then I would look at the actual peach again & compare it to the other videotape.

I would judge accuracy not by comparing the appearances of the two videotapes to each other, but by comparing each with the original appearance of the objects before the camera.

I'm intentionally omitting all kinds of details. These might include things like this: the camera & other equipment should be unimpeachably good; the lighting should be correct; the objects chosen should have a complete range of colors & textures; one or more of them should be in motion; etc. etc. There might also be considerations about statistical methodology. I'm leaving all this out, not because it's unimportant but because it's irrelevant to my point. All I'm trying to get agreement on is the basic idea, namely that —

In testing for accuracy, one compares video output with video input, where the input is the original appearance of the thing before the camera.

So far, I'm fondly imagining that everyone agrees.

Act II
If I want to judge two pieces of audio gear for the accuracy of their sound (not just for which is my preference), I would take them into a setting where I could hear some sounds with my own ears and then hear the reproduction of those sounds on each piece of audio gear. I would judge accuracy, not by comparing the sounds from the two components with each other, but by comparing the sound of each with the original sound in front of the microphone. (This is crucial & non-negotiable. To understand why, see the P.S.)

If I want to judge two methods of audio recording for accuracy (not just for which is my preference), I would, as above, try both methods in a setting where I could compare the results of each to the original. Using the very best equipment for every other component, I would listen to the piano (say) in front of the microphone while taping it with both machines. Then I would play back one audiotape. Then I would listen to the live piano again & play back the other audiotape through the same equipment.

I would judge accuracy not by comparing the sound of the two tapes to each other, but by comparing each with the original sound of objects before the microphone. (Again, non-negotiable; see P.S.)

I'm intentionally omitting all kinds of details. These might include things like this: the microphone & other equipment should be unimpeachably good; the background noise level should be very low; the sounds should have a complete range of "colors" & textures; etc. etc. There might also be considerations about statistical methodology. I'm leaving all this out, not because it's unimportant but because it's irrelevant to my point. All I'm trying to get agreement on is the basic idea, namely that —

In testing for accuracy, one compares audio output with audio input, where the input is the original sound before the microphone.

At this point, I'm imagining everyone agreeing in principle. And if so, I ask, "Then how can you judge anything in audio if you haven't compared live with reproduced?" And the answer is, of course, You can't.

Note that this has nothing specific to do with
      Analog vs. digital
      Tubes vs. transistors
      Condenser microphones vs. ribbons
      Electrostatic speakers vs. dynamics vs. horns
      Lp's vs. CD's
      Direct-to-disc vs. tape
      78 rpm vs. 33 vs. 45
      Triode amplifiers vs. pentodes
      Diamond styli vs. sapphire (my childhood)
      Cactus needles vs. steel (before my time),
or anything else! It has to do with the basic question of how one evaluates accuracy in audio equipment. Which means, of course, that in reality it has everything to do with all of the above!

"But," someone says, "I don't really care about accuracy. And I've never heard a valid comparison in my life; it's too much work. I just care about whether I like the sound." To which I reply, "More power to you! But then don't pretend to tell us about accuracy of audio equipment or recording techniques."

P.S. Here's an important point from my article "Rules of the Game," (HiFi News & Record Review (England), Jan. 1983)

. . . An informal listening experience underlined the importance of having an identified direct feed as a reference. In December, 1981, twenty listeners heard a half-hour of music direct, while the feed was also recorded on a professional digital recorder. After a short break, everyone listened to the playback.

In such a comparison, any difference must be a degradation. [Repeat: In such a comparison, any difference must be a degradation.] Yet two listeners preferred the digital playback to the live feed!

This suggests a 'thought-experiment': What if we had a literally perfect recorder but we weren't aware that it was perfect; and what if, instead of listening to the direct feed and the digital recording, we had listened only to the playback of the two recorders. Those two listeners would have preferred the digital machine to the perfect machine, and the experiment would conceal the fact that they were preferring a degraded sound. This problem can be avoided only by using a labeled direct feed as a reference. . . .


When will digital audio be good enough?
The Research Department of the British Broadcasting Corporation, which for over 60 years has been a fount of important work in audio, published a paper almost a decade ago about how good digital has to be before people don't hear what the BBC calls "quantizing distortion." (The research cleverly simulated having more bits than were really available.)

