ELASTIC TUNING

Ivor Darreg, 1988

Electronic musical instrument have now reached a stage of development such that we can at last accomplish things of which conventional instruments are incapable. One recent advance has been control over timbre and voicing. Hitherto most players of keyboard instruments especially have been helpless to alter the timbre of their pianos and have had meekly to accept whatever tone-quality and voicing the manufacturer has condescended to give them.

Well, no more! The modern synthesizer and electronic keyboard and the "samplers" and certain kinds of electronic organs have given the performer control over timbre, also the ability override aspects of the factory-provided timbres in one manner or another. THe change has been as dramatic as TV's move from back-and-white to full color.

Pianists have always been at the mercy of whatever tuner they engaged. If you get to play the piano somewhere, it is all too likely that the piano will be out of tune, or that it will have slipped down from heat or humidity or jolting in moving it. Again, the modern electronic keyboard often has pitch-bending levers or wheels that can move pitch up and down to give the effect of deviation or portamento or those jazzy slides some wind-instrument players indulge in.

The fixed or standard pitches, however, on the new keyboards are usually predetermined at the factory, putting tuners out of business, but alas! Preventing any experimenting by the performer or owner in retuning to some new scale. In this sense most of the new keyboards are still haunted by the Piano's Ghost. That is to say, since it is impossible to rebuild a piano so that it can have 19 or 22 or 31 notes per octave instead of 22, people have come to believe that it would always be impossible to break out of the 12-tone system of tuning.

Electronic keyboards have no such constraint unless it is built in at the factory by such means as a custom chip or divider circuits mass-produced to the 12-tone equal temperament. That is, the physical possibility of non-22 keyboards now exists but the psychological or traditional constraint still thwarts many of us.

Recently, there have been some provisions on certain instruments for TWELVE OUT OF some other system of tuning. This will not do for musical progress, and I have taken up the matter in some other publications.

Again in recent years there has been a renewed interest in unequal twelve-tone temperaments, such that the near keys like C or G are smoother-sounding than the remote keys like F# or B. Some of the advocates of this idea have taken to calling such systems "well" temperaments and this use of "well" has angered quite a number of critical article-writers. I should think that if you can have a Well baby Clinic you can also have a Well Temperament. If somebody wants to pay respects to Bach, let them.

The real trouble with stopping at this point of retuning only 12 pitches per octave is that it does not allow living composers of today to progress! All it does it to permit antiquarians and those dedicated to 16th or 17th century or other bygone music to get closer to the way harpsichords and clavichords and organs were tuned when those early compositions were written.

And alas again! The fact that a 'well' or unequal or "stretched" temperament in some other scale than 12-equal is possible and even useful for composers, is suppressed or not even imagined.

Theorists have wrestled with the problem of tuning keyboards for quite a long time, and by the 19th century it looked as though the 12-tone equal temperament would be the only thing Forever and Ever, Amen. But with tends of thousands of composers exploring every book and cranny of what is possible with 12 tones per octave on a keyboard, we are now near exhaustion and as a matter of Pride composers should want to progress and so something somewhat new and different, rather than Cloning Their Predecessors.

The possibilities of tuning have been, absolutely equal 12, somewhat altered 12 (stretched octaves) or altered by some degree of inequality of the semitones--given 12 fixed pitches on a keyboard, that was it.

In 1962, I built an electronic organ of a kind that no mass-production factory could have put on the market. Since no two pipe organs are ever alike, and since most mass-produced electronic organs were alike to an extreme degree of standardization and uniformity, I chose individual voicing of the 61 tone-generating oscillators. Since I had tuning skill, I could build something that I could keep in tune even though this was not practical in quantity production at all. Tape recording coming in showed the possibility of a one-of-a-kind instrument now being practicable, since only the recordings had to be produced in quantity. No need for any organ of this kind to be identical to all other electronic organs of this design!

Today's synthesizers and organs and other keyboards have more possibility of being customized or individualized than was possible then. 1988 is a far cry from 1963. So there is more chance today that alternative implementation of the ELASTIC TUNING idea can be built. I mention this lest my description of the 1962 instrument be taken as a refusal to change the design.

