<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; =
charset=iso-8859-1">
<META content="MSHTML 6.00.2719.2200" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Add the B( and you have a C7 =
chord</FONT></DIV>
<DIV>Joe Goss<BR><A
href="mailto:imatunr@srvinet.com">imatunr@srvinet.com</A><BR><A
href="http://www.mothergoosetools.com">www.mothergoosetools.com</A></DI=
V>
<BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; =
BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
<DIV style="FONT: 10pt arial">----- Original Message ----- </DIV>
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: =
black"><B>From:</B>
<A title=mathstar@salemnet.com =
href="mailto:mathstar@salemnet.com">Alan R.
Barnard</A> </DIV>
<DIV style="FONT: 10pt arial"><B>To:</B> <A =
title=pianotech@ptg.org
href="mailto:pianotech@ptg.org">pianotech@ptg.org</A> </DIV>
<DIV style="FONT: 10pt arial"><B>Sent:</B> Saturday, September 21, =
2002 7:53
PM</DIV>
<DIV style="FONT: 10pt arial"><B>Subject:</B> Ghost Tones by =
Request</DIV>
<DIV><BR></DIV>
<DIV><FONT face=Arial size=2>I mentioned ghost tone testing in an =
earlier post
and a few people emailed me to ask huh? What? How? So, post the
following. I am also attaching this as a Word document, if anyone =
cares to
print out a cleaner copy with better layout.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>
<P align=center>GHOST TONES IN PIANO TUNING</P>
<P align=center>Based on Information and Ideas Stolen from Various =
Reliable
Sources</P>
<P align=center>By Alan R. Barnard</P>
<P align=center></P>
<P align=center> </P>
<P>Introduction</P>
<P>I'll try to outline the science of strings, briefly, and their =
ability to
generate "ghost tones" at specific partials for accurate tuning =
checks. Most
readers will already know some or all of this information, but it =
really helps
me to understand the theory behind our practices so I assume others =
will
appreciate having it presented. If you are pretty new to all of this =
or
limited in knowing the physics of strings and/or music theory, =
<I>please sit
at a piano as you read this.</I> Experiment with the relationships I =
describe;
I think you’ll "get it."</P>
<P>I’m writing this all out of my head. If I get anything wrong, =
email me
pronto, please, and I will post a corrected version of the whole =
thing.</P>
<P>Intervals for "ghost tuning" are ones that occur naturally in one =
string so
this is not a discussion on temperament or evenness-of-scale testing =
(running
3<SUP>rd</SUP>s or 10<SUP>th</SUP>s, for example) for which ghosts are =
not
appropriate. The reason for this is an "energy" problem (more on this
below).</P>
<P>Also, my presentation of theory is based on "perfect string" =
science. No
string, and especially not a heavy steel string, can begin vibrating =
precisely
at its termination points due to the stiffness of the material. This
generates, of course, the "inharmonicity" that provides job security =
for
professional tuners and is never quite grasped by amateur "tooners."
