Using audio whilst analysing

[jonathan-annett]

I'd like to capture audio in real time using the microphone/soundcard, and use it to drive a display immediately. I'd then like the audio buffers to be placed in a 30 second queue, and after 30 seconds, play the recorded audio back using the default playback device. This would be a continuous process, and would function like an audio delay.

[jonathan-annett]

What's the simplest way to get the presumably discarded buffers so they can be reused for later playback?