The conclusion of this paper was that 22 bits per sample were required to digitize music of the widest dynamic range. Well, we're not there yet, folks. The highest bit-rate machine I know (Nagra/dCS, $55,000, which we auditioned against a live feed in the Caltech Music Lab in November, 1996) is nominally a 24-bit machine; but according to the manufacturer its dynamic range is "only" 17.5 bits. Very good, of course! But not 22.

If my memory's correct, the BBC found that 22 bits were necessary in their tests so that "quantizing distortion" would be inaudible to 99% of listeners. That means, of course, that 1% still heard it.

P.S. In audio, one cannot take for granted the competence of the engineers doing the fundamental designing. Philip Greenspun has pointed out, tongue only partly in cheek,

"If it is built in Japan, audio equipment is designed by engineers who couldn't get jobs designing video equipment. If it is built in the US, audio equipment is designed by engineers who couldn't get jobs designing high frequency electronics or computers."

Years ago, I gave an informal talk at Caltech (where I teach the interdisciplinary course "Projects in Music & Science") about the then-current state of digital audio; and I mentioned problems copying digital tapes. A colleague came up afterwards and chided me, saying that digital copies couldn't be anything but perfect, as in the computer world. I asked if he understood the format that was being used in audio, and told him how many bits per square inch were being asked of the magnetic tape material. He immediately said, "But that's beyond the limit of what's possible!" QED


"Perceptual Crossover"
Once I was at Doug Sax's place when they were setting up to record piano from a studio where Lincoln Mayorga was practicing. Doug said, "I've got two mikes in there and I'm comparing them in mono. Which do you prefer?" I listened for a while and picked one. Then he mixed them up—changed channels repeatedly so I couldn't follow—and again asked me to choose; and again I did so.

Then he grinned and said, "There's just one mike there. You're listening to the same thing through both channels. The only difference is that one channel is two-tenths of a dB louder than the other. Each time, you picked the one that was louder."

Had I been asked whether the levels were different, I would have denied it absolutely. (Doug agreed about this.) I don't think it's possible to pick a 0.2 dB difference on music, especially without the opportunity to play back the identical thing repeatedly, which of course is impossible when the source is live music-making.

Yet the louder one sounded better. I've coined the term "perceptual crossover" for this. In other words, the difference is in one area (level) yet subjectively appears in another (sound quality). There are many other examples, too, so this is a useful concept, I think.


"Can a good hifi-system enhance my appreciation of my classical CDs?"
(a) My response to this posted query: Yes, definitely. Proper choice and matching of components will maximize not just musical enjoyment but "musical pleasure per dollar," if we can imagine such a criterion. A well-chosen system will perform better than a poorly-chosen one of many times the price. On the other hand, "proper choice and matching of components" is not easy, and it's very hard to come by good advice. The magazines are useless. Their information is often wrong; and when not wrong, is rarely of use because it's out of context. And retail hi-fi salesmen are profoundly ignorant (with the rare exceptions, of course; but they are *really* rare, and one has to be an expert oneself to tell which they are).

Choosing the right system for *you* means for your budget, your room, your taste in music, your taste in sound, and so on. Two people with the same amount of money to spend may nevertheless need two entirely different systems. For instance, one client of mine was an academic radiologist in Colorado who listened to just two kinds of music: (1) string quartets of the classic period at moderate volume (when his wife was home), and (2) requiem masses with the volume "cranked up" (when his wife was away). Another client was a serious ballet student who also listened to two kinds of music: (1) the classic ballet scores, and (2) heavy metal. The same system will not satisfy these two! (Unless the system were extraordinarily expensive, which would have been beyond the reach of either.)

Or take another two: When I asked them what they like about live musical sound (this is one of the questions on the three-page questionnaire I've developed over the years), one of them said, "Well, I sit in the 3rd row at the Symphony, and you know how you can just swim in the sound when you're that close? That's what I love." The other client said, "I love, not so much the power of it, but the clarity of volume changes and textures." These people could not be satisfied by the same system, again unless it were ultra-expensive.

Then there are the clients who tell you they don't want to spend much money, but who then reveal that they want something that's all but impossible at any price, such as flawless reproduction of the the female voice, or terrific choral sound, or the "sock in the gut" of the bass drum.

Enough of this.

By the way, I find that the matching of components takes about 2/3 of the time, and the choosing about 1/3.