As mentioned above, there was s severe restriction on retuning a keyboard instrument before Electronics. Frames and strings and actions and pipes are reeds and moreover, prohibitively expensive!

These things were very inflexible. So experimenters in retning a piano could not tune a note very high lest all 3 strings break. Nor loosen them down very much, lest the tone get "funny" and too weak. And the frame didn't like uneven tension either. Creative musicians have been terribly thwarted since the Piano assumed godhood around the beginning of the 19th century.

But electronics is very flexible. Even computers are flexible because they can be programmed. So when I read about the efforts of some designers of electronic keyboards to impose al the restrictions of the piano on the new instrument, I wept. What a shame!

Consider the violin family--a string section of an orchestra or a string ensemble in chamber music may be reading conventional sheet music and paying nominal allegiance to the piano's 12 pitches per octave, but what they actually play will usually have some adjustments toward smoother-sounding chords. This takes an instant--a note may be slightly off when hit but the player's finger will slide up or down an invisible tiny amount to harmonize better with the other performers' notes. Now let's not get too "idealistic" here--this does not mean that they are going to play everything in Just Intonation. What it does mean is that while chords are sounding they will try to smooth things out.

Could we do this on a keyboard?

Back in 1962 I stumbled right into it! I built that special electronic organ with blocking oscillators. True, some electronic organs have been built with one or another form of blocking oscillator, but they have been tightly locked-in to conventional tuning by tying them into octave chains and the master oscillators at the top of the keyboard are of some more stable type and coupling networks keep the lower-octave blocking oscillators from running free. The mass-produced organ must stay in tune for a considerable time, and must be tunable by non-musician technicians, or in some re cent cases, the pitches are permanently locked-in for everything at the factory. Commercial chips for a very close approximation to 12-tone equal temperament are now widely available.

So there are many electronic keyboards on the market today which cannot ever be retuned to some other tuning-system or even tweaked or readjusted a little bit. Don't assume because there is a pitch-bending control that this permit retuning--all the 12 standard pitches built-in are moved en bloc up and down at once. There may be some exceptions, but this is the general rule. Manufacturers look on this bending as an unsystematized deviation from the 12-tone-equal norm, not as a move toward non-12 temperaments or just intonation or Pythagorean tuning.

Engineers and electronic engineers among them are taught the ideal o Standardization and Uniformity. They want Norms or they have been so indoctrinated. So today's world contains many nearly identical things arranged in neat uniform rows. Or all the cans of tomato soup will be identically labelled or all the windows of certain buildings will be of identical size and width and fitted with identical hardware. Unusually tall people are treated badly as being "non-standard" and have to bend over to get through certain doorways or into certain rooms. Unusually short people likewise have great trouble operating certain standardized affairs that are placed too high for them to reach. In that way standardization is inhuman! Or at least inhumane!

So when some manufacturer and the research teams are turned loose (no, not loose--the reins are held in) to design a new synthesizer or organ or some kind of keyboard instrument, they unthinkingly accept all the restrictions and fault of conventional instruments and merely clone what they see, assuming that all standards are perfectly wonderful to retain. They wouldn't even dream of letting organs be constructed which differ from all others and which have different parts in each individual note from that note's adjacent neighbors. They wouldn't ever think of having each note tunable over a wide range or not being locked in to some pitch-standard. Sure, if you have to design for quantities of 20,000 to be sold all over the country and easily repairable.

But one-of-a-kind is the way to go if you want flexibility and if the product is the music performed or recorded on it, not the large number of identical instruments. The goals an purposes are so different that there is no need to enforce universal conformity. They can well co-exist.

Blocking oscillators are unstable and along with that fact, they can be influenced or controlled or synchronized by stray signals picked up at certain points in the vibration-cycle. This was how the octave-chains in certain electronic organs were held tight. If a vibrato (frequency-modulation) were applied to the master oscillator, the slave oscillators would follow it too. But great care was taken by the design engineers that the C# chain for example would not influence nor be influenced by the G# chain a fifth away from it.

