Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum FAQ - Tuner-Technicians Forum Tuning by Ear or Machine

I add only that, IMO, you don't have to be good at anything more that aurally tuning unisons to tune well by machine. There is huge amount of skill level in making the jump from clean unisons and computed stretch, or drift, and coming up with an appropriate amount of stretch on your own. Actually, its this belief that I think has put quality tunings in reach of non-professionals. I find the limiting factor, in my case, is tuning stability.

As an aside, I find my tuner who tunes for the BSO and some of the local colleges, uses zero, or no stretch, in the bass and only adds it on the right side of the keyboard. He tunes aurally, FWIW. I have gone back twice with my ETD to sample.

My beef with machines is they don't seem to deal well with those bass notes whose upper partials ring loudly vs those they can hear, that I can not. Sometimes the fundamental is beatless, but the partials take on an annoying ring. I forget whether this is falsing, but find ears help where the machine does not, even to the point of leaving the fundamental imperfect to fix the beat in the upper partial.

Chris

I've never had problems like that with my hearing. I don't do a lot of tunings in a single day, almost never more than three, and I am never affected by the unamplified sound of a piano, except maybe if I stick my head under the lid of a concert grand when someone else is playing.

[1] This is, of course, the assertion of many who wish to bypass the work and practice time involved in learning to tune aurally. Just buy a machine, spend a bit of time on unisons and hammer control and I are a tuner. I’ll not say it can’t be done, but it is rare. So rare that I’ve not found a piano tuner yet that I would hire who has not learned to tune a piano acceptably by ear without the aid of an ETD.

Now, once those fundamentals have been learned and become second nature, I’d agree that many, if not most, tuners can enhance their work, make it more consistent and certainly less stressful by incorporating the wonders of modern technology. But I’ve evaluated too many tunings that have gone astray simply because the tuner could not identify the problems by ear and could rely only on what the machine told him (or her).

Working out an appropriate amount of octave stretch is not really the problem. As the skills and fundamental knowledge of basic piano tuning are acquired things like applying an appropriate stretch to a given scale become second nature. The problem with depending on an ETD to be the last word in things like octave stretch is that inharmonicity is not a constant throughout the scale. Indeed, in most pianos it is quite erratic. Which note, or set of notes, is the ETD going to use as a basis? The ear, once trained, handles these irregularities easily. It constantly adjusts and adapts the stretch to compensate for even the poorest example of string scale design. Only the Verituner claims to accommodate these irregularities — the jury is still out on this.

{2} You might well find the stretch going in the opposite direction down in the low tenor and bass. That is, the fundamental pitch should be going a little flat of the theoretical ideal.

The problem is that the inharmonicity curve should not really be a curve at all but, when plotted on a log scale, a straight line. In the real world, however, it is not. Nor can it be. It will be highest at C-88 and will decrease to some minimum down somewhere towards the lower end of the tenor bridge. (This minimum would occur at the lowest unison on the tenor bridge were it not for the ‘foreshortening’ of the last unisons typically found on most traditional scale designs.) From this minimum it typically goes up some with the last few unisons. Then into the bass where anything can happen. Usually it goes back down because, for some inexplicable reason, most traditional piano makers insist on using quite small core wires on the upper bi-chord wrapped strings. Finally, it goes back up — usually rather dramatically — because those same manufacturers like to use rather larger than optimal core wires down in the mono-chord section. In between there is little uniformity since no effort at all was made to smooth out the inharmonicity curve through the bass section. Probably because inharmonicity itself was not at all understood until well into the twentieth century and few really new scales have come along since that time.

While all of this becomes second nature to the ear of the experienced tuner it becomes a veritable minefield for the ETD which plods straight through the whole mess oblivious to how the end result actually sounds.

[3] The closer the scaling gets to ideal the better the machines work. Unfortunately most string scales are far from ideal. In the end the tuning and the piano must sound good to the ear. While I have great respect for the progress that has been made in developing the modern ETD I’ve not yet seen one that can really replace the human ear in this regard.

As a disclaimer I should mention that I learned to tune by ear and worked that way for approximately the first twenty years or so of my tuning career. I then began to incorporate various ETDs into my work, came to like them and have worked with one ever since. Having said that I would much rather trust my piano to the average ear tuner than to a more experienced technician who could not be bothered to learn the most basic fundamentals of his or her craft — the simple technology of piano tuning.

