Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Aural Temperament requirement of the PTG exams ending?

I never said that the situation you are talking about was Reverse Well, so you wasted a lot of gas on that one, Emmery.

You have it now, A443. Reverse Well is alive and well. Perhaps I have been the reason that now so many tuners are saying that temperament is not as important as unisons and octaves. It only matters if somebody actually knows what they are doing and deliberately tunes a non-equal temperament. Then, it wouldn't work, couldn't work and shouldn't be tried but if Reverse Well is the result, then it is completely in consequential and ignored for what it is.

I stand by that.

I spelled it all out. Sorry it took so long.

I don't typically engage myself in these debates; you do fine enough arguing your points. I don't check other tuners tunings, I just retune them--so I have little in terms of actual observations to be able to state what other tuners are doing or not.

Just FYI, in general, I dislike color in my tunings, but I did start putting your EBVT III on one of my pianos today. It occurred to me, all of a sudden, that there could be a potential pedagogical advantage to learning ear training and the recognizing of pitches/intervals (ie perfect pitch) through a WT. So, I'm going to explore this concept/possibility with a pianist/teacher over the summer.

It always seems to me that those who think Reverse Well doesn't exist never looked for it or couldn't recognize it for what it is if they did.

The late, great, Owen Jorgensen who some people are trying to discredit now that he cannot defend himself, always taught his students ancient temperaments with pure and equal beating intervals first. That way, when they finally got to ET, they easily understood it and could do it well and properly. You will never find a more skilled aural tuner than from among those who graduated from Owen Jorgensen's Piano and Allied Arts program at Michigan State University.

The EBVT III takes after that idea of pure and equal beating intervals and so does the ET via Marpurg sequence that I developed. One clown on here described the latter as "Two steps backwards and one step forward" but many people who formerly could not pass the PTG tuning exam have now used it successfully and are now RPT's because of it.

I am teaching a young man now using both. My goal with him is for him to graduate high school as an RPT and for him to completely take over the tuning and care of the pianos which had all been neglected and tuned in Reverse Well at his high school in September and for him to be paid the fees that would have otherwise gone to an incompetent aural tuner.

He will get the job based upon the portfolio of work and the sound from the pianos that he creates entirely by ear and through his diligent work on them this Summer. He already knows how to recognize Reverse Well. He did before he ever came into contact with me. He listened to how the pianos had been tuned and didn't think they sounded right based upon what he had learned already through his own initiative.

Bill it does not show up where a A fork is used, so that makes yet a large part of the world.

If it exists here it is then in another form, but there are no so loose tuners those days in my experience; all know how to check FBI vs SBI, .

Bill,

Many thanks for your detailed write-up.

I have some questions about the errors and multipliers. You wrote:



OK, this I understand. There is 1 cent tolerance, after which 1 point is deducted per 0.1 cents error, up to a max. of 20 points. Similarly, in the last section, unisons, you once again listed the tolerance (0.9 cents), the error size (1.0 to 1.9 = 1 point, 2.0 to 2.9 = 2 points, etc.), and the multiplier.

But as far as I can see, in all of the other sections (temperament, midrange, bass, treble and high treble, stability) you only listed the tolerance and the multiplier, but not the error size per point.

So, my question is: for all these other sections, what is the error size per point? Is it the same as in the pitch (0.1 cents = 1 point) or as in unisons (1.0 to 1.9, 2.0 to 2.9, etc.)?

Thanks,
Mark

Isaac, Reverse Well is just as easily possible from a C (Do in Europe) fork. To say that all tuners know all the checks is a very broad statement. The fact that so many don't know them who come to take the tuning exam here is currently being discussed on the PTG Examiner topic (exclusive to PTG examiners). I didn't just pull that out of thin air. I know it is true from personal experience.

The same thing has been said, however on here such as in, "How could anyone do that with all the checks and everything?". I know them, you know them and probably most people who would write about their disbelief that Reverse Well is a very common error know them but certainly, those who actually tune Reverse Well such as at Lucas Brookins' high school do not know them.

