For the bass I just used the WAVESTABLES #6, basically the square wave table. I modulated it with a slow triangle lfo, I also used the second oscillator playing an octave above the bass. From there I filtered it through the WASP filter. From there to the AD envelope with a 2 SAMPLE-&-HOLD modules modulating both the attack and decay of the AD envelop. The envelop's general attack and decay it set to approximate an upright bass. The Sample and HOLD Modules add that human-like imprecision to the sound.
Also the beat is being divided by the Fibonacci sequence. It sounds much more natural that way.
The same two S&H modules are summed in Kurt's GREAT DIVIDE module, then quantized to a major scale.
The main solo voice is composed similarly but I have the rbss, a single S&H, and two slow triangle lfo unsynced, moving at slightly different rates. All that is summed in Kurt's in GREAT DIVIDE as well. At is also quantized to a major scale. I used WAVETABLES #1 also modulated by a slow triangle LFO and filtered with a STATE VARIABLE FILTER
I use only one SEQ16 set to B, G and Eb, with a ii-V-I at the end in C#-F#-B. This was quantized and run through a CVADDER to make the key changes.
The twinkly sounds were just random quantized pitches from another SEQ 16 going through the BBG DELAY. I used a triangle wave from the TH555 going through a wavefolder and LPG. Here's I would have much preffered CIRRUS and RAINS, as that would have produced a far deeper patch than just some random twinkles. Imagine RAINS strumming along with the chord changes! Imagine CIRRUS granularizing this entire patch. That in and of itself could have produced a far more interesting beginning and end to the patch.
The congo, drum sound was my least favorite thing in this patch, but I don't have a drum sequencer or TOPOGRAF. I just used FMOS sinewave through a wavefolder and a LPG with another set of random control voltages summed via POLAMIX to try an simulate a drum set. Hopefully MARBLES/PEBBLES will make this portion a lot better.
NOTE: All the filters are CV controlled via feedback off the VCA. This gave me further control over the resonance. I actually forgot to play around with that during the recording. I also use the end of cycle out to as the second CV input on each filter. I forget how I modulated the WAVEFOLDERS.