Kevin Flanagan Electro-Acoustic Compositions



REAL PLAYER REQUIRED
tracks optimised for 56 kps modems

PLAY 1. PMOOD1
3.36
2. PMOOD2
3.48
3. PMOOD3
3.46
4. HOOT4a
3.56


1.
****************************************
instr 8 ; pluck-sawtooth
idur = p3
iamp = p4
ifrq = p5
icut1 = p6
icut2 = p7
iresgn = p8
ipan = p9
; irvbsnd = p10
kcut expon icut1, idur, icut2
kenv linen p4, 0, p3, .5
aplk pluck iamp, ifrq, ifrq, 0, 3, 1 ; 0 3
abpf butterbp aplk, kcut, kcut*.2
alpf butterlp aplk, kcut
amix = alpf+(abpf*iresgn)
outs amix*ipan*kenv, amix*(1-ipan)*kenv
gavrb = gavrb+amix ,(amix*irvbsnd)
; dispfft amix, idur, 1024

endin.

2.
****************************************
; Planet Position (X, Y, Z) & Velocity (VX, VY, VZ)
kx init 0
ky init .1
kz init 0
kvx init .5
kvy init .6
kvz init -.1
ih init p5


; Star 1 Mass & X, Y, Z
imass1 init p6
is1x init 0
is1y init 0
is1z init p8

; Star 2 Mass & X, Y, Z
imass2 init p7
is2x init 0
is2y init 0
is2z init -p8

; Calculate distance to Star 1
kdx=is1x-kx
kdy=is1y-ky
kdz=is1z-kz
ksqradius=kdx*kdx+kdy*kdy+kdz*kdz+1
kradius=sqrt(ksqradius);
Determine acceleration due to Star 1 (AX, AY, AZ)
kax=imass1/ksqradius*kdx/kradius
kay=imass1/ksqradius*kdy/kradius
kaz=imass1/ksqradius*kdz/kradius

; Calculate distance to Star 2
kdx=is2x-kx
kdy=is2y-ky
kdz=is2z-kz
ksqradius=kdx*kdx+kdy*kdy+kdz*kdz+1
kradius=sqrt(ksqradius)

; Determine acceleration due to Star 2 (AX, AY, AZ)
kax=kax+imass2/ksqradius*kdx/kradius kay=kay+imass2/ksqradius*kdy/kradius
kaz=kaz+imass2/ksqradius*kdz/kradius;
Update the velocity
kvx=kvx+ih*kax
kvy=kvy+ih*kay
kvz=kvz+ih*kaz

; Update the position
kx=kx+ih*kvx
ky=ky+ih*kvy
kz=kz+ih*kvz

aoutx = kx*kampenv
aouty = ky*kampenv
outs aoutx*kpanlfo, aouty*(1-kpanlfo)

endin

3.
****************************************

instr 2; Bells

idur = p3
iamp = p4/11
; scale iamp to the number of oscillators
ifqc = p5
ipan = p6
ae1 expon iamp, idur, 1
ae2 expon (iamp* .67), (idur*.9 ), 1
ae3 expon iamp, (idur*.65 ), 1
ae4 expon (iamp*1.8 ), (idur*.55 ), 1
ae5 expon (iamp*2.67), (idur*.325), 1
ae6 expon (iamp*1.67), (idur*.35 ), 1
ae7 expon (iamp*1.46), (idur*.25 ), 1
ae8 expon (iamp*1.33), (idur*.2 ), 1
ae9 expon (iamp*1.33), (idur*.15 ), 1
ae10 expon iamp, (idur*.1 ), 1
ae11 expon (iamp*1.33), (idur*.075), 1
a1 oscili ae1, (ifqc* .56), 11
a2 oscili ae2, (ifqc* .56+1),11
a3 oscili ae3, (ifqc* .92), 11
a4 oscili ae4, (ifqc* .92+1.7),11
a5 oscili ae5, (ifqc*1.19), 11
a6 oscili ae6, (ifqc*1.7 ), 11
a7 oscili ae7, (ifqc*2 ), 11
a8 oscili ae8, (ifqc*2.74), 11
a9 oscili ae9, (ifqc*3 ), 11
a10 oscili ae10,(ifqc*3.76), 11
a11 oscili ae11,(ifqc*4.07), 11
amix = a1+a2+a3+a4+a5+a6+a7+a8+a9+a10+a11 ; mix
outs amix*1pan, amix*(1-ipan) ; stereo outs with panning

end


1.

This portfolio consists of compositions constructed from two differing sets of compositional parameters. The first, the Pmood series, is built around a group of instruments created in the Csound program. The original idea was to use a pitch series derived from the Charlie Parker piece 'Parker's Mood', hence the name. This was used both as a pitch class and as a rhythmic basis.

