Output much shorter than input #2
Description
Even with more sensitive parameters (100ms, -10dB) per #1, I'm still only getting a fraction of my input audio back out. My wavs directory is 199MB (22.5khz mono 16bit PCM wavs) but my wavs_split_final comes out to 36MB. My half-hour input file is split into 25 chunks, and an eight-minute one has only 10 chunks.
That's not enough information I realize to probably give much help, but I don't have time to debug just right now. So I figured I would at least open the issue for discussion (i.e., suggestions for what to try), and also so a fix can be documented for others who might have a similar issue in the future.