[CAUT] Sacrifice (was tuners- technology)

[Fred Sturm]

> This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

---------------------- multipart/alternative attachment
On 3/2/05 8:31 AM, "Wimblees@aol.com" <Wimblees@aol.com> wrote:

> Doing unison tuning by ear is described best by Virgil Smith. He says tha=
t the
> human computer can analyze the beats much better than the ETD. This is ev=
ident
> when tuning octaves. As he put it. "...octaves were right on when only on=
e
> string of the upper octave was sounding, but was flat when all three stri=
ngs
> were sounding".

    Yes, this is what Virgil says. That a unison goes flat overall when it
is tuned (ie, if one string is tuned =B3perfectly,=B2 when the other two are
tuned to it, the overall pitch of the unison will be flat of where the
originally tuned string was). And about nine years ago Jim Coleman confirme=
d
this claim through measurements he made with a new RCT (RCT had just come
out). I questioned Jim when he asserted he had made such measurements, as I
had been unable to measure this phenomenon with my newly acquired SAT. Jim
shared his data (I think this discussion was on Pianotech), explaining in
some detail his methodology. Bottom line: through careful measurement he ha=
d
confirmed that a tuned unison was between 0.1 and 0.2 cents flat of the
average pitches of the three individual strings. In his experiment he had
tuned each of the strings to within a measured tolerance of 0.1 cents of on=
e
another.
    I more or less accepted that at the time, thinking that RCT was perhaps
more precise than SAT, hence you could read such a small effect using SAT,
but I had to say that a difference of such a small magnitude was not going
to be significant in the actual tuning of a piano. My own take being that
Virgil was using this purported =B3fact=B2 to justify stretching octaves.
Remember he was asserting at the time (he has since recanted) that he tuned
by listening to the beat between the fundamentals of the two notes of an
octave, and made that utterly beatless. He also said (and wrote) that this
produced pure and beatless triple and quadruple octaves.
    Your mentioning this unison/octave =B3apparent phenomenon=B2 led me to be
curious. Having an RCT of my own now, I decided to try to duplicate Jim
Coleman=B9s experiment. First step is to tune each string of a unison within =
a
measured tolerance of 0.1 cents =AD not a real easy task. Among other things,
it is difficult to get repeated readings for a single string that are withi=
n
0.1 cents of one another. It requires playing at an utterly consistent leve=
l
of volume. To have any credibility at all, one needs to be able to do at
least three sample readings of each string, and have all of them be within
0.1 cents of one another.
    But, yes, I was able to do this, and proceeded to read the unison, with
the same care and the same number of samples. And then I went back and
repeated every step (re-measuring each string individually, etc). My
results: I did not confirm Jim=B9s data. I found what I consider to be
completely random results. Sometimes the three strings played together woul=
d
be flat, sometimes sharp, sometimes the same. I am by no means saying that
Jim did anything but a very careful and credible job, as I know him to be a
very careful and utterly honest person. But my results were, shall we say,
varied to such a degree as to lead me to believe that it would need a great
number of repetitions of the experiment to persuade me that there was any
measurable difference between the pitch of three strings sounding together
and the pitches of the individual strings.
    I realize that Virgil has taught this in classes, and that he has
demonstrated, and that people have been persuaded by listening to his
demonstrations. I suggest that it is quite possible that, in many instances=
,
they heard what they thought they did. First, it is next to impossible to
tune a unison within a tolerance of 0.1 cents, and I would say that it is
utterly impossible without the use of a machine. It=B9s a problem of
resolution =AD 0.1 cents is at the threshold of where a pitch produced by a
piano string can be measured. They just don=B9t produce pitch that clearly
defined. Variance in volume, and not that large a variance, will change
pitch more than that.
    So my explanation of =B3how it works=B2 in Virgil=B9s demonstrations is that,
in fact, the unison tuned is not =B3absolutely perfect.=B2 That one of the
strings is likely to be, say 0.3 to 0.5 cents flat of the originally tuned
string. And that the aural resolution of the pitch of three strings of
slightly different pitches will be affected by the factor of phasing
(phenomenon where strings will tend to phase with one another, locking thei=
r
pitches to one another just like PitchLock does), so that it is quite
possible that the perceived (and measured) pitch of the entire unison would
be lower than the original string, because of one string having a lower
pitch. And the unison might sound very clean. A unison within a tolerance o=
f
0.5 cents generally sounds =B3perfect=B2 to most everybody. But I know most if
not all of us can hear a difference of 0.5 cents in context of octaves, M3s
and many other intervals.
    At any rate, I would take Virgil=B9s assertion with a very large pinch of
salt. Maybe there=B9s some truth in there somewhere, but it isn=B9t what I woul=
d
give the status of a fact.
Regards,
Fred Sturm
University of New Mexico
PS I would be interested in hearing the results of anyone else who
replicates the described experiment.

    

---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/caut.php/attachments/a6/3b/96/2b/attachment.htm

---------------------- multipart/alternative attachment--