The instruments were of three basic types. Some, such as 'Grain' and 'Pluck' were constructed around suggested basic exercises from Eastman and Boulanger. A second type, such as 'Binary Stars' and 'Low Pulse', were taken from existing catalogues of instruments (such as the Amsterdam and Mikkelson) and modified for the purpose at hand. A justification of this could be that no composer working with acoustic instruments would be expected to invent or even construct any part of his orchestra (unless he was Harry Partch). Therefore no loss of face would be attached to the assigning of parts to an existing concept of, say, 'cello'.

A third type of sound used was that of synthesised analogue instruments, such as 'Low pluck' and 'Bells', from the Csound catalogue. These were modified along the lines of the other instruments with the addition of split signal for stereo, which then allowed the manipulation of spatial elements such as panning over the duration of the note with an envelope preloaded as an adjustable LFO. There was also the need to bring them in line with the audio, control and sampling rates being used in the rest of the composition. Each change in the comparative balance of these parameters brought about a slightly altered sound quality.

The problem that became apparent with these was the clash of familiar sounds with a more electronic ambient landscape. Each newer version used them less and less as the synthesised textures originally conceived as background gained in importance; the bells, for instance seeming more out of context as time went on. An example of the first type of instrument was chosen for its apparent simplicity. As features were added, it became one of the more problematic instruments in this series. This was the sound called 'pluck- 2.

This was created from a basic exercise in the Boulanger Csound tutorials, starting with a simple sound and adding a few more interesting features. Panning, reverb, stereo, filters, and a sawtooth waveform (for a more interestingly dirty sound) were some of the modifications added to the template. It proved incredibly flaky, functioning well at certain times but somehow creating a lot of release envelope clicks only audible after being rendered into a waveform and heard on studio monitors. Again, this was chased around for some time, changing the envelopes to linear from exponential, trying the more basic comb filters as opposed to the Butterworth, dumping all the effects such as reverb, and reverting back to the original simple waveform, as opposed to the sawtooth GEN table.

The 'Binary star' instrument is an example of one of the complex 'borrowed' sounds. Taken from the Csound Book's catalogue of possible sounds it came with a disclaimer as to its reliability, or even its ability to function at all.

It is based around non-musical criteria, that of the calculus needed to determine complex planetary orbits of multiple objects; something which most musicians (or at least me) cannot have any hope of understanding. However, if we think of the Keplerian 'music of the spheres', and try to see how the mathematics describing space-mass relationships can be translated into vibration rates, and so into sound, this becomes a brightly colored mobile to be set in motion above our playpen. And looking at the maths involved, that is exactly how it makes one feel. It was this relation to Kepler's theory that led to the exploration of a folder of sounds based on planetary, fractal, and chaos-based maths. Of the dozen or so equations offered to produce sounds,' Binary Stars' proved the most interesting.

My input for this instrument was primarily concerned with practical problems; the first task was to get the thing to run. This input is visible as the areas separated from the main body of calculus by asterisks. After a few days of experimentation, I hit upon the idea of simplifying the envelope to make it less complex. It was changed from exponential to linear, which was then cut back to a much simpler shape again. This cut the amount of calculations necessary to run it by a huge amount. The equation would now run without exploding and grinding to a halt. The fascinating thing about it was that very, very small changes in the parameters of mass or velocity would produce wildly varying and unpredictable results. It thus became an almost aleatoric element in the composition. Once the instrument would run, it was therefore needed to have it conform to some of the parameters already present in the piece. First of all it had to run in stereo, in order to use some of the spatial effects already present in the piece. If you look at the final three lines of the instrument, you can see how this was accomplished by splitting the outputs. This then allowed the use of panning, which was created by using a low frequency oscillator (LFO) and inserting more parameter fields to enable the sound created some illusion of motion in space. This can be seen in the first 10 lines of the instrument. The use of effects such as echo and reverb had already been added to the piece as a whole, at a global level. The stereo effect chosen was that of a rapid oscillation, to contrast with some of the other instruments, such as the 'low pulse', which had a simpler, single slow pan specified over its duration .

An example of third type of instrument can be seen in 'Bells'; this was taken from the Mikkelson catalogue of instruments. It represents an attempt to re-create an acoustic instrument by synthesis. As you can see, the program consists of a series of cascading harmonics with their own envelopes. Initially, this sound proved too complex for my computer to realise in real-time calculations using the DirectX Csound front-end. Before adding parameters for stereo mixing, the expon envelopes were replaced with simpler 'linen' opcodes, but this completely lost any bell-like qualities to the sound, if it ran at all.

Again, the use of newer versions of the WINSOUND front-end allowed the use of the original 'expon' opcodes to create the bell. The problem being, as mentioned above, that the 'analogue' sounds seemed increasingly out-of-place in the evolving ambient soundscape. Although the initial Parker motif was originally stated using the bell sounds, by the final version of the piece (Pmood3) the bells had been discarded save for a single toll at a point of climax. This was also the case with he other 'analogue' sound, 'Low Pluck', which was a reproduction of a pizzicato acoustic bass. It went from quite dense usage in Pmood2 to just a couple of discrete punctuations in Pmood3.

Kevin Flanagan. Jan 2001