So, how could it be that all the tuners in Europe know all the checks?

Thank you for your question, Mark,

The tolerance, error size and multiplier are confusing to many.

Pitch: The tolerance is +/- 1.0 cents. A reading of the examinee's pitch is taken. If it is a negative number, it is converted to a positive number, then 1.0 is subtracted from it. Whatever that result may be, it is multiplied by 10 and that is the Pitch score.

For example, if the examinee's pitch reading is -2.6, the negative sign is taken away and 1.0 subtracted from it. 2.6 -1.0 = 1.6. 1.6 X 10 = 16 100 - 16 = 84. The Pitch score is 84.

That is actually the equivalent of 1 point off for each 0.1 cents beyond 1.0 cents, so that is the way I usually express it but the above is actually how it is done. Therefore, the limit for a passing Pitch score is +/- 3.0 cents.

In that scenario, let's say an examinee's reading is 3.0. 3.0 - 1.0 = 2.0 2.0 X 10 = 20 100 - 20 = 80

For all other scoring, the tolerance is 0.9 cents and the spread (as you termed it), is also 0.9 cents. Therefore, 1.0 to 1.9 cents is 1 point. 2.0 to 2.9 cents is 2 points, 3.0 to 3.9 cents is 3 points, etc. Only in the Stability score is anything greater than 1.0 cents only 1 point, no matter how large the number may be. If the string slipped 5 cents, it is still only 1 point but each point has a 4 point multiplier in the Stability portion of the exam. Therefore, 5 slipped strings anywhere beyond 1.0 is the limit. 5 X 4 = 20 100 - 20 = 80

The Temperament score is fairly easy to understand. Whichever notes within the designated Temperament octave have points, those points are multiplied by 2.5. An error anywhere from 1.0 to 1.9 would still be 1 point.

What sometimes saves a point from counting against the examinee is nullification by aural verification. It sometimes happens if the error is only barely out of tolerance such as 1.0 to perhaps 1.3. If the scored error cannot be aurally confirmed by the examiners, it is nullified. The criterion is whether that note could be improved by either sharpening it or flattening it. Sometimes, there are conflicting choices. Some checks show it slightly sharp, some show it flat, some show it about right. In a case like that, it is nullified.

If the score has a half point due to the 2.5 multiplier, that score is rounded up to the next whole number. Therefore, a nearly perfect temperament with just one scored and verified error would receive a 98 (not a 97.5) but two such errors would result in a score of 95.

The Midrange score is often confusing. It counts errors in the entire midrange, including those in the designated Temperament octave but as it does count those, it only gives them a 1.5 multiplier. Therefore, it sometimes happens that the Temperament is passed but the Midrange not or Midrange passed but Temperament not. I am running out of time this morning to explain how that can happen but I will later if you or anyone wants me to.

The Unison scoring can also be confusing because each unison can have from 1 to 3 errors scored. I have been asked any number of times why there was double and even triple jeopardy but I didn't make the rules. The tolerance is once again 0.9 cents and the spread is 0.9 cents, so any value 1.0 to 1.9 is one point, 2.0 to 2.9 is two points, etc. Left to center, right to center and left to right are each measured and counted.

A unison could have two errors below tolerance from left to center and right to center but if one outside string is sharp and the other flat and the difference is 1.0 cents or more, an error is counted. Example: left to center is +0.5 and right to center is -0.5. Neither is even close to tolerance with the center string but the left and right string are 1.0 cents apart, so that is one point and has a 2 point multiplier, so just one such error would result in a score of 98.

In the scenario I gave earlier, left is -1.0 and right is +1.0 so both strings are just 0.1 cents out of tolerance but both of them get a point and are multiplied by 2. The left to right difference is 2.0 and that gets two points. 4 X 2 = 8

Normally, a unison that has just one string 1.0 cents off does not sound terribly bad but does have an audible but slow beat. It is probably what we may often hear with our own tunings when we would check the unisons over a final time and it would signal us to correct it. You could have 10 out of 24 such errors and still pass.

