Shapemaker a tool designed to help cut shapes from a corpus of tiny sound files (perhaps made automatically with FluCoMa or Aubio). The tool was designed for students on the MSc Sound Design and Audio Crafting course at the University of Edinburgh.
This tool was built for sound designers and composers, by a sound designer and composer, not by computer programmers. When you understand the principles of how it works, it's your design skills that will be key to getting the best out of it. Your decsisions on timing, sound naming and structuring end up having direct consequences on what kinds of sounds come out, though you may of course begin playing with the tool in a fairly arbitrary way by simply loading a folder of files in and messing around.
There is also the possiblity to perform with the shapeMaker using a live microphone, or to use a sound file to trigger playback through a curve of sound files. At present this functionality is pretty basic, but could be improved.
Crucially, shapeMaker does not use any clever audio analysis tools to understand your audio better, nor does it try to understand your sound for you. Instead, you organise it into folders and use your ears, organisation, taste and design skills to make decent sound outputs in the end. The idea is that you might generate lots of audio from this and then mix things together later.
If you're looking for tools that will help you at an algorithmic level, then watch developments at FluCoMa and further developments by amazing python programmers such as Chris Tralie, James Bradbury or Brian McFee.
Features are developing with this as students request or design them.
shapeMaker detects if a file is two channel or mono. When mono, the sound is panned according to the path you design, if it is stereo, the line is bypassed and the stereo sound is played.