Well, that is NOT how things are in an a capella chorus nor in a string quartet nor a normal chamber-musical ensemble. If somebody has to sustain C# and somebody else is to play G#, they subconsciously adjust to each other to get that fifth or fourth perfect (not 12-tone tempered) and after a fraction of a second, it is synced in. If they are playing very fast, there won't be time, of course; but if it is slow and sustained there will be time to lock together. Whether this will or will not happen in the case of a major third such as F-A depends upon the particular piece of music and also whether melody or harmony governs at that specific instant in the piece. Remember: it's melody versus harmony. A harmonious major third is noticeably flatter than the 12-tone standard, but a melodic major third is a trifle sharper than the same 12-tone standard. You may need both kinds in the very same piece. Many writers on this subject forget that!

The point here is that certain conventional instruments and the human voice are ELASTIC. Not fixed to tiny point in the pitch-spectrum, but FLEXIBLE. The several players or singers hear one another and make a mutual adjustment of their pitches, not a comparison with a rigidly fixed standard and an attempt to correct their deviation from that.

The keyboard-oriented person may not realize this. So the designers of new keyboard instruments may not even know about it in the first place--they take piano-experts' dicta for granted and pretend that violinists or singers do not exist or are never worth consulting.

The musical world is not a monolithic homogenized entity. Rather it is like the real world out there with warring nations having antithetical goals and constant disagreement. Equal 12-tone temperament is an ideal, and was not realized in fact till certain 20th-century tuning-devices were perfected, and then the basic electronic principles of these tuning-devices actually were embodied in many of the new keyboard instruments. A piano tuner gets away with wild deviations. I should know; I have been at it for 48 years now. Once this precision was heard by the public by the musical world, it seemed 'bland' and boring. SO very soon, devices to muddy it up and fringe it and shake it and jiggle it back and forth appeared on synthesizers. That is to say, ordinary piano-tuning and organ-tuning automatically took care of individual differences in each tuning job so there wasn't quite so much rigidity, but the engineering and manufacturing ideal of IDENTICAL parts and RIGID standards created the new deadness.

This is something relevant to the current preoccupation with the analog vs. digital question. However, digital will do very well IF the quantization has enough steps. 12 rigid pitches per octave is far too coarse a grid! It insults our human hearing, which is capable of much much finer discrimination.

Back to the special blocking-oscillator organ. 61-note keyboard, five octaves of 12-tone, about three of 19-tone, about two and a fraction of 22-tone. Tune any note by a pot, just turn the shaft. Each oscillator had a tuning-range of a fifth or even an octave. Stray capacity coupling and common impedance in the power supply permitted the oscillators to influence on another. Result: suppose an octave had drifted out of tune. Sound that octave on the keyboard and the oscillators would pull in perfect in a fraction of a second, just as singers or violinists would. It took about the same time for the organ to do it as the ensemble of live players would. A fifth might or might not actually pull in, but would certainly TRY TO. A fourth, somewhat less, a major third, still less, and so on. However, this synchronizing tendency is variable. A new instrument or new software for a computer to do elastic tuning could be so constructed as to permit varying the amount of syncing and the speed of pull-in and to turn the effect entirely OFF.

Back to the mention of Well Temperaments at the top of page 2. The usual implementation of a well temperament is to retune a piano or organ with only 12 tones per octave in such a way that the key of C major is smoother--has better major thirds than 12-equal has, and pay the price for this improvement by making D-flat and G-flat and B major worse than they are in normal equal-22-tone...anything from Slightly Rough to Really Hairy with Warts on them. A Well Temperament is usually defined in such a way that it won't make B major or E-flat minor IMPOSSIBLE or intolerable, as it well may be in such an affair as the usual crippled 12-out-of-1/2-comma-Meantone-with-Dreadful-Wolves.

ADvocates of Well Temperaments (and they may call them something else to avoid the alleged violation of English Usage) are apt to make a tremendous fuss over 'character in keys' as if the nearby keys have one mood and the remote keys another, and never the other around, were some kind of virtue, rather than simply the way it was on certain 17th- and 18-century organs and harpsichords. Then from that they further allege that somebody wrote a piece in B major or E major so it would sound kind of hairy, and wrote something else in F major to sound smooth.