Inharmonicity, its relation to tuning, string length, size of piano, =
unwanted
noise in single bass strings (longitudinal inharmonicity), and other =
topics,
are all very interesting but beyond the scope of the present =
discussion. </P>
<P>So we will assume perfect string harmonicity in this discussion =
although,
ironically, it is inharmonicity that makes "ghost tuning" useful and =
interval
testing necessary in the first place!</P>
<P></P>
<P>The Science Behind It:</FONT><FONT size=2> </P></FONT><FONT =
face=Arial
size=2>
<P>Every string vibrates in a complex way including ...</FONT><FONT =
size=2>
</P>
<UL></FONT><FONT face=Arial size=2>
<LI>End to end, at its fundamental pitch (1<SUP>st</SUP>
partial).</FONT><FONT size=2> </FONT><FONT face=Arial size=2>
<LI>In halves, with a node (dead spot) in the center of the string, =
one
octave above the fundamental. This is the 2<SUP>nd</SUP> partial =
(or, to
make it confusing, some call it the 1st overtone)</FONT><FONT =
size=2>
</FONT><FONT face=Arial size=2>
<LI>In thirds, with two nodes, for the 3<SUP>rd</SUP> partial
(2<SUP>nd</SUP> overtone). This interval is a 12<SUP>th</SUP> and is =
an
octave + fifth above the fundamental.</FONT><FONT size=2> =
</FONT><FONT
face=Arial size=2>
<LI>In fourths, with three nodes, for the 4<SUP>th</SUP> partial =
(3rd
overtone). This is the double octave.</FONT><FONT size=2> =
</FONT><FONT
face=Arial size=2>
<LI>In fifths to produce the 5th partial. This interval is the
17<SUP>th</SUP>, or two octaves + fifth.</LI></UL>
<P>For a brilliant and impressive demonstration of this, ask a =
brilliant and
impressive guitarist to play "harmonics" for you. Especially at the
2<SUP>nd</SUP> partial, you can actually <I>see</I> the string vibrate =
in two
halves with a node at the twelfth fret (1/2 the string length). While =
it is
thus vibrating, you can lightly rest your finger on the string at this =
node
without dampening the sound in the least. The 7<SUP>th</SUP> fret =
hamonic is
the 3<SUP>rd</SUP> partial, the 5<SUP>th</SUP> fret is the =
4<SUP>th</SUP>
partial, etc. Do you see the implications of this for hammer shaping? =
Damper
placement and length? Hammer strike line? Scaling, in general?</P>
<P>A cellist or bassist could also demonstrate this very well. Violins =
you
could hear, but not see, what is happening. Did I mention that I am a
guitarist (classical)? The jury is still out on "impressive," and =
definitely
found me not guilty of that "brilliant" thing.</P></FONT><FONT
size=2></FONT><FONT face=Arial size=2>
<P>Here is a chart of these intervals and their musical tone =
equivalents for
the note C28. The key numbers listed under "Note" refer to the key =
that plays,
at its fundamental, the same pitch as the partial of C28. </P></FONT>
<TABLE cellSpacing=1 cellPadding=7 width=638 border=1>
<TBODY>
<TR>
<TD vAlign=top width="33%"><B><FONT face=Arial size=2>
<P>Note (Key)</B></FONT></P></TD>
<TD vAlign=top width="33%"><B><FONT face=Arial size=2>
<P>Partial</B></FONT></P></TD>
<TD vAlign=top width="33%"><B><FONT face=Arial size=2>
<P>Interval</B></FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>C28</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>1<SUP>st</SUP> </FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Fundamental</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>C40</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>2<SUP>nd</SUP> </FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Octave</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>G47</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>3<SUP>rd</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Octave + Fifth (12<SUP>th</SUP>)</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>C52</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>4<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Double Octave (15<SUP>th</SUP>)</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>E56</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>5<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Double Octave + Third (17<SUP>th</SUP>)</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>G59</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>6<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Double Octave + Fifth (19<SUP>th</SUP>)</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>A#62 (B</FONT><FONT face=Harmony size=2>(</FONT><FONT =