Del

I think a lot has been said about the finer points of tuning and how an ETD effects these points but not a lot has been said about other reasons why one tuning sounds better than another. It could have nothing to so with the method used. It could be the level of skill the tuner has with the physical aspects of tuning. It doesn't matter if you can put the piano in tune, if you can't set the strings so that they stay in tune even through fairly hard playing. If your hammer control is no good, then your stability probably isn't any good. You can bend those pins until the note sounds in tune, but if you don't know how to set the string, as soon as someone gives the note a good shot, or even as you tune the unisons, the note will go right out. I think that the hardest thing about being a great stable tuner is fine hammer control. Being able to move the pins in very exact increments in the block to achieve the highest level of stability.
The honest truth is that I could probably go to someones home and tune a sub par temperament, but if the octaves and unisons are good then they probably won't even notice if the stability of the tuning was good. (I really don't do this.) However, if I went and tuned someone's piano and did a lousy job of pin and string setting, they will be calling me two days later because they have unisons wining all over the place. My point is basically the same as all of the other technicians here. I have heard fine tunings done with both methods, but neither is any good without a mastery of hammer control and good stability.

Thanks for all the replies. I just had my first piano tuning, and was impressed/surprised that the tuner did it all by ear. I had assumed that "tuning" a piano was just a matter of setting each key to the right Hz frequency (like the A above middle C is 440 Hz) and that was it. Thanks for all the information!

I wrote:
[1] I add only that, IMO, you don't have to be good at anything more that aurally tuning unisons to tune well by machine. There is huge amount of skill level in making the jump from clean unisons and computed stretch, or drift, and coming up with an appropriate amount of stretch on your own. Actually, its this belief that I think has put quality tunings in reach of non-professionals. I find the limiting factor, in my case, is tuning stability.

DF writes:
[1] This is, of course, the assertion of many who wish to bypass the work and practice time involved in learning to tune aurally. Just buy a machine, spend a bit of time on unisons and hammer control and I are a tuner. I'll not say it can't be done, but it is rare. So rare that I've not found a piano tuner yet that I would hire who has not learned to tune a piano acceptably by ear without the aid of an ETD.

Or, Del, it is the opinion of a consumer, who is the only one who has to be satisfied with his work. I guess I won't win a popularity contest with the aural tuners for the above comment. Take heart, I am an "amature at large", anyway, so most probably just moved on to the next post.

My needs are a basic tuning with some stretch and that is my idea of quality. I have found what I previously described to work quite well and volunteer it here. YMMV.

[2][3] Thanks for the feedback. You know your way around this stuff and it makes for some interesting reading. It doesn't, however, change my view that, as you so sophisticatedly described, the subtleties of inharmonicity and scale go *beyond* what the "average ear tuner" can hear, as well as what ETD's can currently be used for. These are truly subtleties. I, for one, am keeping my imperfect scale. Thank you.

Chris

If you tune aurally and check on a variety of intervals, the accuracy with which you tune is remarkably good, at least as good as claimed by ETDs. It's probably at least as good as the best tuner's hammer technique. And the fact that you are tuning with a number of intervals guarantees that all those intervals sound good, which means that you have the proper stretch.

The advantage is that there is nothing to calculate. It's just the way that it is.

Stretch is not something you do, it's something that happens.

Setting temperment and stretch have nothing to do with one another. If you are achieving your intervals in the center of the keyboard aurally, you haven't even begun to accomodate inharmonicities in the bass, for instance. You will simply have a nearly even 440 - 880 octave and a clean temperment, assuming you got that far. If you guys honestly end up with a frequency that is slightly wider, and yet in proportion with the inharmonicities on down the keyboard, then my hat is off to you. A simple description of your technique would reveal how you are accomodating stretch. If you call yourself done after the temperment is set and you haven't played a good number of bass notes, then you can't possibly have got it right. If you go deliberately wide by smidgen, then you just took a shot in the dark as to what that particular piano might need. Maybe you go back to your temperment to add stretch latter, I don't know and am not saying it isn't possible. I am, however, much less in a position to fight technology as someone who doesn't have the time to be a diciple of beat rates. FWIW, I don't even have the time to tune with the computer and am using the person I mentioned. Kids do that to you

Franz Mohr,for many years before,has challenged
techs to visit him at Steinway in New York and give him a "perfect" machine tuning.

That *stays* in tune - once the concert pianist starts hammering.

Don't know if anybody ever did the trick.

But curiously - there still ain't no concert tuners on any of the world's major stages using a machine - I know!

