Vocal tract effects (formerly RE: dB SPL testing)

[JOHN THADEN <JJTHADEN@xxxxxxxxxxxxx>]

In a recent installment of "RE dB & SPL" Barry Schaede wrote:

>. . . the testing has led me to discover an interesting 
>effect . . . . Using a regulated air supply I . . . was unable to get 
>lower notes of a Suzuki D harmonica to sound at all.  The windsavers 
>would resonate . . . but the reeds would not.  I did have success 
>with the higher blow notes . . . . What was the difference?

It is interesting, isn't it?  The mystery (and many others) was cleared up
for me by the article  "Pitch control in harmonica playing", by Robert B.
Johnston (1987) in Acoustics Australia, vol. 15, no. 3, pp. 69-75.

He also used a regulated air suppy.  Besides a water manometer (to measure 
pressure), his device had one more important feature: the effect of changes
in the vocal tract geometry was simulated by a variable-length cylinder in
the air supply.  The air came into it through a narrow hole and the length
was varied by sliding a tube inside another like a trombone.  The tubes,
by the way, were 8 mm (5/16ths inch) in diameter.  He varied the length of
the cylinder from 0-500 mm (0-20 inch

>My initial thought was that the resonant frequency of the 
>air supply I was using was too high to sound the reed.

Good guess.  From Johnstons results, it seems to be the resonant frequency
of the "vocal tract" that is important.  Your air supply had too big or too
small a "mouth" to play those reeds.

>I also had the thought that perhaps it was the interaction of the windsaver
>that caused the problem.  The upper reeds on the Suzuki's are not valved.

The math that describes vocal tract effects is rewarding to study, but a bit
much to describe here (and to re-read and recomprehend now!).  Almost anything
you do to a reed (changing its gap, certainly filing it, and probably valving
it) will change the parameters that describe its behavior.  Hence, the optimal
vocal tract geometry to make it sound will be different.  But the effect you
described here is more basic.  The low reeds choked because your air supply
geometry happened to be "wrong" for playing low notes.  Change it, and it
could well be the high notes that crap out.

>I can reproduce the  no-sound  effect by blowing myself. . . . it seems to
>be more of a trying to force a note to overblow that cannot.

Nice observation.  Since the first requirement of an overblow is that
you choke off the normal blow reed (called the closing reed), by proper
combination of vocal tract geometry and air pressure), the valves may well
have stopped your air supply from popping a nice overblow.
                              
>I truly don't have an answer to this question but I am intrigued and welcome 
>any informed input on the subject.

Hope I helped.  Johnton's article is a gem.  He fully explains both bends and
overblows/draws.  I'd recommend you order it through your library interlibrary
loan office (My copy is so unreadable it will not sustain another round of
copying).

Regards,

John Thaden

Oh yes and as for the results of the soaking.  No discernible difference 
in volume between a wet Suzuki and a dry one.  This using a regulated air 
supply on the number 7 and 8 blow reeds of a key of D valved Pro Master.  
I will say this, tiny differences in proximity affect the dB readings much 
more than the relative humidity of the interior of the harmonica. If 
there is a difference it's less than 1 dB and not discernable with my 
test equipment or the human ear.  The disclaimer being of course that I 
didn't go valve any of my Lee Oskars to conduct this test nor did I run 
out and buy a 4 track to measure the levels. Different methods, different 
results.  Enough already.  Sorry for the bandwidth.  FJM