face=Arial
size=2>)</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>7<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Double Octave + Seventh</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>C64</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>8<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Triple Octave</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>D66</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>9<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Triple Octave + Second</FONT></P></TD></TR>
<TR>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>E68</FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>10<SUP>th</SUP></FONT></P></TD>
<TD vAlign=top width="33%"><FONT face=Arial size=2>
<P>Triple Octave + =
Third</FONT></P></TD></TR></TBODY></TABLE><FONT face=Arial
size=2>
<P>Important or Interesting Comments on the Chart Above</P>
<P>Please note distinctions between the terms partial, interval, and =
chord
member. The note G37 in relation to C28, for instance, is the
<I>3<SUP>rd</SUP> partial</I> of C28, a <I>chord fifth</I> in a major =
or minor
C chord, and an <I>interval of a 12<SUP>th</I></SUP> from C28.</P>
<P>It is also interesting to note the natural science of music theory =
that a
single string demonstrates. In our example, the first 10 partials =
include 4
C’s, 2 E’s, 2 G’s, a B</FONT><FONT face=Harmony =
size=2>(</FONT><FONT
face=Arial size=2>, and a D. So we have a C major chord in the =
natural
harmonics of the string, i.e., C-E-G. Add the B</FONT><FONT =
face=Harmony
size=2>(</FONT><FONT face=Arial size=2> and you have a C7 chord; =
add the D and
you have a C9 chord. The same relationships hold for any string =
played.</P>
<P>Note: It is only a coincidence that the chord seventh, here, is the =
7<SUP>th</SUP> partial and the 9<SUP>th</SUP> partial is a chord =
ninth. In
fact, the 3<SUP>rd</SUP> partial is a chord fifth, the 5<SUP>th</SUP> =
partial
is a chord third. Don’t be confused.</P>
<P>It is the relative strength or weakness of each sounding partial =
(and the
higher ones not here named) that, mostly, account for the string’s =
timbre
(pronounced "tam·br"), that is to say, it’s particular character =
of sound.
Middle C has different characteristic sounds in pianos, harps, =
guitars, and
cellos because of the way string materials, lengths, masses (weights), =
&
tensions—as well as differing soundboard and resonating chamber
constructions—impede or encourage the strength of each partial. =
Similarly, it
is the reason an oboe, a clarinet, a flute, and a saxophone can play =
the same
note (fundamental) yet sound distinctly different. This likewise =
explains why
two human voices singing the same note sound different, especially if =
there
are significant physical differences between the two persons, e.g., a =
tenor
and a soprano each singing middle C. Further examples include organ =
"stops,"
reproduction of sound through two different hi-fi speakers, the same
instrument played in two different concert halls, etc.</P>
<P>Demonstration</P>
<P>At a tuned piano, hold down the C28 key (without sounding it). Use =
no
pedals. Strike and instantly release C40 so the damper immediately =
stops this
note. You will continue to hear C40 because, through what is called
"sympathetic resonance," the energy you put into the piano has caused =
C28 to
vibrate at it’s 2<SUP>nd</SUP> partial (1<SUP>st</SUP> =
overtone)--the C40
pitch! To prove that this is so, strike and release the C40, as above, =
then
release C28 and note that the middle C tone of C40 stops instantly. =
Instantly,
that is, if your C28 damper is working properly.</P>
<P>Just for fun, hold down C28 as above and quickly strike and release =
any or
all of the notes above C28 listed in the chart above. They will all =
sound
until they decay or until you release C28.</P>
<P>Now, holding down C28, try striking and releasing a key <I>not</I> =
in the
above chart. Any tones that are picked up by the C28 strings are from
<I>overtones</I> (partials above the 1<SUP>st</SUP>) of that struck =
key, not
from its fundamental. These overtones, have less energy than the =
fundamental
of that key. This explains, in part, why sympathetic vibrations picked =
up by
the C28 strings are relatively weak and somewhat out of tune, in this =
case.