Once, I was curious to know what a unison with a left string at -1.0 and a right at +1.0 would sound like. I tuned a unison that way to find out. It is raucous!

Bill I said that most countries here use a A fork, so if you see a lousy tuner that only use 5ths and do not check FBI (it could happen some decades ago) he obtained a reverse Well based on A , not C

Now training is of a general good level, because of state diplomas, relation with ministry of education a more tight control , Europiano etc.
The basic checks for all/most intervals are learned, without using mixes as 6:3 4:2 generally but with a basic control on any interval if wanted.
Just with that it is difficult to go far from an acceptable ET.
For what I have seen bad tunings are more the result of tuning in 45 minutes.

Bill,

I appreciate your taking the time to write such a detailed response.

I understand how the temperament could fail and the midrange pass, or vice versa. It's because they use different multipliers. The one for the temperament (2.5) is more punishing than that for the entire midrange, or at least the temperament part of the midrange (1.5).

Thank you very much - I think this gives any reader enough information to score/rate his own aural tunings with an ETD. In my case, I'm particularly interested in evaluating my temperament, as I'm currently experimenting with different sequences. I want to get a feeling for the relationship between a certain degree of success in my aural tests (e.g. inside M3, outside M6) and the resulting score. I hope that this will show me where to improve.

Oh, and I also seem to remember that we have an old thread somewhere that details the exam quite rigorously, methinks it was written when Patrick Wingren took the exam. (My aim is not to study towards any exam, but to use a reputable yardstick in my self-evaluation.)

[Edit: I realise full-well that I will not be measuring my efforts against a master tuning, but against what an ETD thinks. Nevertheless, I think it's a good starting point to improve my accuracy.]

Then the scoring is verified by the other two examiners who are in the room...

You stated that what adam posted is an "explanation" for how
RW temperament can result from an ETD only tuning. How do you come to that conclusion? You have no way of objectively knowing how far out that tuning was at the recording studio to make that kind of connection.

Modern ETD's do not make enough of an error on ET templates/calculations to produce Well or Reverse Well temperaments....not unless they are deliberately programmed to with offsets to give RW. That postulation is so silly that if it were true, there would be no reason for having charts showing offset values from ET (to be used on ETD's)to produce other temperaments.

One of the biggest arguements in favour of ETD use for even aural tuners is the fact that it acts as a foolproof barrier preventing a temperament from accidnetly resembling something other than ET.

It would be possible if ETD create a temperament by stacking 5ths. But that would be interesting to have that sort of ETD, that could tell you how much is left when the last 5th is attained,
You also could ask for a progression of 5th sizes.

I think it is a good idea. The same with M3ds also why not.

Thank you, Bill and Ron for your posts !

You're only broadcasting around the world what you don't think.

Mark,

What you are saying you will do is what is highly recommended. You really can't do better than that. It is true that Master Tuning Committees will ponder an ETD generated tuning, even from such touted best sources as the Verituner for up to 4 hours but really all they are doing is splitting hairs.

If you accept it as true that the ultimate ET is an infinitely unobtainable goal, then it stands to reason that a committee of three will spend up to four hours refining what is already quite acceptably perfect. No one can ever achieve that level of refinement on his own nor can yet any device do it. But really, the quest for that perfection is only to provide a super human model so that when a humanly created tuning is compared with it, the very fine errors that sometimes do get scored can be properly nullified.

The difference from what the Verituner could do on a well scaled grand and the "Master Tuning" on the same piano would not be recognizable by even the finest musician, now would the "Master Tuning" actually make that piano sound "better", at least as far as temperament octave and midrange tuning go. There would also be very little difference if any tat would matter elsewhere. The real challenge and difference that can be made is in what everyone says these days and that is in the quality of the unisons.

(You really have to consider that if I have been tuning pianos in a temperament that would FAIL the PTG exam for some 25 years, then PERFECT equality in temperament is truly not the most important nor the ultimate goal. It is just what the PTG tuning exam is all about and nothing more than that).

No doubt, if the preliminary tuning generated by the Verituner were run against the ultimately decided Master Tuning, it would score no errors Nd that is because of the tolerances.

