Hello Slestrange,
I am a professional piano tuner who has over forty years' exposure to tuning various makes and types of pianos, whose conditions range from very good to downright horrible. Please note that I do the following in Pianoteq PRO, but it is worth sharing with you the amount of de-tuning I perform on every Pianoteq model: D4, C4, K1, Erard, etc.:
1) Retain the stock settings of diapason (set to A=440Hz), and string length at 2.7m as a general starting point.
2) I reset the tuning in the top 2 to 2-1/2 octaves to grow exponentially (curving upward per note) from +0 cents up to +25 cents.
3) I reset the tuning from the bottom 1-1/2 octaves to flatten exponentially (curving downward per note) by about 10 cents.
Note: The above 2) and 3) corresponds to the tuning of 9' grands to compensate for the natural inharmonicity in the strings.
4) I slightly randomize the tuning by giving one or two clicks on the randomize button in Pianteq PRO to the above tunings, such that one does not necessarily hear any specific notes as being out of tune, but it contributes to the way a piano becomes within a few hours' playing of a concert tuned instrument.
Next, I go in and adjust the unison tuning of individual notes within Pianoteq PRO:
5) I minimize the unison width for the single strings in the bottom octave or so, because single strings should not be out of tune with themselves -- unless they are defective and contain a twist in them. The result of minimizing the unison detuning in the lowest octave is that no single string is able to "self-warble" -- an otherwise undesirable effect in any grand piano. Restated, when one hears the lowest octave single strings warbling by themselves, it is an immediate clue that something is dreadfully wrong in any piano library.
6) Up about an octave from the bottom, where two strings first occur, I nudge a small "bump" in the unison tuning. This is to simulate the difficulty of maintaining absolute tuning of two unison strings over time. I generally taper off the degree of unison detuning, until I hit the area where the three strings come in -- and nudge an additional small amount of unison detuning, and then taper the effect off as one goes up the keyboard.
7) Next I wish to account for an amount of unison detuning that occurs in the uppermost two octaves, simply because these are the most difficult strings to tune correctly in unison.
After having performed 5), 6) and 7), I tend to hit the randomize button again.
Lastly, I wish to slightly unison detune the middle three octaves, simply because these are the notes that are played most often, and tend to go out of tune from simply playing the piano.
All of this sounds like a lot of work, but I can do all of the above steps within a few minutes -- and then freeze these settings for introducing into other presets.
The "trick" to all of the above is to perform the steps sparingly. None of this should be done to the extent that the listener is consciously aware of what is happening. On the other hand, if one does wish to make the piano sound deliberately detuned, then simply apply the above steps with much greater effect per step. In fact, as the creator of the "SuperTonk" preset in the new D4 model, this is what I did to the various aspects of tuning.
Hopefully this might shed some light on how real pianos do not sound as antiseptic of many sampled pianos (whose notes' detuning I am unable to modify in this manner).
Cheers,
Joe
EDIT: Upon reading your original posting, I believe your concern (about the lowest notes not being in tune with the highest notes) has to do with something called Inharmonicity. You may wish to move the stretch tuning slider, to compensate for the overtones of lower strings not corresponding the the fundamental pitches of so-called correctly tuned notes. This is essentially what I mention in Step 2) above; that is to say, to increase the stretch tuning. I manually stretch the tuning in the uppermost octaves by up to 25 cents (half a quarter tone) sharp -- to compensate for what you may in fact be hearing.
Last edited by jcfelice88keys (24-05-2012 06:08)
