I just came on to this topic in today's posting.  The topic of pitch
raising and overpull has been discussed extensively on the Pianotech
discussion list.  Search their archives:

  http://www.ptg.org/
  http://www.ptg.org/tech.htm
  http://www.ptg.org/archive/pianotech/

New wire will stretch for a period until it is stable.  Any wire
pulling around a bearing point will not, in and of itself, cause the
string to go flat by the wire bending at a new point.  The natural
curve of the wire remains until a force is applied to sharpen this
termination, then it will go flat.

You've heard of tapping the strings to the bridge, this was done for
two purposes: (1) to remove the natural curve, and (2) to seat the
string on the bridge.

The 'tapping the string to seat' procedure has fallen into disfavor
because it forces the wire into the bridge surface, denting the face.
So only a gentle tap at an angle to remove the natural curve and define
the termination.  _But,_ this procedure could exacerbate an existing
problem: false beats.  False beats are cause by loose bridge pins,
not the string lifted off the surface.  An added sideways force
could damage the pin's footing as well.

I fashioned a tool from a screw driver to brace against the pin as it
applies a force on the wire to sharp the termination (a JPG photo
image is available).  Also, lifting the wire at the agraffe or capo will
better define the termination and improve the sound.  The next time you
hear false beats, apply pressure on the bridge pin; it will disappear.
Fitting the strings to the hammers (mating) will also clear up the
tone.  Removing the natural curve, leveling the strings and mating
them to the hammers are prelude to voicing.

So what causes the string to go flat during a pitch raise?  Plate
compression and case distortion, not so much board deflection as
thought previously.  Then there is the friction factor of the strings
pulling through the bearing points (over time) if they don't pull the
string segments to equal tension.

So during a pitch raise on an older piano the overpull is applied as
10% bass, 30% middle, 36% treble, meaning that if a note is 30 cents
flat, the pitch is raised to +3c in the bass, +9c middle and +10c
treble.  This will result in all the notes being within a few cents
for fine tuning.

On an older piano the strings are pretty much stretched so the only
reason for the piano going flat is plate compression and case
deformation.  Then as the piano is played, any string segment which did
not pull up due to friction will do so as the board is excited during
playing.  The pitch of a new piano could be raised slightly sharp for
the same reasons but loading the board on a new stringing job is
another topic.

On a newly strung piano, once it is at pitch -- by removing the natural
curve at the hitch pin, rear duplex, bridge pins, front termination and
counter bearing bars -- the piano will be more stable and fewer tunings
will be required to maintain pitch.

As for inharmonicity and % of breaking strain, these are considerations
for the scale designer.  The tuner will take inharmonicity into account
for how much of a tuning stretch will be applied throughout the piano.

I was unaware of the effect EMF's had on piano tone during manufacture.
Will they have a similar affect during restoration?

Regards,

Jon Page <jonpage@attbi.com>, piano technician
Harwich Port, Cape Cod, Massachusetts