Not a very satisfactory answer! In effect, I'm saying that Yes, a good system will make a difference, but you can't find advice that is both true and useful! But unfortunately that's the situation. Nutty, isn't it?

My articles may be useful. Specific recommendations from older ones may be out of date, but the point of all my stuff is that the reader can learn the process by which equipment may legitimately be chosen.

(b) May I add a remark to my previous posting about reproduction of musical sound? Think about musical "attacks" -- the sound from the first moment the piano hammer hits the string, or the first moment the bow starts the violin string going, or the first moment the drumstick hits the drumhead. Attacks are usually louder than the succeeding body of sound on the same note. Because attacks are so short, we do not hear this greater loudness the usual way. Rather, it comes across to us as "presence," or "liveness." Nevertheless, the amplifier and speakers are for those instants required to supply more power.

If they can't, then a crucial element of the music vanishes. And then you have a system that can play perhaps quite loudly on the average--that is, for every part of the notes except the attacks--while pooping out on the attacks. This system will sound dynamically constrained, not free. Of course this same system will be able to handle the attacks all right when playing softly; but this may require *very* soft playing indeed.

Have you ever heard a recording you own sound better on FM than on your own system? This is because virtually all FM broadcasting removes the dynamic peaks associated with the attacks. Therefore it removes them as a source of overload to your amplifier or speakers. Therefore the amp or speakers may be "happier." But *you* won't be happy, because either way, you're not getting correct attacks.

OK, how much power do you need? There's a big caveat here, which I'll get to in a moment; but let's say that you are using 2 watts of amplifier power to each speaker to get an acceptable volume level overall. You will certainly need 60 watts of power and, if the recording is one that preserves attacks exceptionally well (a matter of microphone technique and other things), you may need as much as 200 watts of power per channel for each instant of the attacks.

If you're using 5 watts per channel to get the average overall volume level you want, then you will certainly need 150 watts and you may need 500 watts per channel for the attacks.

Whether you need 2 watts or 5 watts for the overall volume you want (or 1/2 a watt or 10 watts) depends on your room's size and acoustics, the efficiency (or "sensitivity") of your speakers -- and of course, how loud you want to listen.

Here's the caveat: The *published* power rating of an amplifier tells you almost nothing about how much power it can actually deliver to a speaker. Not that it's incorrect. No, it's simply that the standardized conditions under which these measurements are taken are irrelevant to the real world. In a test published some years ago, for instance, one amplifier (correctly) rated at 20 watts per channel played louder through a particular speaker (for a given distortion level) than another amplifier (correctly) rated at 100 watts per channel!

Nevertheless, this gives you the idea, I hope, and it's an important one if you want to get pleasure from your collection of recordings:

Other things being equal, you want a more powerful amplifier.

Get one whose rated power per channel is at or near the maximum your speaker is rated for. (I usually specify an amplifier of considerably greater power, but obviously I cannot recommend this to the layman, though it's interesting to note that, for reasons beyond our scope, an "over-powered" amp is less likely to damage a speaker than an underpowered amp.)

By the way, choral music is one of the *most* demanding in this way!

(c) I offer below an anecdote from my book, To Hear Ourselves As Others Hear Us, following which I respond to Mr. E.'s post.

First, the anecdote: We pianists are unfortunate in not carrying our instruments around with us. But we are not the only ones with omnipotence fantasies; for many musicians feel their art will come through despite defective recording or inadequate playback.

This is wrong, as I found out when I sent my first album, an Lp, to a young pianist friend who was studying at Peabody Conservatory. In her return letter, she was enthusiastic about the Scarlatti; but her embarrassment was almost palpable when she discussed the Beethoven Sonata Opus 111. She said the first movement seemed to have too many climaxes, and too much banging.

I did not hear the banging or the multiple climaxes. But I knew I might be hearing what I thought I had done, or what I had wanted to do, not what I really did. I decided to listen again in a few months.

The summer came before I got around to it, and my friend wrote again. At home in Colorado for the vacation, she had listened on her father's system, which was much better than her dormitory-room player. Now she did not hear the extra climaxes or the banging. Now she loved the performance. But of course she was puzzled by the change from the same disc heard at school. Reading her letter, I suddenly realized what had been wrong. Unlike most recordings, this one has a wide dynamic range. When the music got above, say, *forte*, it had overloaded her dormitory system. Any such passage came out equal in loudness to any other such passage. Hence the multiple climaxes. They were graded dynamically in the playing and on the recording, but couldn't be distinguished by the system.