You have to be sure bout what you are doing, however. If, for example, you only want to examine the Midrange (C3-B4) you have to tune the piano as precisely as possible to what the ETD says and you need to have a very stable environment. Then, because the partial selection for the exam will be different from the ETD calculated tuning, you need to read and program the results of the ETD calculated tuning into the Master Tuning program that the ETD has on it. That also has to be done with the utmost precision and it requires skill and experience to do it.

Then, you have to tune that range aurally, read and store the results into the Examinee page of the ETD and following the ETD manual instructions, lay those results over the Master tuning and use the "Score" function.

The Score function will instantly add all of the Master Tuning Temperament octave values together and divide by 13. It will also add all of the Examinee values together and divide by 13. It will subtract the the dividend of the Examinee tuning from the dividend of Master tuning and apply that small number to each note of the Examinee tuning and subtract it from the Master tuning value for each note. That will provide the basis for a scored point or not for each note.

When you move the ETD to the bottom note of the Midrange or Temperament octave and press "note up", it will stop on each note for which there has been a scored error. When you find a scored error, you can use your own known aural checks to determine if it is an error or not. Since your "Master Tuning" will not be as perfect as one may be, you could find that a scored error does not seem wrong at all and you may nullify it.

Unless you are doing all of the above, however, you should not be too worried about whether your tuning matches what the ETD says is right. You have to have the proper controlled conditions before any kind of evaluation such as that is anywhere near to being valid.

I assume that you are a musician and have a good ear. If all of your intervals sound right, you are quite probably well within the passing range of the PTG tuning exam.

So, as I see it, technicians such as Chuck Behm and Jim Iallegio who truly do excel in other areas of piano technology, are only asking to be included among the ranks of piano technicians who demonstrate minimum sets of skills.

They are truly professional piano technicians and I have no doubt that what they do when they tune a piano meets and far exceeds those standards. [/quote] …but…
This brings up a point I have made repeatedly ever since the idea of standardizing the testing of technicians was first being discussed: there are two distinct sets of skills involved here. We need to recognize this fact and develop two different types of exams to evaluate them.

It is rare, I think, for a single individual to excel at both tuning and at rebuilding/restoration. There are some, to be sure, but there are many more who are clearly more proficient in one area than they are in the other. And yet we insist on clinging to one test that is not ideal for either.

I would suggest that for those whose skills lean more toward tuning the current tuning exam is probably adequate. I’m sure it can be improved—can’t everything?—and it may well need to evolve with the times. I’ll leave that for others to debate. But I am continually surprised—and dismayed—at how much trouble some folks have in passing those portions of the exam intended to demonstrate a working knowledge of how pianos work and basic technical skills when I consider both of these to be utterly simplistic. So, I should think, do most of those whose skills lean more toward major repairs, rebuilding and restoration.

The obvious solution is to develop two separate but equally qualifying exams to more adequately test the skills of each of these specialties. Each of these exams should be structured so that ii is challenging to each examinee within his or her skill sets. That is, one of the exams should be biased to challenge the tuning specialist while the other would be biased to challenge the more technically oriented examinee.

It would not even be necessary to change the initials tacked on at the end of one’s name: RPT could indicate either a Registered Piano Tuner or a Registered Piano Technician.

ddf

Del,
I tried to get something like this in the 1980's. The committee tore it to shreds. I wanted to have RT for Registered Technician, and RTT for Registered Tuner Technician.

The rebuilder test would be peer review. A committee would view examples of the applicants work over a couple of years. There would be a checklist to grade with. The RT would test for three separate skills:
Refinishing.
Rebuilding; including replacement of board, bridges, pin-block and strings.
Action rebuilding; which would include fitting an entire new keyframe, action, hammers and dampers into a grand.

This was before we switched RTT to RPT.

The excuse is that it would be too difficult to formulate and administer a reasonable exam geared primarily toward technical skills. To which I say, "poppycock!" It would be difficult, to be sure, but so was coming up with reasonable tuning exams. Mostly it is a lack of leadership, vision and political will.

ddf