Nedko Arnaudov wrote: > ears and also caused everything (the mixer output) to silence out (I > guess adding normal sample value to a NaN produces another NaN). > No need to guess :) http://www.validlab.com/goldberg/paper.pdf R