I asked this question already to Richard at Facebook, but perhaps this place is the right one;

I have a long list of numbers that I like to convert to audio with Csound, at flexible frequency.

To play it at the sampling rate, I can see that I need to put them into a function table (GEN23 ?) and just play with an instrument (oscil ?). If oscil is right, then what is the frequency?

To play them slower than the sr, I need to stretch the data and fill it linearly extrapolating from sample to sample. Can Csound do this, or should I do this outside Csound, and then import with GEN 23?