Divide an audio stream into tiny segments, using the signal's zero-crossings as segment boundaries, and discard a fraction of them (i.e. replace them with silence of the same length). The technique was described by Trevor Wishart in a lecture.
Parameters: the filter drops drop out of outof chunks. mode can be
1 to drop chunks in a simple deterministic fashion (e.g. always dropping the first
30 out of a set of
40 segments), or
2 to drop chunks randomly but in an appropriate proportion.
number of waves to drop within each region of
number of waves within which