The tuning of the Great Highland bagpipes (GHB)

Hi all,

Last weekend I attended a friend's kid's birthday party where I had a
conversation with a piper and his pipes. I pumped him for info about the
scale and its precise tuning. Alas, he *only* knew how to play the thing,
:-) and knew very little theory. But of course I did learn that the drones
were one and two octaves below the low A of the chanter, and that the
chanter plays only 9 notes which are called by pipers (and notated on the
staff as): low G, low A, B, C, D, E, F, high G, high A, but in fact sharps
are assumed on the C and F. It would be a D major scale if it weren't for
the fact that the drones are A's. So it's approximately a Myxolidian mode,
but it actually evolved to allow playing in 3 different pentatonic scales
against the drones. G A B D E, D E F# A B and A B C# E F#.

Also, the low so-called A is no longer 440Hz (if it ever was) but is today
typically 466Hz (which makes it a Bb). We won't worry about that.

This piper was kind enough to offer to obtain the tuning information and
mail it to me. In the meantime, I looked up Manuel's archive and found the
following:

! bagpipe2.scl
!
Highland Bagpipe, from Acustica4: 231 (1954) J.M.A Lenihan and S. McNeill

7
!
9/8 !B [I've added these note-name comments. Dave K.]
5/4 !C(#)
27/20 !D
3/2 !E
5/3 !F(#)
9/5 !G
2/1 !A

I looked at that 27/20 D and said to myself, "You've gotta be kidding!".
With those omnipresent drones why wouldn't it be a 4/3 (a comma flatter) or
even an 11/8 (32 c sharper). And why not 7/4 for the G's. It's even
conceivable that the F# might be 13/8 (44c flatter).

If you plot the above as a 5-limit lattice it *is* nice and symmetrical

F# C#
(A) E B
G D

I went searching on the web for more information and found plenty about
bagpipes but not the precise tuning. I learnt of the existence of the
newsgroup rec.music.makers.bagpipe and, after scouring their FAQ to no
avail, posted the question there.

I received an excellent well-researched reply, that raised alomost as many
questions as it answered. It seems GHB tuning has changed since 1954. But
I'll leave that for another day, so you can think about it for a while.

Regards,
-- Dave Keenan
http://dkeenan.com

[Dave Keenan <d.keenan@uq.net.au>]
Dear Ewan, I just received in the mail, from the piper friend who got me
started on this, the official tuning of the Australian Pipe Band College
(below). This was specified by Allan Chatto in 1991. They give only the
frequencies in Hz, I added the ratios:

Hz Ratio to low A
A' 932 2
G' 839 1.800
F 777 1.667 (F#)
E 699 1.5
D 622 1.335
C 583 1.251 (C#)
B 525 1.127
A 466 1
G 420 0.901

[Ewan A. Macpherson <emacpher@umich.edu>, Kresge Hearing Research
Institute, http://www-personal.umich.edu/~emacpher/]
'Official tuning', eh? I've never seen something like this before!
There's definitely no standard pitch these days. For solo playing it
really doesn't matter (except timbrally), but of course bands have to
agree within themselves. It depends what make of chanter you're
playing, what make of reed, the ambient temperature etc! 466 Hz is a
bit low these days.

[Keenan]
I probably overstated its officialness and standardisation. It does allow a
range of frequencies for the low A reference. You can see the whole thing
as a GIF at
http://dkeenan.com/Music/AFPBAtuning.gif

Those G's are clearly 9/5's and no hint of a flattened high A. Behind the
times eh?

[Macpherson]
Well, I wouldn't cry if the flat A were abolished! People may not be
aware exactly what they're doing to get 'that sound'.

Of course just because this is the official tuning, doesn't mean people
are actually using it. For instance I just measured high and low Gs
from the Victoria Police Pipe Band (winners of the World Championship
last year) from a 1993 recording. [Victoria is a state of Australia. DK]
The low Gs were actually a hair flat of 7/8 and the high Gs were bang on
7/4. Their high A was about 10 cents flat which is normal. Bands don't go
tend for the extreme flattening.

I'll try to find you a quote from a book by Royce Lerwick (a
rec.music.makers.bagpipe regular) explaining the flat A rationale in a
band context.

[Keenan]
At least the D is 4/3 and not that whacky 27/20.

