David Boothe wrote:
> 
> Either I am not using it correctly or there is a bug. 

Problem is that table f1 (which you use for noise
audio source), ends up at level zero:
    f1    0  32768    21  1      ; white noise

- you need to set the "level" parameter, ie
    f1    0  32768    21  1   1  ; white noise

This is somewhat counterintuitive behaviour - other 
Gens are scaled to amp 1 (one) unless told otherwise. 

Btw, I noted that the uniform Gen 21 gives positive-only 
values, is *this* right? - Gens should by default give 
normalized values (plus/minus max), not? - that's what 
the other gen21 distributions do...