Overloading also makes the reproduced tone ugly; but because my friend was thinking in musical terms, she heard the ugliness created by her system *as though it were created at the piano*! This shows why record reviewers who don't have good systems risk misjudging interpretation along with sound quality.

(d) Mr. E. [an audio salesman] wrote: "...I've found [that] for pop, speakers with good bass - large woofers - two way will usually suffice; for classical the mid/high frequencies are very important and one can get away with a small bookshelf speaker of good quality (two way) of course a larger three way is usually best for everything including the "big bangs" in things like Tchaikovsky's 1812. It's hard to hear audible differences in amplifiers but don't get a very low output amp if you want to enjoy classics without having to continually adjust volume. Try to buy from a dealer that has a "comparator" so you can compare different combinations at the press of a switch as the brain doesn't "remember" slight differences in sound.... A strategically placed rug can do more than $500...."

I'm happy to believe that RE does a fine job for his customers, and I certainly agree about judiciously placed rugs. But generalities like these are neither useful nor correct. "Large woofers" may or may not give "good bass." Some kinds of classical music--organ, piano and orchestra, for instance--require the low notes for full appreciation. "Three-way" speakers may or may not be better than "two-way" (the BBC's own prime monitor speakers are two-way). Amplifiers differ greatly in their sound quality. And comparators are notorious for degrading the sound of everything connected to them so much as to make valid comparison impossible.

Even at the lower ends of the price range--and reasonable sound for a CD-only system in a small room might begin as low as $1200 or so here in Los Angeles--it is simply essential to talk about the virtues and defects of individual components and their usefulness in complete, carefully-matched systems.

(e) Mr. P. G. wrote: "[D]oes anyone really contend that the nature of the imperfections in a system determine its suitability for one kind of music over another? Imperfections, once a fairly low standard has been exceeded, equally affect all kinds of music, and equally affect all discerning listeners."

My response: Yes, I do "contend" that the "nature of the imperfections" in a system determine its suitability for one kind of music, and more important, for one listener, over another. Of course!

If music-lover A listens only to guitar music, then it will be not too difficult to make a system (call it system A) that plays loud enough (in an ordinary-size room) to reproduce the actual volume level of the original instrument. If music-lover B listens mostly to piano, or orchestra, or chorus, then system A will not work well because these are much much louder than guitar. Therefore one must design system B, which can play louder. If system B is to cost the same amount as system A, it must give up some other virtues in order to get the capability to play louder. System B will need, for example, a bigger amplifier, and perhaps loudspeakers which can produce more acoustic power.

Very broadly speaking, there are tradeoffs between power, on the one hand, and delicacy, on the other hand. System A can put more of its money toward delicacy because it needs less power. Music-lover B, who loves piano, or orchestra, or chorus, is just as enamored of delicacy as music-lover A, the guitar listener. But B can't have the delicacy because he needs the power.

The wild card is money. If B is willing to spend a *lot* more money, then he can have both the delicacy and the power.

Many other examples are possible in which a power difference is not in question. For instance, a tweeter (treble speaker) with a somewhat rough response (that is, one that exaggerates some pitches) may be ruinous to reproduction of female voice. Of course it will not be good for any instrument or combination! But it may well go much less noticed on say, brass or percussion, and may in fact give a kind of excitement to the sound. It's not accurate! But its inaccuracies may well be minimized by one kind of music, while being ruinous for another.

This is the kind of thing that I have worked with for 20 years (when not practicing), and I assure you that they make a big big difference.


Beethoven's ear
M. S. posted the following: "Beethoven in his later years probably due to his deafness had to rely on what he learned in his youth: cadential harmony, standard sonata and fugue forms, themes & variations. Deafness posed great challenges to his creativity. For instance, the Gloria in the Missa culminates in a fugue."

My response: Every note of Beethoven's composition testifies to the extraordinary accuracy of his inner ear. Not just the harmonies but the sonorities are perfect. (You know that string-quartet music, say, played by a brass quartet will not work because of differences in the sound qualities of the instruments.) Tovey, that universal expert, says of late Beethoven "He does not make mistakes of sonority," or words to that effect.

Realizing that Beethoven wrote sonically flawless works--the Opus 131 quartet, the Opus 126 Bagatelles, the first bars of the Ninth Symphony; pick anything at random--makes it utterly impossible to believe that he "had to rely on what he learned in his youth." With respect, the idea is ludicrous.

