Some considerations about bag pipes - long ? (was Bagpipe samples)
2004-12-06 by Gérald RYCKEBOER
I go on with the previous thread as it seems some of you folk are
intersting in ;-)
Well just as a short introduction, if I do use computers for music
now for more than 10 years, I'm a piper first and play mainly
traditional music ; that explains that pipes are "my cup of tea" ;-)
Something else is that my technical knowledge is nothing else than
fragmentary (made of what I could pick up here and there along my
expericences and problems ;-)) as I don't work and play in a
technological environment.
Some short general things : there're quite a lot of pipes that can
be completely different one from the other (sound, organology, blowing,
tuning, playing, scale). The pipes that everybody knows (which are
these related in this thread) are the great scottish Highland bagpipes,
but, even in Scotland, there are other ones, called "chamber pipes" or
"small pipes" which are quite less powerful and can have quite
different sound. Now another kind of pipes are a bit more commonly
known, because of the "celtic fashion", which are the irish uillean
pipes, and that have nothing to do with the great pipes (they have even
a keyboard on it to play accompaniement notes and chords - with the
wrist or one hand fingers).
Bagpipes had been common - shematically - in all Europe (from the Oural
mountains to the Mediterranean sea), North Africa and Middle East up to
India.
But although there are big differences between all of these
instruments, all the pipes have jointly two things : a bag, allowing
continuous sound (on the drones first, but also on the chanter) and
drones, for a steady accompaniement of the melody.
Well I'll not develop on pipes technicals problems, but just talk
of some of them what can help to understand the instrument and its
playing :
First two problems that make phrasing difficult on the instrument
- there's no dynamic modulation : velocity is 0 or 127, nothing between
… (but same problem with the harpsicord for example) ;
- sound don't stop - so it's legato sound (except with a few kind of
pipes like uilleann pipes, some small pipes and bulgarian pipes that
can stop the chanter) ;
So these technical limitations explain the big use of ornementations
(grace notes) to suggest staccato notes or dynamic modulation, the
scottish pipes have certainly the more tricky ones (we can have 5 grace
notes, or even more, to attack just one melody note). And unlike in
classical music, these grace notes are not melodic ornaments but are
sound effects for rythmic and dynamic purpose (make "stronger" notes to
suggest dynamic stress), and are most of time very short (some 1/100 of
second).
- depending of the pipes, never more than two diatonic octaves with a
modal scale and eventually some accidents (Bb mixolydian for the
scottish great pipes from Ab to Bb one octave higher) and never equal
temperament (because of drones) ;
- there's always harmonic "alchemy" (acoustically speaking) between the
drones and the chanter's which interact, and something as a "sound
bubble". Something else is that the sound (harmonics) change with
pressure variations.
This last point still makes the pipes not easy to record and can result
in a "schizophrenic sound" ; never had a good recording of any pipes if
drones and chanter are recorded separately ( and even difficult to play
as the piper plays "on the drones" (or better "in" the drones) as
tuning reference) … But for the sampling recording purpose it's always
possible to stop the drones or the chanter and record the other (with
corks ;-) - and even some pipes can also play drones and chanter
individually, like the uilleann pipes, but the problem can be to be "in
tune" (what can be done is to record first the drones and then record
the chanter's scale with the drones in the headphones) ; but the
harmonic interaction between drones and chanter will be broken … well
it's a big limitation in the pipes sampling credibility.
So Gramophone's and Sacha's suggests are possible but with a dodgy
problem with the loop points of the drones : I've made a lot of
editing on acoustic instrument, but the really most difficult job is to
do it with pipes because of the drones whose sound is always changing
- lightly but quite enough to give problems (especially with the great
Highland pipes whose drones are very "present" - in front). In such
case I've added some synth's sound (as suggested also) to attenuate the
break.
With the split keyboard, this drones problem can be reduced with a long
drones' sample.
I have to test second's Sacha's suggestion to trigger the drones with
chanter's notes which is attractive but the problem of the looping
points of the drone will be more acute here because with this way of
doing the drone will be retriggered at each new melody's note (and
grace note).
Also another way of doing would be to record each note of the chanter
with the drones (not a big job : as the scale is never bigger than two
octaves). But the problem of the drones will still be there (and I feel
it even more problematic that way). Something else to test.
As a conclusion, the number of parameters in action in pipes'
playing give a real challenge for real convincing sampled pipe music,
although technically possible - IMO it would need a dedicated
instrument (but it's anyway the case for a lot of acoustic instruments
isn't it ? ;-)
I 've had the project of making some sampling experience with my
different pipes (I play 7 different ones) and I can think of having it
sooner under construction and let know the list when it'll be done, if
some of you that are interested in (but don't be too much in hurry, I
have still a job's waiting list :-D
--
Gérald