The other reason is that the C28 strings are naturally sympathetic =
only to
their own partial series, as per the chart above.</P>
<P>Try the experiment with G35. This is the first natural fifth of the =
C major
chord built on C28. So one might expect it to produce a loud =
sympathetic ring
in C28. But it doesn’t. In fact compare it to what is generated by =
G47 at the
12<SUP>th</SUP>. G47 is in the partial series (chart) and G35 is not. =
The
energy put into C28 by G35 is at G35's second partial, not at its =
strong
fundamental, and therefore has the pitch of G47—but not with the =
amount of
energy the G47 key will impart—at its fundamental pitch—if =
<I>it</I> is
struck.</P>
<P>The beats between strings of different intervals (or out-of-tune =
unisons)
are caused by the interactions of closely--but not =
identically—aligned
partials of the respective strings. Two strings in a unison provide =
the
easiest example: When they are in tune (and presuming they have =
identical
mass, tension, and length—with no nicks, kinks, rust, or other =
flaws), every
partial will, in theory, line up exactly. Each partial will be exactly =
in
phase, meaning the highs and lows of their generated sound waves will =
occur at
precisely the same instant. If the strings differ in any way, =
something will
change, some of the partials will be out of phase.</P>
<P>When two tones are near each other but not in phase, their =
respective sound
waves will sometimes reinforce each other (we hear louder) and at =
other times
interfere with each other (we hear softer). These are the proverbial =
"beats"
we tune with or try to tune out. Beats are a real, physical =
phenomenon—a
change in the way sound reaches our eardrums—not a psychological =
trick of the
brain, as some believe.</P>
<P>As an interesting aside, multi-engine aircraft can also be tuned =
with
beats. When the beats get slower, the prop speeds are closer; when the =
beats
stop, the two propellers are spinning at exactly the same speed =
("rotational
velocity," to be precise). If the plane has more than two engines, you =
have to
"tune" two of them, then start and "tune" the third, etc. The same =
principle
applies to tuning strings: you can really only tune one at a time. =
Even if you
aren’t muting out other strings, you are only listening to and =
tuning one
string at a time. Also, if the propellers differ in some way (not a =
good
idea), would you have trouble tuning them to beats? Yes. It would be =
like
tuning two badly mismatched strings in a bass unison (Oh, Mercy, =
don’t you
know: Been there, done that, got the souvenir T-shirt.).</P>
<P>If you want to see this effect, take some stones of differing mass =
or size
and find some still water (a pond, not a river). Start dropping stones =
near
each other and watch how the ripples interact.</P>
<P>This little phenomenon also explains why string "level" is =
important to
tone quality in unisons. If the strings are not in exactly the same =
plane with
respect to the hammer face, one string will start sounding slightly =
ahead of
the other, i.e., out of phase. So if the hammer shape is uneven, the =
hammer
"travels" and strikes at an angle to the string plane, or the strings
themselves are not exactly level, it is not possible to align all the =
partials
of those two (or three) strings. Even if each string is precisely,
stop-the-lights, in tune at the fundamental pitch, out-of-phase higher =
partials will make the note sound whiney, or worse. (Hello Betsy Ross =
and all
your cheap little friends.)</P>
<P>Ghost Tuning ("Finally," Many Will Say) or, More Precisely, Ghost =
Tone
Interval Testing</P>
<P>This part, I will just present as "how to’s" and examples. The =
idea, in
each case, is to isolate exactly the partials we are trying to match =
in the
two notes of the interval without having to listen to—or =
through—the
fundamentals or any other partials of the notes.</P>
<P>Octaves</P>
<P>Please note that this is not a treatise on tuning, per se, nor on =
what size
octave to use in what part of the piano or anything like it. This is =
simply a
way to isolate the pitch of string interactive beats—for training =
purposes, if
nothing else. Often times, the "best" tuning for a bass note, =
especially in
poorly scaled or tired pianos, will be somewhere <I>between</I> a =
perfect
<I>this</I> and a perfect <I>that—a process known as "evening out =
the
growls."</P></I>
<P>To test a 6:3 octave, hold the two octave keys down, silently, =
(remember,
no pedals) while striking and releasing the 19<SUP>th</SUP> (two =
octaves +
fifth) from the lower note. This will excite the lower string of the =
octave at
it’s 6<SUP>th</SUP> partial and the higher one at its =
3<SUP>rd</SUP>. The
resulting tone will have a beat or a slow roll unless the two octave =
notes are
exactly in tune at the 6:3 interval or so far out of tune that they =
are not
picking up the vibrations of the "energy" key, the 19<SUP>th</SUP>. =
</P>
<P>If our octave <I>sounds</I> like an octave and this test produce no =
beat or
roll, you have a <I>perfect</I> 6:3 octave. For many pianos, =
especially
smaller ones, this makes a fairly clean, nicely "stretched" octave in =
the
bass. The double octave, in this case, will have about a ½ second =
roll that
you can isolate by using the double octave note as your energy =
key.</P>
<P>In our chart example, the C28-C40 octave 6:3 test will use the G52 =
key.