Yes, it is almost literally incredible that anyone should have an inner ear of such powers. But his music testifies that he did.

And I completely fail to see how his culminating a movement in a fugue shows that "deafness posed great challenges to his creativity." With respect, I must confess that since musically speaking, Beethoven could put the whole lot of us in his pocket and not notice our weight, I find it unbelievable that anyone would have the nerve to condescend to him. Condescending to Beethoven!


Piano student's preparedness
In a recent post, another list-member presented me with the bouquet of saying he'd enjoy studying piano with me, and then the brickbat of saying I was all wet to claim that a student's preparedness for a given work is "testable." (Originally, another professional musician had raised the question, "Why are the 'Pathétique' and 'Moonlight' Sonatas so hard to teach?" I responded, "Partly because so few students are ready to learn them." For some reason, this offended people. List-members who are excellent chefs would think me ridiculous if I walked into their kitchens and said, "I know nothing about cooking but I want you to teach me to make saumon en croute today." Experienced drivers would think me addled if I said, "I've never driven but I want you to give me my first lesson today -- on the freeway.")
     The current post addressed the question of whether one could "test" the resulting performance. This was not what I said. I said you could test whether a student was ready to learn the piece.
     Learning the same work can have different functions for different students. For student A, already mature in technique and musicianship, learning and memorizing the 'Pathétique' to add it to repertoire might be the exercise of a few days. For student B, who takes six months on the task (with other pieces at the same time), it can be a laboratory for acquiring new technique. As always, this carries the risk that if the technique is acquired not quite fluently, some glitches will always remain. (They are terribly difficult to remove, no matter how much more accomplished one gets.)
     For student C, who cannot read music and cannot play, but who is crazy about the piece, it can be a laboratory for learning everything. This is not hyperbole. When I was in college, there was a guy who loved the Bach D Minor Concerto but knew zero. He bought the score and a basic music book, looked at the score and said "Hmmm... the first note is... [reference to the book] D. And D on the keyboard is...this note. OK, this first note of the score is...THIS [playing the note]. Now what do these two flags on the note mean? [Looking in book] Oh, it's a 16th note. OK, now what's the next note." And so on. I am not making this up. If anyone claims that this fellow was "ready to learn" this piece, I will give up on this line of discussion.
     For other students, the piece can have other roles; but a definition that allows for all possibilities is the following:
     A student is "not ready" to work on a piece if there's a good chance that the resulting performance will have built-in defects.


The piano as a tool
In response to an irritated post, I wrote: Why is it not fussiness or affectation for a pianist to try to assure that he or she has the finest piano in the finest possible condition for each performance? Because the piano is, with regard to the performance, simply a tool. If the tool is faulty the performance will be impaired.

A piano technician wouldn't by choice use poor tools. If the socket on the tuning hammer were made of inferior metal, so that it soon wore out and slipped on the pins, it would be useless for tuning; or even if it were usable, it would waste effort. Would it be fussiness or affectation or arrogance for the technician to get a good tool? Would it be a statement that the technician thought himself or herself better than others? No, it would mean that that person just wanted to the best work possible.

Note that even with a defective tuning hammer, the tuning might get done correctly--eventually. But musical performance, by its very nature, allows no excuses, no delays, no second tries. Otherwise one gets what one does so often get, 'just another piano recital.' Routine is not why people go to concerts; they go to be moved, to be exalted, to have life illuminated by great art. If one isn't at least attempting to do this, one has no business playing the great works.

No performer can guarantee to play a great recital. But it's hard enough to make music even on a perfect piano. Any imperfection in voicing, regulation or anything else--or in the fundamental sound quality of the instrument--makes it that much harder, that much more likely that one will not reach the heights.

Attempting to assure a perfect instrument is not an arrogant statement that one is oneself perfect. It is a way of putting oneself on the spot. "Well, you've got a perfect piano. You have only yourself to blame if the performance isn't good."

Of course most recitals are played on pianos in far from excellent condition! But that doesn't mean that this situation is desirable! The pianist who works actively to have the best possible instrument in the best possible condition is showing an ethical responsibility to the composer, the audience--and himself.

James Boyk Home
Articles by James Boyk on Audio & Music
Recording Credits
James Boyk's Audio Consulting Services