All, I was thinking of taking a defined channel's audio input, manipulating the audio, and feeding the result back to the caller as requested by dtmf. I was hoping to use C# and mod_managed. Assuming the functions can be called, in some way, using C#. Anthony gave a brief suggestion of reading raw audio data from switch_buffer, in a way similar to reading from a file. (Which I yet do not understand). Again I would need to be able to both read and write audio to a channel. Can anyone provide me a high level overview of audio buffering in FS. Are two audio buffers created, one for the audio input of a channel and one for Audio output? Are they blocking if not read? What portion of the code would I best reference to understand how this could be accomplished? Is the state of any such buffer(s) defined in channel variables? Etc. Bob