Once the guitar is clamped to the shaker table, the
process begins with a reading of the instrument’s frequency response. A
reference accelerometer is attached to the aluminum table, and it feeds
into a computer that flattens the frequency response of the system.
Three accelerometers, attached to the guitar with wax, monitor the
frequency response of the instrument at the peghead, bridge, and neck
joint. The guitar is shaken at a moderate rate with a full-range audio
signal to record the "before" response into the computer, and then the
power is ramped up to full intensity. After 45 minutes of shaking, the
force is reduced and another frequency response plot is captured. The
"before" and "after" response plots are then merged, and the graphical
output shows only the change that has taken place (see graph).
Does it work? Are the results predictable? Does the
computer plot show something relevant to how the guitar sounds after
treatment?
To shed some light on these questions, we tested four
guitars: a Sobell owned by Henry Kaiser, a Taylor owned by Laurence
Juber, a recent Gibson Roy Smeck Radio Grande reissue owned by Jackson
Browne, and a Martin HD-28 loaned to us by the factory.
The Sobell is a stiff guitar design to begin with—-I
describe it as being 1/3 archtop and 2/3 flattop. Kaiser said he hoped
it would come out sounding as though Martin Simpson had been flailing
away on it for several years! The verdict? Kaiser gave the shaking a
thumbs up. The Sobell developed more low end, and the high end became
more detailed and open. The computer plot did not show radical changes,
but the changes did correspond with the audible results. The guitar
gained nicely overall and had the "played hard" factor that Kaiser had
hoped it would develop. My feeling was that the guitar felt more
responsive to a lighter touch.
The Martin we tested is a scalloped-brace dreadnought
with a fairly loose top and lots of bottom end. It was passed around
the office at Acoustic Guitar so people could get
used to its "before" sound—which was found to be quite good right out
of the box. It didn’t sound like a guitar that particularly needed
playing in.
The Martin changed more in terms of overall
sensitivity than frequency response, and that is admittedly a somewhat
subjective view of it. Jeffrey Pepper Rodgers and Dylan Schorer,
working with engineer Steve Bird, set up as controlled a recording test
as was practical, and their report is as follows:
"We recorded it before and after the shaking, with
stereo mics in the same position relative to the guitar, run through a
mixing board set flat and into a DAT machine. We used new strings (same
brand!) both times, and recorded the same series of songs, played as
consistently as possible. The difference between the before and after
tapes is hard to pinpoint and not necessarily consistent from song to
song, but we did feel that the guitar was ringing out more in the
second session. Even before the shaking, we had to set the mics to the
highest bass roll-off to get a decent sound; afterwards, if we had been
trying to get the best sound on tape (rather than to recreate the way
we recorded it the first time), we would have opted to move the mics
further away from the guitar to cut back the boom. The guitar sounded a
bit bigger.
"Our ‘before’ and ‘after’ comparison is obviously
subjective and fraught with variables (the humidity, how hard the
guitar was played during the sessions, our expectations, our memories
of what the guitar sounded like before), but the bottom line is that
this guitar did seem to be opened up by shaking. Afterward, it seemed
to make more sound with less effort—a subtle but nice change."
These impressions of the HD-28 shake were very much in
line with my own. The guitar was loose and boomed out on the low end to
begin with; it did not sound like a typical new, stiff guitar. But what
I noticed was an increase in overall sensitivity. Nice old instruments
sometimes feel like they are playing themselves; it takes very little
effort to bring out a great sound. The shaking increased that factor in
the Martin just as it did in the Sobell; these were remarkably similar
results in two very different instruments.
The most dramatic change was in Laurence Juber’s
Taylor 514, a small-bodied cutaway with mahogany back and sides and a
Sitka spruce top. Before the shaking, Juber liked its
balance—especially for recording—but felt that it definitely sounded
like a new guitar. Afterward, Juber says, "It gained another octave of
low-end tone, and the highs became much more complex. My wife, Hope,
noticed it immediately when I brought the guitar home. It was like when
an adolescent’s voice breaks and becomes mature." Juber also reports
that his use of the guitar in studio work has greatly increased, as its
sound is very mic-friendly. Also interesting is that Juber feels that
the guitar is continuing to mature more rapidly than he would expect
since the Timbre Tech treatment.
Finally, Jackson Browne’s comment on the shaking of
his new Gibson Roy Smeck was short and to the point: "Wow!" He
elaborated that the guitar now sounds much more like his vintage
Smecks. Even the old ones sound different from one another, but now the
new one sounds as if it has been played in. Browne now wants to try the
shaking treatment on a little-played 1935 Smeck.
Timbre Tech has also treated solid-body electric
instruments, again with interesting results. Aerosmith and Eddie Van
Halen have recently had instruments shaken, and Jerry Donahue reports
spectacular results on two of his Fender Custom Shop Telecasters. With
solid-bodies, it seems that the most dramatic results are with
bass-wood–bodied instruments, and run-of-the-mill new production
guitars get more improvement than instruments already judged to be
excellent. The changes in solid-body instruments would seem to indicate
that it is not just simple flexural patterns that open up with
vibrational aging of guitars. Internal sonic wave patterns in woods
also change with shaking, whether done naturally over time or
accelerated on the shaker table.
Steve Rabe of Timbre Tech has been experimenting to
see if it is possible to "move" dead spots to less musically intrusive
frequencies. By emphasizing or de-emphasizing particular frequencies
during the shaking treatment, it might be possible to suppress wolf
tones and bring life to the dead spots. It might be further possible to
program what kind of improvements might be made to an instrument. For
instance, by not exciting low frequencies and concentrating the shaking
in the mid- and upper frequencies, it might be possible to balance out
a too-boomy guitar.
I have talked to musicians who felt that guitars age
according to how they are played. An instrument strummed in open
position will age one way; the same guitar fingerpicked will turn out
different. As far out as this may seem on the face of it, it may just
be true. Since the Chladni patterns are frequency-specific for a
certain size and thickness of plate (think "top"), a musical style that
emphasizes one key or frequency range over another will excite a
particular set of vibrational patterns. Thus different vibration
programming might age guitars according to the music you want to play.
Will that be the Doc Watson tone file or the Reverend Gary Davis?
Richard Thompson or Lightning Hopkins?
What about predictability? So far, the results on the
guitars Timbre Tech has shaken seem to indicate that the most
noticeable results are on newer, stiffer instruments—just the ones in
need of the most aging. There is not enough data yet to come to any
conclusions about different effects with the different woods commonly
used on acoustic guitars. We might expect, though, that the aging
effects noted naturally might just be accelerated with the shaker
treatment—for instance, there might be more opening up of a thicker
spruce top than a thin cedar top.
The results heard by musicians and engineers seem to
track fairly closely with the computer frequency response charts made
before and after treatment. However, improvements in overall
responsiveness do not show up so obviously with current measurement
techniques. Further testing using laboratory mics with even more
sophisticated computer programs may reveal the increase in acoustical
efficiency and sensitivity that all observers have felt and heard. As
in the rest of life, the right answers are right in front of you if you
ask the right questions.
Rabe is working with scientists at MIT to try to
determine if there are observable changes in the wood that has been
shaken. Samples of different woods will be shaken and compared with
control samples of matching cuts from the same planks left unshaken.
Hopefully the combination of acoustical testing along with electron
microscope viewing of the wood cells will result in documentation of
what the musician’s ear knows. For now, the ears have spoken: "This
works."
Excerpted from Acoustic Guitar magazine,
February 1997, No.50
