At college we spent 3 years of torture learning all the maths, physics, putting in ET scales, passing test and exams etc etc, but right at the end in the final semester we were told to forget about all that rubbish; "you tune a piano this way". Tune the fourths and fifths so that you hear 2 beats, not 2 beats a second, but 2 beats. In other words the higher up the scale you go the intervals will appear to beat at the same rate even though they are actually beating faster. Of course this all depends on the sustain of the piano. The higher up the scale you progress, the shorter the sustain is. Now apparently this was the method of those tuners of yore who knew absolutely nothing about the maths and theory of equal temperament. AF ----- Original Message ----- From: BobDavis88 at aol.com To: pianotech at ptg.org Sent: Sunday, February 08, 2009 10:49 PM Subject: Re: [pianotech] Do fourths beat faster? Do ascending fourths beat faster? The answer is "Sometimes." Ironic that this discussion on fourths and fifths is just about to produce more heat than light, just at the time that something useful is beginning to emerge.... Hang on, hang on.... Given a choice, I'll obviously rather have my tunings sound good than understand them, but intellect and intuition are not mutually exclusive. They are both human attempts at understanding, but they do not change the nature of things; they are just approximations which we find helpful. I have huge regard for the human wetware, so I like to think of intellect as informing intuition, to make sure it has all the correct inputs it needs to function well. If the result is good, it doesn't matter how we describe it, but keeping both channels open surely makes the job more interesting. Sometimes our doing is ahead of our understanding; sometimes the other way around. That said, the math must itself be based on correct inputs. Calculations based on frequency tables may be correct to six decimals, but are not the information we need for tuning. (Because of inharmonicity, if A4 is 440.00 Hz, D4 will not be 293.664776, but something closer to 293, and their 3:2 coincidence will not be exactly 880 and 880.99xxxx, but something different depending on the inharmonicity of each note). Inharmonicity is exactly what this whole conversation hinges on, being the cause of fourths and fifths staying close to the same up the scale. Fourths will typically, but not necessarily, increase in speed very slightly as we ascend through the temperament area, but in no case will their beat rates come close to doubling each octave, and in most pianos, they stay close to the same above (and below) the temperament. This is counter-intuitive only if we don't have a full understanding of the effects of inharmonicity. I just re-read Dan Levitan's excellent articles which begin in the August 1994 Journal. I'm no math whiz, so for me they're heavy going. It took him several articles to explain this, so I can't summarize his reasoning in a sentence or two, BUT, some of the conclusions are that 1) Inharmonicity would cause ascending fourths to be wider (beat faster) and fifths to be narrower (beat faster), but 2) it also expands octaves even more, which more than counteracts this. 3) as inharmonicity increases up a good scale, the ever-wider fourths, ever-narrower fifths, and ever-wider octaves keep counteracting each other, keeping the beat rate nearly the same, at least at the lowest coincidence, depending on the tuner's choice of octave width. 4) Although temperament beat rates are not the same from piano to piano, most well-scaled pianos, even with different primary inharmonicities, produce very similar beat speeds, and similar (but not necessarily identical or linear) rate increases on the temperament thirds, because of the scaler's attention to the rate of change of that inharmonicity. 5) Some of our temperament tests which rely on equal beating, such as the inside M3 - outside M6, are unreliable at the finest level, since the speed of each of those intervals depends upon octave width, and they change either at different rates, or, in the case of tests using a minor interval and a major one, oppositely. Taking the D5-A5 fifth as an example: the inharmonicity of the third partial of D4 (A6) is greater than that of the second partial of A4 (A6), which would make it sharper, and therefore the fifth narrower EXCEPT that the A5-A6 octave is wider than the difference. On a typical scale, if A4=440 Hz, its 2nd partial (A5) will be about 2 cents sharp of 880, or about 881; its 4th partial (A6) will be around 10 cents sharp of 1760, or ~ 1770Hz. This will cause the 5th to be slower than if the octave were an exact 2:1 ratio. The third partial of D5 (A6), will actually wind up very close to 1770. I hope Dan will forgive (or correct) my (very) condensed (incomplete) version of his fine work. Bob Davis ------------------------------------------------------------------------------ Who's never won? Biggest Grammy Award surprises of all time on AOL Music. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://ptg.org/pipermail/pianotech_ptg.org/attachments/20090210/bb510b93/attachment.html>
This PTG archive page provided courtesy of Moy Piano Service, LLC