Well, Timbucktwo....perhaps: your home or mine! :rolleyes:

...er..... YOURS!!

norbert

Temperament and inharmonicity may be separate phenomena, but they certainly come into play with each other. Your octave from 440 to 880 may actually be from 440 to, say, 881, due to inharmonicity, so every interval in between is affected. By ear, you are tuning so that the relationships between intervals are correct, that is, so that there is the proper relationship between the way that the intervals sound. If an ETD only deals with the pitch numbers, it may be correct by the numbers, but the relationships won't be right.
I don't even bother much with counting beats exactly when I tune. I'm not good at judging fractions of a second, and the beats vary from piano to piano. The temperament on a cheap spinet is not the same as that on a concert grand. But it's no big deal to get the intervals correct on both of them. (It's usually easier on the concert grand, though!)

I won't jump in the middle of the ETD vs Aural debate. I've had tuners that use an ETD (Accutuner) and tuners that do it by ear only. I will say that I have been most pleased/satisfied with tunings done by aural-only tuners.

I would also point out, fwiw, that in Germany I have not met a tech that uses an ETD. They are all of the old school over here I guess.

But I'd also note that technology/design improvements in ETD in recent years has been pretty impressive. I'm talking about the Verituner specifically. It will not, imho, replace aural checks/tweaking but it has a lot of potential to save the tuner's ears for where they are really needed. And that's why I think that this particular ETD is gaining acceptance on this side of the pond. Several highly regarded technicians in the BENELUX have started to use it and spread the word. Whether their enthusiasm will sway their conservative, hide-bound colleagues to the east remains to be seen.

It will be interesting to see if Verituner's competitors can come up a design that will take ETDs to the next level.

JP

Just what am I supposed to be saving my ears for, if not for tuning?

If you don't tune by ear, you aren't getting practice listening.

As far as I am concerned, saving my ears means leaving when the sound is too loud. If it were too loud when I'm tuning by ear, it would be too loud if I were tuning with and ETD.

(I've got two shows tomorrow before the play-offs of my bocce league, where my team went from dead last in the first half to tied for first in the second half. Maybe the play-offs are important, but not as important as the shows, and I don't need to save my ears for them. )

If you take the double negative here literally, this is true. If what Norbert means is that no concert tuners on any of the world's major stages use machines, that is absolutely false. There are many tuners regularly tuning for major pianists on the world's major stages absolutely using the Reyburn cyber tuner and the accutuner. I don't know any using the verituner, but it wouldn't surprise me. I don't know what the percentages are, but a concert tuner using a machine is not an annomaly.

Jim Coleman and Virgil Smith had a contest at a PTG convention a few years ago. Jim,tuning electronically, beat Virgil, tuning aurally. The next year Virgil won. The pianos were voted on by several tuners and pianists. Jim is also an excelent aural tuner and teacher as is Virgil.

Keith:

At least in Europe it's apparently a total no-no.

Perhaps over here it's different.

It's not a battle for me.

I don't even tune myself.

But I can tell when the tuning is good and - holds!

norbert

I dont think the relationship between machine, and aural tuning staying in tune, is relevant. Hammer technique, pin setting, and string rendering, are what keeps a piano tuned. If those cant be done solidly it wont matter what you use for temperment and octaves , unisons and so on.

16 years later do you all hold the same opinion?

I downloaded Entropy Piano Tuner on the app store. I tried tuning my piano with it and it made it sound worse.

About 7 years ago, when I played a recital at a piano festival in Feuchtwangen, Germany, the tuner was using an ETD on a cellphone. I only had a glance, but I think it was TuneLab. Tuning was fine; piano was a Hamburg B.

I know several technicians who regularly tune for concerts in the greater Washington, DC area who use ETDs. One uses a Sanderson Acu-Tuner, the other Reyburn CyberTuner.

As to HOW they use it, from what they tell me, it is a combination of aural tuning and stop the lights moving visual tuning.

I'm working on some upgrades to EPT right now. There are a few things to note when you use it, if you believe the entropy theory it's based on:

Mute all but one string when recording, unless there's a noticeable tonal difference between the strings. Otherwise, make sure each unison is perfectly in tune.

Using the "EntropyMinimizer" algorithm, make sure to set it on "Infinite" accuracy, let it run for a while, then change some notes around randomly in the UI and see if it returns them to the original tuning it found. I've gotten the total entropy down quite a bit compared to the "High accuracy" (150 steps) setting which I am pretty sure stops the optimization prematurely in most cases. Doing this on the PC is much easier. On my laptop, this took a few minutes of calculation.

I'm working on a few ideas to help it find a better tuning without intervention, and later if I have time I have some ideas to speed the computation up.