[Macpherson]
Just as evidence that the sharp D has not vanished from the face of the
earth, I measured the D from a recording of Dugald MacNeill, who is the
director of the College of Piping in Glasgow and the author of the
Piping Times article from which I got the pentatonic info. It came out
at a ratio of 1.344, which is 8 cents flat of 27/20, but 12 cents sharp
of 4/3. Since there's apparently no strong dissonance minimum there, I
expect it would be hard to 'find' 27/20 exactly.

[Keenan]
Those modern 7/4 G's are a big problem if pipers want to play with 12-tET
instruments. When the pipes' low A is tuned to the 12-tET Bb (466.16Hz) the
errors of 12-tET range from -4 cents (B) to +31c (7/4 G).

Aha! We can minimise the absolute errors by making the pipes' low A 13.5c
sharper than the 12-tET Bb so the errors range from -17.5c to +17.5c
(better but hardly tolerable). I get 469.85Hz as the optimum low A for this
purpose. 470Hz in round figures. It could also be acheived (within 0.2c) by
tuning the pipe C# to the 12-tET D.

Could it be that pipers are tuning the high A to the 12-tET Bb under this
condition thereby making it approx 13.5c flat? But if so, the question of
why still remains.

[Macpherson]
An interesting direction of inquiry! However, piping is, in general, a
pretty closed world. Some of the top soloists probably wouldn't be
caught dead playing with other instruments. I'd guess the pressure for
ET-compliance is close to zero. I'd like to measure the tuning of
pipers who play in folk groups or folk rock bands. From listening, I
suspect they just barge ahead without worrying about the tuning
nicities too much. Some pipemakers do make A-440 chanters for pipers
who have to play with orchestras, or who just prefer the lower pitch.

[Keenan]
I realise now, that even if it were a consideration, the (optimised) error
range has hardly changed.
When the G's were 9/5 the optimum low A was 465.90Hz, giving 12-tET errors
of +-16.6c.
When the D's were also 27/20 the optimum low A was 465.64Hz, giving 12-tET
errors of +-17.6c).

But why go higher than 470Hz, as some do?

[Macpherson]
The rise in pitch has been attributed to the belief that a brighter
sound will give an advantage in pipe band competitions. There seems to
be a growing feeling that this has gone far enough. Again, the pipes
are generally tuned to whatever pitch they happen to be working best at
on a given day - not to any fixed standard.

[Keenan]
I'm sure I speak for many when I say: Thanks Ewan, for your excellent
article and for taking the time to discuss these things for the benefit of
myself and the alternative tuning list.

[Macpherson]
You are most welcome. It's nice to have some interest from outside of piping.

[Keenan]
Manuel, how about some more bagpipe entries in the scale archive, based on
Dr Macpherson's research? We should also make it clear that the 12-tone
scale given in 'bagpipe.scl' is (I assume) for the Uilleann (pron. ILL-en)
pipes.

The Ewan's index to piping sites is at
http://www-personal.umich.edu/~emacpher/pipes.html

But of particular interest to this list are Ewan's own bagpipe research
reports via
http://www-personal.umich.edu/~emacpher/pipes/ewanstuff.html
only some of which have been mentioned in this thread.

On this page Ewan writes:
"If I had a wealthy patron (hint, hint) I'd spend a lot of time studying
the acoustics of the bagpipes and the use of expressive timing in GHB music."

Regards,

-- Dave Keenan
http://dkeenan.com

[Dave Keenan]
Any luck with finding that quote from Royce Lerwick's book, explaining the
flat A rationale in a band context.

[Ewan Mcpherson]
Yeah, here goes:

"... The trick, however, is to set the high A a bit flat from a true
octave - about 10 cents flat. This allows room for overblowing, as you
will not be blowing steady even after years of mastery.

Setting up exactly an octave apart means you will vary sharp with every
step and pothole. The most annoying sound in the world is a sharp high
A. It is better to be ranging a bit flat - even in many minds better
than a true octave if you could blow absolutely steady, because of the
weird harmonics up there - than to ever come close to crossing the line
into the sharp zone. 10 cents is not all that much, and actually, many
big name pipers and big bands made their fame set up as much as 25
cents flat."

- Royce Lerwick, The Pipe Major's Handbook, 2nd ed. p 63.

This all seems pretty reasonable except for the mysterious comment
about "weird harmonics".

-- Dave Keenan
http://dkeenan.com