Don’t forget to immediately dampen the "energy" key after playing =
it. We don’t
want to listen to the sound of the G52 key at all! We just want it to =
put some
energy into the other strings, via the soundboard and bridges. Also, =
if your
"energy" key is above the dampers, be prepared to mash your right =
thumb or,
better, a chunk of felt onto that unison as soon as you release the =
key.</P>
<P>To hear the double octave in our example, use C52 as the energy =
key. If the
octave is "in tune" and you hear no beat using C52 for energy, guess =
what! You
have a <I>perfect 4:2 octave.</I> The fundamental of C52 is the =
4<SUP>th</SUP>
partial of C28 and the 2<SUP>nd</SUP> partial of C40. This is, of =
course, the
approximate octave size we want for the temperament and on up into the =
treble.</P>
<P>I will note, at this time, that it gets harder to hear octave ghost =
tones
as you go higher in the piano. Sometimes the lowest bass notes are =
hard to
excite with "ghosts," also. I don’t know why, but I expect it is =
because of
their mass. In fact, my theory is that "tubbiness" in big and fat =
and/or tired
and dirty strings is just a reflection of their inability to generate =
or
sustain higher partials. (Comments welcome.) Anyway, ghost tones are =
really
most useful in the lower tenor and bichord bass regions, in my =
experience.</P>
<DIR>
<P>In the "To err is human, to really screw it up takes a computer."
department: I had to correct my own error in the last paragraph. My =
spell
checker wasn’t the least bit bothered by the phrase "… to hear =
octave ghost
towns ...")</P></DIR>
<P>Tuning a Steinway D or a Böesendorfer Imperial? Try C64 with the =
C28-C40
octave. See how close the piano will let you stretch for 8:4 =
octaves.</P>
<P>Unisons</P>
<P>When string pairs are very noisy, especially in the bass, it is =
sometimes
helpful to tune or test with ghosts. The usual culprit or "bad boy" in =
the mix
is the 15<SUP>th.</SUP> The energy key for this is at the octave + =
fifth. In
our example, if we were tuning two strings of the C28 unison we would =
strike
G47 and try for no beat, no roll. Not that this always produces the =
"best"
unison: As always, <I>listen</I> carefully—then judge, nudge, or =
fudge, as
necessary. </P>
<P>Other Intervals</P>
<P>Though not as useful as for octaves and unisons, other intervals =
can be
tested with ghost tones. The beat we listen for in fifths is found =
<I>one
octave</I> above the <I>higher</I> note of the interval. The beat of =
the
fourth is found <I>two octaves</I> above the <I>lower</I> note of the
interval. For example: The F33-C40 (fifth) roll can be isolated and =
energized
by striking the C52 key. For the F33-A# (fourth) beat/roll, strike =
F57.</P>
<P>Ghosts for other intervals are, in my opinion, hard to hear, hard =
to
remember (where they are), and not any help to the tuner. </P>
<P>Well, that’s it for now, boys and girls. This is Uncle Al, the =
kiddies’
pal, saying: So long, and thanks for listening.</P>
<P>Disclaimer</P>
<P>If this information is correct, coherent, and useful, I wrote it.
Otherwise, it’s someone else’s fault, entirely.</P>
<P>Alan R. Barnard</P>
<P>Salem, MO</P></FONT><FONT size=2>
<P> </P></FONT></DIV>
<P>
<HR>
<P></P>_______________________________________________<BR>pianotech =
list info:
=
https://www.moypiano.com/resources/#archives<BR></BLOCKQUOTE></BODY></HT=
ML>