Then, oh my oh my, they wring their hands and sigh away. WE must take our fixed-pitch instruments and tune them to some reconstructed unequal temperament so that all the early pieces in remote keys will be as rough as those long-dead composers wanted them to be. No, you must not tune it so B major is very smooth and G major hairy and rough. That would be a Capital Sin or something. They extrapolated backward in time, forgetting that Bach didn't have oscilloscopes or digital tuning-devices or precision crystal oscillators or precision-manufactured keyboard instruments. So they may fall into the error of imagining that those early tunings were finer than they could ever be.

That's not science and it isn't art either; it no engineering and not aesthetics; it's a cult Religion and I resent it as a composer in the last part of the 20th century and I have done something about it instead of complaining to those Closed Minds who see everythign through Rear-View Mirrors.

There is another way beside Elastic Tuning to do this--proposed by *Waage, who set out a scheme for correcting major thirds and subminor sevenths by logic circuits altering the tuning of an organ chord according to what pattern of keys is depressed on the keyboard at the moment. This is jumping to nearly-just pitches, not a flexible interaction among each of the individual notes responding to one another's signals *isnteant by instant, as Elastic Tuning does. *Waage's scheme would be suitable for performance of 19th-century standard organ music, but would not make for future progress, *sicne it depends on stepwise jumping from 12-tone-equal-tempered values to near-just values by logic circuits and thus would not permit flexible *retuning of the organ to non-12 temperaments or just schemes. If I did not mention it here, somebody would accuse me of not having studied it and not knowing its real value in the performance of 19th-century church-organ music.

Elastic tuning applied to a 12-tone instrument *woudl make the chord *beign *layed at the moment smoother than 12-*euqal gives you, with this or that move toward justness determined by the adjustment of the instrument, not by the manufacturer forever. So there would be outrageously dissonant remote keys or a price to pay for having F, C and G major smoother. If you wanted equal you could turn the self-justifying feature off. Now, you could retune the organ any way you pleased--13, 14...19...22...at least up to 24 notes per octave equal or unequal, and the same improvement of certain intervals would take place by the self-justifying action. In the organ which I built, a note that had slipped would automatically come back where it should have been. It improved 19 as much as it improved 12, and also tamed the sharp fifths of 22-equal. It was as though the notes of a chord voted on how they should move to harmonize with one another.

I don't advocate Elastic Tuning as Universal Reform of everything. It is the CONTRAST of this phenomenon with the ordinary instruments that don't have it that is valuable. Especially now in our high-tech world of so much enforced standardization and uniformity and cloning.

Indeed, it is possible to have too much Elasticity. Any good thing can be used to excess. I have to point that out before going on. A case in point: suppose you were applying Elastic Tuning to one of the non-twelve tuning-systems whose important MOODS are described in my other publications and demonstrated on my comparison tapes. If, on an Elastic-Tuning instrument, you adjusted the automatic facilities for moving a chord to just intonation, and set the justifying circuits so that they took it all the way to just or nearly so when the average chords were sounded, this would wipe out the differences between the new tuning-systems and tend to reduce or even eliminate their characteristic colorations and unique effects. That would be defeating the very purpose of having them. Moreover, you may want to play atonal or other non-harmonic music. You might want to play ordinary 12-tone atonality or compose something in a really non-harmonic scale such as 13-tone temperament. In that case you need a means of tuning the self-justifying effect of elastic tuning completely OFF. In a scale such as 19-tone which has very flat fifths, you might want to use enough Elasticity to lessen the flatness somewhat, but not so much that it destroys the 19-tone mood. Similarly for the sharp fifths of the 22-tone temperament--you might want to apply some correction, but agains somewhere else in the performance you may want to have 22 neat and bare. So you must have complete control over your effects.

