I'm trying to detect answering machine tones.... so I don't care
about the pitch... just that the pitch remains near constant for some
Are there any shortcuts to avoid full decode and still get things
like dominant frequency / excitation level for a sample? Or other
ways of detecting possible tones?
Looking for hints / help you all might have...
but also, if there is a developer available (ASAP) that could wrap
this into a C library, we would be very interested. (We have C
algorithms for uncompressed PCM data you could start with.)