Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum HISTORICAL ET AND MODERN ETs

The Pianoteq developper, Philippe Guillaumme (whe is/was a piano technician) made a CHas tuning file that takes in account the piano model ih .

Here is a real application of the tuning formula in a virtual piano :

http://www.forum-pianoteq.com/uploads.php?file=CP33%2005.03.2010%202%20CHASIH%20demo.mp3

I need no more to be happy (and the Pianoteq software is doing VERY good on that !)- you have a trial version that is complete if you wish ... go to the Pianoteq forum or ask me if you want the CHas tuning file for Pianoteq.


Same music but with standard tuning , for comparaison purpose

http://www.forum-pianoteq.com/uploads.php?file=CP33%2005.03.2010%201.mp3

Best regards.

Thanks for these mp3s Isaac. The CHAS tuning really seems to make the chords 'breathe' so much better. I'm sold on this technique. I think I will attempt to tune it when I next have some time. See what it can do for the old Erard!

That sound is very, very good. The treble 12ths are wider, however, than what Alfredo specifies in his tuning sequence, using a 3:1 check.(Sorry, Philippe. Just being specific about the details of the tuning.)

(By the way, as Olek says: PianoTeq has a demo available for the standard edition with some notes silenced. But there is not a demo for the pro version, which lets you tune each note individually and adjust the amplitude of each partial in each note, etc.

I recently found, in any case, that the ability to adjust each partial's amplitude makes the program very good for listening to beats while pitching each note or the unisons. You can mute all of the partials you don't want to hear, so that you can listen to just the 3rd and 1st partial while tuning 12ths, say.)

Euh Jake, I apologize, because as adepts of the Chas ET concept, we should not have contradiction publicly (all those may stay confined to the ashram !)

But to me the 12ths in the high treble are tempered , I can hear it , but you gave me doubt so I measured with Tunelab.

Put the display at zero move for instance at G7 (Tunelab is tuned to the G7 frequency)
Play C 6 - the display goes right , showing a 3d partial of C6 being higher than the actual pitch of G7

measure the offset by tapping on the display with the mouse till it is stopped.

I measured 2.71 cts offset form a pure triple of the fundamental frequency . (it is well tempered, in that case)
I wonder how do you test the 12th, if you hear the M6 faster than the 17 th that mean that the interval is shorter than pure.
If you find the 3d partial of C6 higher than the fundamental of G7 that is, a tempered 12th.

Which notes did you find over "pure "twelve ?

Anyway, iH plays also a role, and modify the way the notes are pitched. G7 with a high iH, may tone higher than its fundamental frequency. fundamental of C6 may be lower than 1/3 the frequency of G7 , that is normal, to me, the 3d partial is what provide the beat with the 12th so its frequency is what counts)

In the recording I dont feel the high treble as asking for more stretch, to me that is one of the main advantage of the Chas ratio, as it bring a high coherence. When we stretch "artificially" just to please the ear, we cut in the resonance, as adding stretch octave after octave finally provide a high treble that have lesser sympathetic resonances.

Whenever it is possible to have a natural stretch all along (and I wonder if it is even possible out of the Chas ratio) the instrument get sonorous but stay smooth.

I also have find that in the basses the 5ths are yet tempered (the same than in the medium, while in a more classical tuning, the fifths get very large in the bass and even invert sometime.

I am unsure that the formula applied get exactly to the result from aurally tuning Chas, but the main things are really there, I verified the beat rates, the intervals, on a midi keyboard, and they have, to me the same relations than with an aural tuning.

May be you expect the fundamental frequencies to be less than 1/3 ratio, it may be the case where the ih is lower.

Despite the limitations, (a few notes missing, no direct tuning) I enjoy much the demo of Pianoteq -i am not equiped enough with a decent midi keyboard, but I have read the users comment, saying the software acts more as a musical instrument because of real time computation of tone (compared with sample based setups) I suppose that in the future modeling will be the tendency, for digital instruments. The possibilities are amazing (as on Pianoteq the strike point changing, the hammer voicing changing the length of strings , and now , allowing us to test different tunings... etc..).

I also appreciate the interest of Philippe Guillaume for the Chas concept, and his work to put in reality something that we could not do only with pianos...

