I’ve been researching audio libraries for Processing recently since I will soon be starting the development of a specialized music application for personal use. I considered using MaxMSP, but Processing seems to suit this project a bit better. If you’re not familiar with Processing, it is an IDE designed for designers, artist, musicians, or anyone interested in exploring new ideas. Although it is mostly used for visual projects there are several examples of music software, like Tiction, which I wrote about in an entry titled Sound For Dali’s Melting Clocks. One of the libraries I’m investigating is called jm-Etude. It’s very easy to implement and use, and makes a few of the features in jMusic, a Java music composition project, accessible in Processing. Here’s some audio from a quick sketch designed to create a random sequence of notes. I also randomized the durations from whole notes to sixteenths, excluding tuplets for the time being.
Randomized Pitch Durations
I have been using my Sequential Circuits Pro-One in the studio pretty often recently, so I decided to use it instead of of a newer keyboard for a recent performance. Unfortunately it did not work the way I had planned. Although I crave unpredictable behavior to stimulate the creative process, it is not exactly as pleasing during a live performance. The aging instrument did not stay in tune for very long and I was having trouble with the keyboard action. I imagine that it’s time for a thorough cleaning of the potentiometers and keyboard contacts. Despite these troubles, while scanning through clips recorded during the set, I came across this nice low frequency filter sweep and ran it through a ping pong delay. 
A variation of this clip is probably going to end up in a new track that Nils Westdal and I are working on. After a recent session Nils left his Boss OC-1 pedal in the studio for me to mess around with. The first thing I did was plug the Rhodes into it. The cool thing about the OC-1 is that if you play intervals into it it gets confused and randomly switches between notes. The results are unpredictable, but it’s also possible to get consistently unpredictable results. I know it sounds like an oxymoron, but duplicating certain intervals at specific dynamics allows you to mold the output into something usable. This is what I was going for in this example. I edited the recording to a suitable length, added some delay, and then automated a filter to manipulate the texture.
I time stretched this recording of my dog Tia barking on the PCM-D50 using the built in DPC or Digital Pitch Control. I don’t know why they call it pitch control when it doesn’t change the pitch, but rather time compresses or expands the audio by percentages, from -75% to +100%. The intent is for musicians to practice difficult passages, but I think it has a nice eerie, metallic effect on voice and other sounds. For comparison I’ve posted her barking with out the time expansion as well.
In order to get this example of feedback I ran my suitcase model Rhodes, which I confine to my studio, through an outboard processor with some nice amp modeling and cranked up the gain. For the processing I used an eleven year old Yamaha A3000 sampler. The A3000 allows you to edit and apply processing to an incoming signal and it has some pretty nice sounding effects. Once I had adjusted the processing I held down the sustain pedal on the Rhodes and let the amplification do the rest. Tapping or gently knocking the instrument also produced some nice sounds. Here’s a snippet of the results.