Cryptwarbler, live at Red Ink Providence RI, using a variety of bends.
A13+D6 / Percussion: Bb3 - hammer rapidly to stagger a warped suspended cymbal sample. D0, D1+V & D5+G (after D6+A13 off) in various combinations to cycle the sample at various speeds, pitches, directions, to improvise on. The timings affect each staggered sample differently, scattering each staggered sample's begin time, as setup for the next part. Above a resistance threshold of ~120ohms, Vcc+A13 to 17 speeds up the cycle greatly, which sounds like a fast scrub thru its various subsections. Below the threshold, the cycles seems to freeze and loop on one of the subsections. Because the prior scrambling, each sample freezes at a different subsection- which forms a dense polyrhythm as each subsection has a different cycle duration/pitch. Right at the threshold (very precisely, a few ticks on a 10 turn 100 ohm pot will tip it over), the cycles will freeze, jump to a new subsection, and freeze again. How long it freezes for is random, possibly due to slight fluctuations in voltage causing it to tip over. Alternatively, Gnd+A12 to A17 in various combinations lets you exclusively change the timbre: between its 'normal' timbre and one alternative timbre, with white noise in between. You cannot connect an Address line to both Gnd and Vcc at the same time (this would cause a short circuit). However, certain combinations of A+V and A+G produce the same sound, which means that these shared sounds can be used as pivot points, freeing up certain address lines to be switched off or used in the other mode.
'Flute' chord that slowly shifts in harmony. When a D3-D8 (Data) line is connected to 5V, it triggers an event which causes all voices to independently gliss up, gliss down, or not change at all. The musical interval of the gliss corresponds to the D line being switched. D8 causes a two octave gliss, D7 - one octave, D6 - maj 3rd, D5 - maj 2nd, D4 - min 2nd, D3 - quarter tone. Because each voice glisses in a direction independent of other voices, the harmony will shift in a random direction but under a predictable interval. For example, say you two voices, C and G, creating a perfect 5th. Now, switching on D6 causes a major 3rd gliss up, down, or no change. The C can then go up to E, down to Ab, or stay on C. The G can go up to B, down to Eb, or stay on G. Thus we have 9 chords it could possibly shift to: EB, EG, EEb, CB, CG(no change), CEb, AbB, AbG, AbEb. The synth can have 10 simultaneous stereo voices and the L and R channels also act independently, so a harmony with 20 moving voices is possible. Other intervals can also be glissed if the switches are activated simultaneously. All of this makes for an interesting control scheme where the musician is only in control of the interval at which a harmony's voices change and when to change them, but not the direction in which the interval changes.
A1+D14 / Vibraphone: A#3-G#4 - the frequencies follow the pattern Fn = Fn-1 * 2^(1/12) + F0, where Fn-1 is the frequency of the previous note and F0 is the base frequency. Because each successive frequency is a multiple of the base frequency, it sounds loosely reminiscent of the harmonic series. But for each ascending key the keyboard is also adding what it thinks is a semitone in the plus 2^(1/12), so it kind of sounds like the harmonic series is being inflated or expanded. Other bends used: A14+Gnd to change the timbre to a bell-like tone; D4+A9 to add a tom-like tone and also slightly change the bell-like timbre; A2+D7 to make the attack longer; D2+A13 to gliss up; D1+Gnd to gliss down; CLK adjustments throughout to change the character of the effects.