Simple enough idea: Generate words with the Synthtech E950, send them through MultiMod and mix them. I took the eight MultiMod outputs by pairs through DABS, so I could switch them in and out, and mixed them in SSF Jumble Henge. After that I passed the signal through Smoothie Audio SMEAR and then through Qu-Bit Data Bender. I found that four or six MultiMod outs worked best.

Simultaneously, a steady clock triggers a new word, selected at random. Data Bender (when engaged) is clocked by this same signal. MultiMod is not clocked, and the settings are fixed — no CV going into MultiMod. I left SMEAR at the same settings.

I tried something interesting, which was to mix the word coming out of E950 with two of the outputs from MultiMod and then into the MultiMod input. This resulted in an echo effect, due to the feedback of MultiMod into itself. There is no delay or reverb applied.

I recorded several variants.