Here is a real application of the CHAS tuning formula :

http://www.forum-pianoteq.com/uploads.php?file=CP33%2005.03.2010%202%20CHASIH%20demo.mp3

Here is the same piece with a standard tuning (ET both)

http://www.forum-pianoteq.com/uploads.php?file=CP33%2005.03.2010%201.mp3

Best regards.

Thanks Isaac, what has been recorded makes an undeniable difference.

Philippe and Jake, Pianoteq is highly performing indeed and I'm very, very happy. Thank you.

Best regards, a.c.

I, too, like the recording, but: I was just doing the 3:1 check that Alfredo says to use in his tuning sequence. The 12ths are wide when I examine them in Pianoteq. Are we perhaps discovering a conflict: Pianoteq shows 12ths as wide on 3:1, but TuneLab shows them as narrow? We need to know. (We aren't contradicting each other. We're trying to get all of these things right.)

Maybe the Chas tuning was autosettling into a Stopper tuning in pianoteq, who knows? (for the dumb people out there, this is a joke, haha!)

Where´s the problem Jake? If the piano synth sounds good and the duodecimes (twelfths) are on the wide side of a third-partial-of-the-lower-note vs fundamental-of-the-upper-note (or simply said a 3/1) partial match, it is apparently more a Stopper tuning equivalent than a Chas tuning. The only persons who may be concerned about this are Capurso and Kamin (and eventually me of course if my tuning is going to be marketed under Chas name).

Stopper, I politely ask you to keep your market anxiety out of this Topic.

You could contribute, for instance, by describing your only-pure model more in depth (?), by explaining how a beatless 3:1 ratio can be congruent (how can it make sense) with beats simmetries (?), and by sharing your aural tuning sequence for a pure 3:1 ET (?).

a.c.

The problem was...I wasn't just trying to get a sound that I liked--I was trying to be very precise in following CHas.

(Remember that I said that the 12ths were wide--so it's actually neither a Stopper, which specifies just 5thd, nor a CHas, assuming the wide reading is correct. The intervals in the midrange are CHas (I think?), however, so it's more of a stretched CHas.)

But I must say I need more information. It may already be somewhere in this nine page thread. Sorry if I missed it, but two questions that reveal the sorry state of my understanding:

1. To have slightly wide octaves and slightly narrow 12ths, on what partials does one listen for the beats? 4:2 for the octave and 3:1 for the 12ths?

2. When speaking of wide octaves, does this mean just using a 4:2 check on the octaves, which will produce a wider octave than a 2:1 check, or slightly widening the octave past that? In other words, is just getting the upper partials nonbeating the goal, which will widen the octave, or is one making those upper partials beat slightly, so that the octave is widened still more?

Jake:

If you are trying to get wide double octaves and narrow 12ths, just tune 4:2 octaves. Because of iH, tuning stacked 4:2 octaves will always produce a double octave that is wide of 4:1. Whether the single octaves, double octaves or 12ths will sound like they beat depends on the listener.

If you are trying to understand Alfredo's tuning instructions, I cannot help you. I have not been able to reconcile the different things he has said.

Jake,

about the 12th, I recorded today a twelve with a pure 3:1 ratio.
I mean : 440 hz (A-49) and 1320 Hz : E6

That twelve had 5 beats at the 3:1 level , coming form the 3d partial of A440 that was at 1325 Hz (measured with Tunelab)

It was on that piano, I admit it is a highly inharmonic one

To tune that 12th so it have no beta at all, the E6 may be tuned at 1325 Hz.
That is what tuners call a "pure" interval (which by evidence is all but "pure" but that is another story).

Frankly I did not get yet all the theoretical explanations, from Alfredo paper, (but I work on that) I was lucky enough to see him doing, to listen and to analyze that with what I know about tuning - as usual those kind of things take time aint as discovering something and understanding immediately what happens.

Jake you misunderstood about the octave (while the octaves are always wide at some point in a piano, if minimally the 2d partial of the bottom note IS the pitch of the fundamental of the upper , this is what Philippe call "iH correction", if not the octave sound false, and too small.

The equivalence of beats in Chas is between the 12th and the DOUBLE octave (2 octaves stacked)

This lend to octaves that are wide but it is not as say Jeff 4:2 + 4:2, the bottom octave can be short of 4:2 , it depends of the level of iH at that part of the piano (and the level of iH of that particular piano). There is always a tendency to oversimplify those concepts, practically many things are intricate, many technical gesture have to be learned, and appropriated, a good job is the result of many details added.


