ECMC PVC example file inharm2

ECMC PVC and /sflib/x example "inharm2" is a linear mix of 4 source examples: inharm2-1, inharm2-2 inharm2-3 and inharm2-4 In each of the 4 resynthesized tones

   1) the duration of the original bassoon tone is reduced by 25 %
   2) the INHARMONATOR_fundamental_frequency argument is set to 110,
matching the fundamental frequency of input soundfile sflib/wind/bssn.a2
3) Even numbered partials 2 through 20 are defned as the target partials, but no frequency or amplitude alterations are specified for these target partials. Instead, a frequency SHIFT of +45 herz is applied to the 10 target partials
TARGETS_frequency_shift=45
(shifting the frequency of each of these partials up by 45 herz) while a different frequency shift of -95 herz is applied to the remaining (non-target) partials
NON_TARGETS_frequency_shift=-95

Differences between the 4 resynthesized tones in /sflib/x/inharm2:
In example file "inharm2-1" (which produced the first note in /sflib/x soundfile example "inharm2") no attenuation is applied to the non-target partials:

NON_TARGETS_gain_in_dB=-0

In example file "inharm2-2" (which produced the second note in /sflib/x soundfile example "inharm2") maximum possible suppression is applied to the non-target partials, resulting in a simpler timbre than in example "inharm2-1"
NON_TARGETS_gain_in_dB=-96 # int, float or FUNC

In "inharm2-3" 2 out-of-phase functions, each with 3 peaks and 3 nulls, are created and applied to the gains of the target and non-target partials:
 gen5 -L100 3 1 0   | reshape -b-96 -B3 > $SFDIR/tremolo1
 gen5 -L100 3 1 180   | reshape -b-96 -B3 > $SFDIR/tremolo2
 NON_TARGETS_frequency_shift=-95
 NON_TARGETS_gain_in_dB=$SFDIR/tremolo1 # int, float or FUNC
 TARGETS_frequency_shift=45
 TARGETS_gain_in_dB=$SFDIR/tremolo2 # int, float or FUNC
Ihis results in 3 "cross-fades" between the 2 detunings of the bassoon tone. However, for reasons that are not clear to me, much of the original, unaltered bassoon spectrum remains. - - - - - - - - - In example file "inharm2-4" no frequency shifts are applied to either the target or non-target partials. Instead the target partials begin at the original pitch level, but then glissando up in pitch a half step.
gen1 -L1000 0 0 .1 0 .9 1. 1. 1 > $SFDIR/rampup
TARGETS_pitch_transposition_in_semitones=$SFDIR/rampup

Example file "inharm2-5" is similar to "inharm2-4," except that here a descending one octave glissando, beginning a major 7th above the original pitch level, is applied to the target partials and an ascending one octave glissando is applied to the non-target partials: