~the following categories in addition to a fixed-voicing
mode.
The second category, 'smart harmoniser' allows
the user to specify a diatonic scale and voicing (ie
adding a single diatonic third above, or both a fourth
and sixth below input pitch) and will produce the
pertinent major/minor/augmented/diminished intervals
that are diatonic to the nominated scale. Chromatic input
pitches outside this scale are dealt with in differing ways
depending on the individual product. The Boss PS-6
effect pedal, Eventide H8000 and audio plugins such as
Antares Harmony Engine Evo offer this kind of
functionality.
Effects in the third and final category are
program-reactive, meaning they accept the live
instrument input plus an additional 'program' input
(generally a keyboard or guitar playing chordally) from
which it infers the appropriate interval choices to again
produce diatonic harmony. The Boss VE-20 and TC
Electronic Voicelive series are prominent hardware
examples.
From the perspective of an improvising
musician, the more advanced models of commercially
available pitch shifter could be considered to restrict the
ability to improvise freely. The need for a
predetermined key centre or scale/mode is prescriptive
in itself, but beside this remains the simple fact that an
improvisatory musical language may not be strictly
diatonic. Fixed-voicing pitch shifters have the opposite
effect, in that they immediately restrict the performer's
ability to produce diatonically harmonised passages at
will, and at times the harmonic functionality of a fixed
interval voicing becomes less prominent than the more
generalised timbral effect that results from its consistent
quality.
3. DESIGN OF THE CONTROL MECHANISM
AND SOFTWARE
3.1. Control Mechanism
The use of the pitch shifter on brass and woodwind
instruments is by no means unheard of; Don Burrows'
Octavider clarinet on LPs recorded with George Golla,
Randy Brecker's octave-up and perfect fourth effects
and Jon Hassell's stark fifths and parallel triads are clear
examples in the jazz and fusion idioms. These effects
relied on single or multiple presets of the pitch shifter
without necessarily requiring complete real-time control
over voicing and density.
My concept for a pitch-shifting harmonisation
effect for saxophone evolved from my existing usage of
a MIDI foot controller with multiple footswitches, and it
was partially inspired by an organist's bass-pedalling,
partially by the bass and chord buttons actuated by an
accordionist's left hand.
As both hands are generally required for
saxophone playing, in the earliest stage of development
I focussed on readily available MIDI foot control
products from Roland, Line 6 and Behringer as possible
means by which I could select predetermined voicings
based around major/minor/diminished/augmented triads
and inversions. I soon realised that the net effect of this
implementation would be limited to the production of
simplistic harmonic material rather the complex and
varied harmonic language I wished to draw upon in a
solo setting. Ideally, a broad range of chordal
possibilities, including cluster chords and stacked
perfect fourths or fifths as well as octaves, unisons and
simple thirds or sixths needed to be available on
demand. As the permutations of these voicings rose
exponentially, I was forced to consider an interval-array
approach which allowed free selection of voicing using
multiple footswitches at a time, where each footswitch
would be assigned a set interval in relation to the live
performance pitch.
Using a commercially available twelve
footswitch MIDI controller it would be possible to
assign one octave of chromatic intervals below (or
above) the live input to the footswitches and create
various voicings using these in combination. For
example, to produce a simple D minor seventh voicing
consisting of the notes D, F, A and C (lowest to highest)
I could play the C live and then depress the footswitches
representing, respectively, intervals of 3, 7 and 10
semitones below this note.
I began to foresee how certain musical and
physical barriers could hinder success in this approach.
Firstly, only closed voicings (i.e. those within a range of
one octave) would be available without introducing
further mechanisms to spread the transposition range
according to predetermined criteria (for example, a
mechanism to prescribe 'drop 2' and 'drop 2, drop 4'
voicings depending on the current state of a dedicated
switch or potentiometer). This would add an additional
layer of complexity to the system, and another
limb/finger/gesture would have to be responsible for
changing between closed and open voicing schema,
given that both feet are utilised for note selection.
Furthermore, the allocation of intervals to the
physical arrangement of the footswitches on existing
MIDI hardware posed another problem. Products
featuring at least two rows of six or three rows of four
footswitches were ostensibly suitable, yet the pedal
technique required became difficult when two feet were
required to depress multiple footswitches
simultaneously. In a simple paper-based analysis where
I drew diagrams representing common 3 and 4 note
voicings on templates representing the layout of the
MIDI foot controllers in question, systems that arranged
the array of intervals in whole tone rows, or in perfect
fourths/fifths (as in the bass notes of an accordion)
proved equally unwieldy.
3.2. Hardware interface
Considering the above evidence, I accepted the reality
that a customised hardware solution would be necessary
for the success of this project. The physical layout of
existing foot control products did not mesh well with the
interval-array concept I had developed. Ideally, intervals
that were often played together needed to be closely
322 2013 ICMC