ALfredo tunes by listening directly at the slow and lower level of beat rate : octaves at 2:1 , doubles at 4:1 12ths at 3:1 ; 5ths at 3:2 not using the checks that compare 2 fast beating intervals, as the M6 M17th to check the 12th size.

I tuned the last piano using those checks (after having followed the same method Alfredo uses) , and seem to me they are suitable for that and that the result is similar.

Simply it may be easier to tune directly and compare the beat progression of intervals that are next each other, than to use totally different intervals, that add complexity for the listening (but in the end the evenness of tone is helping, all intervals may be progressive so one can use whatever interval he prefers.

Coming back to the theory, I simply noticed that the method works perfectly, and that a "hot spot" is there, efficient and audible.
In the end the harmony is raised, the piano sound a tad larger, that have always be my intention to tune in the harmonic spectra of any piano, taking care that the notes spectra add each other nicely en raise the resonance of the instrument.

One of the main problem for tuners is to tune the high treble so it does not sound too low, nor too much stretched. Most of the time the top end of the piano follow a kind of added stretch that is not really lining with the notes under. The same occur in the basses. Enlarging the mediums helps to get there more in line, but only the Chas method allow to get some crispness in the treble (due to the good coupling of many intervals with the top note)without accentuating the stretch curve (it accentuate , but naturally because of the ih raising much in the treble).

Voila, thats all I can say at the moment. There is even more in that, I feel so, but I cant say much more.

I had yet today a pianist that call me saying he like the way the piano plays now , I appreciate it; not easy to do new things...

On the records made with Pianoteq, what I hear is may be not noticed by anyone, but I know what I am looking for.


The Chas tuning is highly balanced in a stable equilibrated way. Other ways to balance a tuning are available, providing harmony, but very often the zone where the notes are all perfectly in tune in the main harmonic modes is not so large, 4- 5 octaves span may be, then the piano may have registers, and the tuner have to even/regulate those registers so they are appropriate to the acoustic of the place.
I am yet to listen Chas when used in concert, and with orchestra, in a room , I go for it. To me if the justness seem to be raised in the ear of musicians most probably it will be good to play with orchestra also.

The effect is so particular that I cant predict what it will give in that situation.

Yes, I see now. But where I still get lost: given that the double octaves and 12ths should beat equally, how much should they beat?

Am I seeking a compromise to adjust for iH, or is there instead an intentional slight roll that I want to install? What should I do, in other words, if the piano has very low iH, so that the double octave beat and the 12ths beat are almost indetectable? Should I slightly repitch both to create a more perceptible, slow rolling that is identical on both the double octave and the 12ths?

Hi Jake,
It might depend on how the other notes relate to each other down the register so as to stay consistent with the amount of stretch. You have also to consider how the "whole piano" sounds. That is one of the advantages of checking double octaves, etc.

I tuned a few pianos on Saturday and experimented with a compromise between a double octave and an octave and a 5th. The wider the octaves become up the register relative to the tempered section, the less narrow the 5ths become, working their way to near pure.

Frankly, when you get either correct, the instrument will sound good. It then becomes more qualitative and a matter of taste as to how much to stretch.

Without really good unions anyway, all the octave stretching and temperament choices begin to lose their value.

If you have time, post a few samples of your tunings.

Glen

Jake this is a very good question ! on pianos with high iH all mean to absorb that iH and find something harmonious are good. But if the piano have low iH , to have the same beat value in an octave it may be less stretched than usually.

I did not have that situation yet, so I cant say. As the idea is to have an equilibrium between the double octave, the 12
th, that relation induce also a symmetric relation
2 octaves lower.

I tend to thing that the global stretch may be lowered then.

There is a roll, but it is so slow it couple before finishing its cycle. anyway it is not noticed as a roll, just as the beginning of one. It may be ascertain by checks using other intervals, as the M3 and the double octave, or the M3 and the octave, and comparing the beat speed. But with differences as small as 0.3 bps it is very easy to miss, and easier to go for a kind of tone , or a behavior of the interval, and stick to it.

Anyway, the difficult part is to begin, after that the good spot is in the ear and in the piano resonance. (that part is less reproduced with Pianoteq , what I hear is some resonance, but not as much reinforcement as at the piano- but it is still partly there so it may be possible to tune it.

