The simple answer (to "why do that?") is that each of these two languages (Python and Csound) have their own unique strengths, and by combining them, you can do things that was not possible (or practical) with each one alone. The list of strengths and weaknesses can be made long, with lots of technical information, but I will try to give an example: Csound has the largest library of processing routines (opcodes) that I know of, this makes it a very flexible tool for a creative process involving audio processing. For example, if you can't get the sound you are searching for with a particular filter, you can always try to replace it with one of the other 30+ filters in Csound (do check the number of filters available, I did not count, but there's a lot of them available). Python is a good language for "general purpose programming", if you could call it that. For example, things like handling lists of note numbers constituting a melody or a chord. You would also want to use it for organization or batch processing, e.g. """process all the audio files in this folder, using this filter, but use this other filter if the filename has the string "vocal" in it""" As an example, I use Python and Csound together (both ways, pycall opcodes and csoundAPI) to do realtime algorithmic improvisation, letting the software search, sort and recombine the material I provide live (audio and midi). This allows for musical interaction between performer/composer and software/composer. For an example of music, see: http://oeyvind.teks.no/results/ArtisticDocBrandtsegg.htm#_Toc187835717 I don't know if you will find it "provocative", but it is an example of a way of performing music enabled by the use of these tools (Csound and Python). If you are in a hurry, for your purposes you might start listening to the piece at 7 minutes 50 secs (for 10 minutes), and the piece starting at 52 minutes 50 secs (lasting for another 10 minutes) Oeyvind 2008/6/2, Jim Aikin : > > I'm aware that Python code can be embedded in an instr, and conversely that > Csound can be run from a Python interpreter such as Idle. My question is: > why do that? Can someone give me a couple of examples of provocative musical > outcomes that might arise from doing either of the above? >