I created this piece today by starting with a few percussive micro samples from a recent field recording and using them as percussion. After programming a simple pattern I ran it through Pluggo’s Harmonic Filter and resampled the results to get two musical loops; one with a low frequency hit that forms the bass in the piece and another with a simple descending melody.
I put a triplet ping-pong delay on one send and an enormous reverb on another then made a quick 1:40 minute mix to hear how it all sounded together. I haven’t explained all the details by any means, so if you have any questions about what you’re hearing post a comment and I’ll give you more information.
Oxotremorine
One of the things I like to do with a few friends is ride bikes to a spot along the Mississippi river, light a fire, and have a few beers. It’s the best happy hour in town. At this particular spot it is not uncommon to catch the University of Minnesota row team practicing. Anticipating this possibility I decided to put a field recorder and microphone in my bag. 
Old monophonic synthesizers have had all sorts of ways to sequence notes; from fairly complex programmable sequencers to simple arpeggiation. Arpeggiators are nice because they sequence the notes that are played on the keyboard in realtime, allowing the performer to improvise while taking advantage of the capabilities of the machine at the same time. Based on the speed of the LFO, the arpeggiator on the Sequential Circuits Pro-One will sequence all the notes in a chord in either ascending order, descending order, or both. Other synths offer a feature to play the notes in a random sequence. In this example the notes are played in ascending order. I cranked up the LFO and added some delay so that the notes blend into each other so you can hear the harmony in the sequence. With this technique you can sort of simulate polyphony on a mono-synth. 
I programmed a couple of parameters in a VST synth to a controller, set a single note (C1) to play for three minutes and eleven seconds, then recorded the automation. The parameters I was manipulating in real time were the shape of the waveform and the frequency. After recording the automation I added a bit of compression, a nice slow chorus to give it a left to right sweep, and a short delay with a lot of feedback for some added atmosphere. The image is just a snapshot from my photos and has nothing to do with the sound, but you’ve gotta love stick figure warning messages. 
To make this 4:15 minutes long recording I found a creaky spot on the wooden floors of my house, rocked back and forth in place and pointed a stereo mic at my feet. I remember doing this years ago, but thought I’d give it another go using a faster sampling rate and bit depth so the quality is not reduced as much when pitched down. Before making the recording I set the sampling rate and bit depth to 96kHz and 24 bit. I pitched it down two octaves and then normalized the results before rendering the final output.