[CAUT] IRe: Unison drift (was Reading low humidity/seasonal SB failure)

[Fred Sturm]

On 3/6/06 4:45 PM, "Dorothy Bell" <dabell58 at earthlink.net> wrote:

> 
> Dear Fred,
> 
> I'll bite: what mechanically would make the unison drift with that
> particular triplet pattern? I tuned a home piano this morning which was in
> exactly that situation after having been tuned at much higher RH five
> months ago. Somewhere recently I read that 'if the pitch goes out, it's the
> humidity; if the unisons go out, it's the tech.' I'm certainly willing to
> take responsibility for all the griefs of man, but it doesn't make sense to
> me that unisons going out in a very specific pattern, in a very specific
> part of the piano, would indicate poor hammer technique. When I look at the
> piano, though, I don't see what would cause the consistent flat-sharp
> changes.  Can you advise? Thanks.
> 
> Dorrie Bell
> Boston, MA
Hi Dorrie,
    The short answer is "I don't know." I do know that the phenomenon
exists, repeats, and I have come to believe that it probably occurs on all
pianos. The very short term effect on the concert grand from a rain storm
shortly before I have observed multiple times. It's rather subtle: the piano
has just "gone sour.'
    But after longer and more extreme humidity changes (in either direction)
a more obvious and quite striking pattern happens. It varies in degree with
model of piano. In my own stable (or should I call that "in-stable"?),
Yamaha G2's take the cake. I'll find unisons in the mid-high treble with
left string 10 cents flat, right string 15 cents sharp, middle string close
to pitch. Lots of them. I have observed this pattern on a wide variety of
pianos, including uprights. It isn't absolute - every single unison
exhibiting it - and it varies from section to section, but it seems to be
consistently there. In all cases, I believe, the treble-most string moves
the farthest in the direction of humidity change (ie, up with rise in RH,
down with lower RH).
    As I have puzzled over this for 15 years or so, I have come to a few
tentative conclusions. First, I have decided that probably the tuning pin
pattern doesn't play a major role. I'll list three pretty strong reasons:
1) The phenomenon happens in both grands and uprights, with the same left to
right relationship. Pins on grands and uprights are mirror image in terms of
distance from speaking length. If that variance in distance were a major
factor, the pitch difference should be opposite as well.
2) Steinway B's (and a few others) have a particular tuning pin pattern
we're all familiar with. If distance from speaking length to pin were a
factor, we should see a significant variance from unison to adjacent unison,
and particularly from the closest of a pattern of six pins to the farthest.
I have looked for such a pattern and have failed to find one.
3) I have observed this phenomenon on a Wurlitzer grand with a Wegman plate
(no pinblock). Hence, no swelling or shrinking of pinblock could play a
role, at least on that piano.
    What does that leave? Well, there is no typical pattern of back lengths.
That leaves the bridge. Particularly the notching pattern, with the treble
most pin closer to the center of the bridge, the bass most pin closer to the
edge. This pattern holds for both uprights and grands. What could the
mechanism be? Well, it's a bit counterintuitive. The way I figure it, the
pin closer to the edge of the bridge migrates (with lateral expansion of the
bridge) in the direction of the speaking length more than the pin that is
closer to the center. Hence, it makes the speaking length (or tension) that
little amount less, counteracting the overall rise in pitch from soundboard
rise, etc. 
    I won't say I believe this is the answer, but it's the best model I have
been able to come up with. Jim Ellis believes it to be more to do with
tension variance due to greater or lesser side bearing of the string
crossing the bridge between the two pins. Either way, that doesn't seem to
be enough to account for it. OTOH, I think most of us have experienced that
strange pitch change that happens when a string breaks. The immediate
neighbor strings go sharp by a significant amount. What can possibly account
for that? Very minute changes in down pressure on the bridge, and side
pressure through those bridge pins. If such seemingly minute changes can
cause a measurable pitch change, then maybe the bridge model I described is
the key to unison change caused by RH change.
    But this is all very hypothetical. I would welcome other models, and
other people's observations. Maybe I am wrong, and this isn't as universal
as I think. It is harder to observe in private homes because I'm not there
enough to know for sure what happened since the last tuning. So my main
observation is limited to the models I have at the university.
    One thing I am pretty certain of: Unisons go out of tune in response to
RH change. It's not always lack of skill on the part of the tuner <g>.
Regards,
Fred Sturm
University of New Mexico