You may try to learn to recognize a "pure 5th" or a "pure 12t" from a larger one and a tempered one. Musically, the output is different, even if the beats may be difficult to perceive at first, the color of the interval change.

Jake:

With a 12th and 15ths having a common note on top, the algebraic difference of their beatrates will always be equal to the beatrate of the implied fourth on the bottom regardless of how the top note is tuned. So if you want them to be equal beating, and the fourth beats 1 bps wide of just, the 12th must beat 1/2 bps narrow and the 15th must beat 1/2 bps wide. 1/2 - (-1/2) = 1.

When tuning equal beating 12ths and 15ths, fourths beat about 1 bps in the middle of the piano, a little faster in the treble and a little slower in the bass. So the 12ths and 15ths may beat up to 1 bps in the treble, and probably slower than can be noticed in the bass.

(Glancing at the thread title and seeing "Historical ET and Modern ETs," I worry that I'm narrowing the focus too much, here. I'll continue over on the CHas pretuning thread.)

Hello, just a personal comment about a fresh acknowledgement.

A young piano tuner and PW poster from Finland, Patrick (pppat), writes:

..."there is something esthetical about the mathematically symmetrical tuning (ET), but there is no question that the color palette of EBVT III makes the (sensible) pianist play the instrument in a different way."...

To me, this is quite something: for the first time a positive comment is spent about the mathematical symmetries of Modern ETs.

What may any young tuner share next?

1 – The first “conventional” ET, 12th root of two, has evolved into “variants”

2 – These ET “variants” have now been described both mathematically and geometrically, they represent Modern ETs

3 – Modern ET theory is finally practicable

4 - Modern ETs beats-symmetries can increase the performance of any piano

About any tuner, I hope his/her understanding about Modern ETs beat-symmetries go from their aesthetic beauty deep down to their astonishing practical effects. Then, he/she will be able to recognize past and present hybrid tunings with a high degree of specific knowledge and professional consciousness.

Regards, a.c.

.

You know, When I have read those kind of comments for the first time, I have find them somehow presumptuous,and I recall I have stated that.

Now that I have tried n and hopefully understand (at last partly) the composition of that tuning, I may say that I understand what Alfredo is trying to state ).

The result of that evening of beats, symmetry and higher resonance, gives the tuner a different ear when he listen to a more conventional ET, and it provide him a reference model to compare with. That is very strange as a sensation, but the level of harmony is then apprehended when listening to more classical tunings.

THen , in some cases the same kind of effect can be perceived sometime only in a portion of the scale. SOmetime it is even larger than the ratio, (un focusing) and in other situations the compactness of the tuning is what jump to the ear.

I believe that it may change the way we listen, eventually the way we tune, indeed and certainly the way we "stretch" a tuning if done in our usual way.

To me, at the same time the Chas is at the best place for resonance and harmony, while being "at the edge", meaning that a step above and the piano will not tone well.

So knowing that kind of limit in spread is a very useful thing, to me , as the tuning itself is simply "normal" the ability of the tuner to have a good tone and stability, and to understand how the piano settle, is only what will make the difference, part experience, part good sensations, part listening, but justness is well taken in account by the ratio itself, that is a quiet comfort sensation, which is appreciable.


I have seen private witnesses and acknowledgments coming from other tuners and they seem to say all the same thing.

I know some good tuners, and it is really strange, but all of them have doubts : do I have well stretched that tuning ? Is the justness well reconciled ? Many of them are unsure of the real justness of the instruments, and that seem to be confirmed by the orchestra instrumentalists that say that the piano is "in tune enough" for them.

We dont imagine how much some of us can be wanderers even after 30 years of career and more ! I'd appreciate comments from other tuners on that aspect (but seem to me that having strong affirmations is also a natural tendency in that trade, and that questioning itself is not much in the trend, nor agreable)


The theory is that as long as the tuning come by the Chas ratio, it may settle in there, for some reason

This may be somewhat difficult to test and prove, but possible, if someone have an idea of a battery of to test it would be welcome

One thing that I find striking is that EBVT and CHas, although obviously very, very different in intention, both seek equal beating on the narrowed 12ths and widened double octaves. They are seeking similar resonances. (But of course EBVT brings out the key color changes, while CHas instead seeks the ET goal while trying to keep the resonances.)