Long empty chunks before actual speech

[tonytorm]

Really often the segments start with a lot of silence, the ending is right but the start is not, how could i fix that only using whisper.cpp?let's say I don't wanta deal with any VAD or more processing prior to feeding the buffers to whisper

[tonytorm]

sorted