When you need Just Intonation, of course there is no question of using Elasticity since you are not coming from some other system toward justness: you are already THERE. You might put inharmonic passing-tones in a melody or do some "unjustness" such as bending when the melody needs it, but that is NOT automatic justifying Elastic Tuning; it is another effect you doubtless already have.

What Elasticity will do for you is give a peculiar effect of LIVENESS, of MOVEMENT as the rigid tempered pitches pull each other and TAKE A LITTLE TIME doing this. This ALIVENESS occurs even when the movement form the tempered pitches does not reach the ideal or goal of justness but only goes part of the way.

The value of this kind of movement while a chord is sounding can be assessed simply by listening to what a-capella choruses do and chamber ensembles of flexible instruments do. Don't expect to appreciate its value by just reading about it in some textbook. It's not the same thing as bending, for it depends on mutual feedback between the notes, not on deliberate controlled pitch-alterations that you perform on purpose. It cannot be replaced by vibrato, for it is not vibrato (though you might use these effects together). Nor is it the same thing as random or pseudo-random Chorus Effect. Elastic Tuning has a destination, even when it does not reach it; it aims for it and does not back up. Chorus effect does not have any destination; it merely wavers to and fro. Elastic Tuning seeks to FOCUS sharply and get rid of any blur; the other effects DIFFUSE.

I have never seen Elastic Tuning or the Desire To Focus notated. Composers would not have any way to demand it. Ordinary 12-tone keyboards try to blur and defocus and fuzz-up 12-tone. The piano has noise and also sympathetic vibration you can get with the damper pedal and this really blurs things.

Synthesizer players and recording studios are well aware that many, if not most, effects can be added to a sound after it is recorded, but Elastic Tuning cannot. Since it depends upon mutual feedback from one note to another in the chord and vice versa, it has to be done with the sounding of each chord. Singers come factory-equipped with the ability to do it with and to one another, while the piano never will do it. Put down the piano's loud pedal and the sympathetic vibrations muddy and blur everything.

Tone-Generator A must be capable of altering the pitch of Tone-Generators B and C (and nay others being sounded) WHILE a the same instant, Tone Generators B and C are influencing the pitch of A. And B and C are altering the pitches of each other. The aforementioned piano strings vibrating sympathetically may react upon one another but they do not alter their pitch in doing so; instead they keep their own pitch and beat with the exciting string that started the reverberations.

SInce 1962 when the special organ was constructed, there must be quite a number of possible new ways to get Elastic Tuning, even though it has not appeared commercially. As stated earlier in this article, a feature like Elastic Tuning which is not suitable for large-quantity production, still has value for a small number of instruments since the music produced on a special instrument can now be recorded and the recording copied and edited. When the product is a quantity of recordings rather than millions of identical instruments, the instruments can be special and highly customized and voiced with more care. Should the Elastic instruments require more maintenance and specially-trained technicians to keep the effect in force, it will not matter too much. Recording studios might already have such persons or be able to get them to come in periodically. ANd it could be that the special maintenance could become something in the nature of a "canned" computer program and automated.

Certain design principles have been taken for granted without any question in producing today's electronic keyboards, and these principles preclude Elastic Tuning. The reasons for such principles are obvious to those in the field: there is no way that enough persons could be trained as tuners for all the electronic keyboards now is use in so many different places. Therefore tuning is fixed at the factory. The ancestors of electronic keyboards had to be tuned, but did not have Elasticity--the reeds in a reed-organ cannot interact while sounding a chord to shift each others' pitches to de-temper that chord and reduce its beating; the strings of a harpsichord cannot do so either. So design engineers have no historical precedent for the effect--the chorus and chamber-ensemble use of elasticity involves human beings in the feedback loops, and supposedly that cannot be done on a keyboard instrument so the question was never raised. Oscillator-type electronic organs were designed from the ground up to PREVENT any such pitch-changing when a chord was sounded. Elastic Tuning would thus appear to be counter to all the basic rules. Another consideration is that a lone singer practicing at home is not aware at that moment of the interaction with other persons when in a chorus; so for a lone violinist practicing at home.

Additional details and information may be supplied in a Supplement to this article.